From f5c84b6781e9450a81ba6868233ecc4edeb35c39 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Cle=CC=81ment=20Doumouro?= Date: Thu, 16 Jul 2026 17:35:43 +0200 Subject: [PATCH] feature(asr-workers,extract-worker,translation-worker): model cache --- datashare-python/datashare_python/config.py | 17 +- .../datashare_python/constants.py | 2 - .../datashare_python/dependencies.py | 16 - datashare-python/datashare_python/objects.py | 54 +-- datashare-python/datashare_python/utils.py | 71 +++- datashare-python/datashare_python/worker.py | 5 +- datashare-python/tests/test_dependencies.py | 91 +---- datashare-python/tests/test_utils.py | 45 +++ workers/asr-worker/asr_worker/activities.py | 130 ++++--- workers/asr-worker/asr_worker/config.py | 26 +- workers/asr-worker/asr_worker/dependencies.py | 56 ++- workers/asr-worker/asr_worker/objects.py | 19 +- workers/asr-worker/asr_worker/workflows.py | 6 +- workers/asr-worker/pyproject.toml | 6 +- workers/asr-worker/tests/test_activities.py | 11 +- workers/asr-worker/tests/test_workflows.py | 2 +- workers/asr-worker/uv.dist.lock | 80 +++-- workers/asr-worker/uv.lock | 52 ++- workers/extract-worker/pyproject.toml | 2 +- workers/extract-worker/uv.dist.lock | 324 +++++++++--------- workers/extract-worker/uv.lock | 314 +++++++++-------- workers/translation-worker/tests/conftest.py | 19 +- .../tests/test_activities.py | 25 +- .../tests/test_hunyuan_translator.py | 71 +--- .../translation_worker/activities.py | 87 +++-- .../translation_worker/config.py | 39 ++- .../translation_worker/constants.py | 42 --- .../translation_worker/dependencies.py | 61 ++-- .../translation_worker/objects.py | 6 - .../translation_worker/processors.py | 22 +- .../translation_worker/sentence_splitters.py | 4 +- .../translation_worker/translators/argos.py | 4 +- .../translation_worker/translators/hunyuan.py | 33 +- .../translation_worker/workflows.py | 22 +- workers/translation-worker/uv.lock | 2 +- 35 files changed, 897 insertions(+), 869 deletions(-) diff --git a/datashare-python/datashare_python/config.py b/datashare-python/datashare_python/config.py index 14b0b38a..12cd9507 100644 --- a/datashare-python/datashare_python/config.py +++ b/datashare-python/datashare_python/config.py @@ -13,7 +13,7 @@ from .objects import BaseModel from .task_client import DatashareTaskClient from .types_ import TemporalClient -from .utils import PYDANTIC_DATA_CONVERTER +from .utils import PYDANTIC_DATA_CONVERTER, SharedResources, close_cm_callback _ALL_LOGGERS = [datashare_python.__name__] @@ -94,6 +94,19 @@ class LoggingConfig(BaseModel): loggers: dict[str, LogLevel] +class ResourceCacheConfig(BaseModel): + size: int = 1 + exit_context_managers: bool = True + + def to_resource_cache(self) -> SharedResources: + eviction_callback = None + if self.exit_context_managers: + eviction_callback = close_cm_callback + return SharedResources( + cache_size=self.size, eviction_callback=eviction_callback + ) + + _DEFAULT_LOGGERS = {datashare_python.__name__: "INFO"} _DEFAULT_LOGGING_CONFIG = LoggingConfig( format=LogFormat.DEFAULT, loggers=_DEFAULT_LOGGERS @@ -109,7 +122,7 @@ class WorkerConfig(ICIJSettings, BaseModel): elasticsearch: ESClientConfig = ESClientConfig() temporal: TemporalClientConfig = TemporalClientConfig() - max_concurrent_io_activities: int = 5 + max_concurrent_activities: int = 5 docs_root: Path | None = None artifacts_root: Path | None = None diff --git a/datashare-python/datashare_python/constants.py b/datashare-python/datashare_python/constants.py index 091f16e0..1cca521a 100644 --- a/datashare-python/datashare_python/constants.py +++ b/datashare-python/datashare_python/constants.py @@ -14,5 +14,3 @@ MANIFEST_JSON = "manifest.json" TIKA_METADATA_RESOURCENAME = "tika_metadata_resourcename" - -DEFAULT_SHARED_RESOURCES_SIZE = 1 diff --git a/datashare-python/datashare_python/dependencies.py b/datashare-python/datashare_python/dependencies.py index 776fa472..c1ad3421 100644 --- a/datashare-python/datashare_python/dependencies.py +++ b/datashare-python/datashare_python/dependencies.py @@ -12,7 +12,6 @@ from .config import LogLevel, WorkerConfig from .exceptions import DependencyInjectionError from .logging_ import setup_worker_loggers -from .objects import Shared from .task_client import DatashareTaskClient from .types_ import ContextManagerFactory, TemporalClient @@ -24,7 +23,6 @@ TASK_CLIENT: ContextVar[DatashareTaskClient] = ContextVar("task_client") TEMPORAL_CLIENT: ContextVar[TemporalClient] = ContextVar("temporal_client") WORKER_CONFIG: ContextVar[WorkerConfig] = ContextVar("worker_config") -SHARED: ContextVar[Shared] = ContextVar("shared") def set_event_loop(event_loop: AbstractEventLoop) -> None: @@ -101,20 +99,6 @@ def lifespan_temporal_client() -> TemporalClient: raise DependencyInjectionError("temporal client") from e -# Setup shared resources -async def set_shared_resources(shared: Shared) -> Shared: - SHARED.set(shared) - return shared - - -# Return shared resources -def lifespan_shared_resources() -> Shared: - try: - return SHARED.get() - except LookupError as e: - raise DependencyInjectionError("shared resources") from e - - @asynccontextmanager async def with_dependencies( dependencies: list[ContextManagerFactory], **kwargs diff --git a/datashare-python/datashare_python/objects.py b/datashare-python/datashare_python/objects.py index 46bad7f1..c985c2fc 100644 --- a/datashare-python/datashare_python/objects.py +++ b/datashare-python/datashare_python/objects.py @@ -1,9 +1,8 @@ import logging import os from abc import ABC -from asyncio import Lock from collections.abc import Awaitable, Callable -from dataclasses import InitVar, dataclass, field +from dataclasses import dataclass from datetime import UTC, datetime from enum import StrEnum, unique from io import BytesIO @@ -12,15 +11,12 @@ import langcodes from icij_common.registrable import Registrable -from lru import LRU from pydantic_core import PydanticCustomError, ValidationError, core_schema -from pydantic_core.core_schema import ( - PlainValidatorFunctionSchema, -) +from pydantic_core.core_schema import PlainValidatorFunctionSchema from pydantic_extra_types.language_code import LanguageName from temporalio import workflow -from .constants import DEFAULT_SHARED_RESOURCES_SIZE, TIKA_METADATA_RESOURCENAME +from .constants import TIKA_METADATA_RESOURCENAME with workflow.unsafe.imports_passed_through(): from icij_common.es import ( @@ -488,47 +484,3 @@ class TaskGroup: @classmethod def python(cls) -> Self: return cls(name="PYTHON") - - -@dataclass(frozen=True) -class Shared: - cache_size: InitVar[int] = DEFAULT_SHARED_RESOURCES_SIZE - eviction_callback: InitVar[Callable] = None - _resources: LRU = field(init=False, repr=False) - _lock: Lock = field(init=False, repr=False) - - def __post_init__(self, cache_size: int, eviction_callback: Callable) -> None: - object.__setattr__( - self, "_resources", LRU(cache_size, callback=eviction_callback) - ) - object.__setattr__(self, "_lock", Lock()) - - def get_resource( - self, key: str, default: Any = None, *, set_if_unavailable: bool = True - ) -> Any: - if key not in self._resources and set_if_unavailable: - self.set_resource(key, default) - - return self._resources.get(key, default) - - def set_resource(self, key: str, value: Any) -> None: - self._resources[key] = value - - def pop_resource(self, key: str, default: Any = None) -> Any: - return self._resources.pop(key, default) - - async def async_get_resource( - self, key: str, default: Any = None, *, set_if_unavailable: bool = True - ) -> Any: - if key not in self._resources and set_if_unavailable: - await self.async_set_resource(key, default) - - return self._resources.get(key, default) - - async def async_set_resource(self, key: str, value: Any) -> None: - async with self._lock: - self._resources[key] = value - - async def async_pop_resource(self, key: str, default: Any = None) -> Any: - async with self._lock: - return self._resources.pop(key, default) diff --git a/datashare-python/datashare_python/utils.py b/datashare-python/datashare_python/utils.py index 052e27d1..86f7d7ef 100644 --- a/datashare-python/datashare_python/utils.py +++ b/datashare-python/datashare_python/utils.py @@ -5,6 +5,7 @@ import json import os import shutil +import sys import threading import time from collections.abc import Callable, Coroutine, Generator, Iterable, Sequence @@ -15,10 +16,11 @@ from hashlib import sha256 from io import BytesIO from pathlib import Path -from typing import Any, ParamSpec, TypeVar +from typing import Any, ParamSpec, Self, TypeVar from uuid import uuid4 import temporalio +from lru import LRU from pydantic import ValidationError from temporalio import activity, workflow from temporalio.api.common.v1 import Payload @@ -40,7 +42,7 @@ ) from .constants import MANIFEST_JSON, METADATA_JSON -from .objects import DocArtifact, DocumentLocation, FilesystemDocument +from .objects import BaseModel, DocArtifact, DocumentLocation, FilesystemDocument from .types_ import RawAsyncProgressHandler DependencyLabel = str | None @@ -594,3 +596,68 @@ def to_payloads(self, values: Sequence[Any]) -> list[Payload]: PYDANTIC_DATA_CONVERTER = DataConverter( payload_converter_class=_PydanticPayloadConverter ) + + +class SharedResources: + def __init__( + self, + cache_size: int = 1, + eviction_callback: Callable[[Any, Any], None] | None = None, + ) -> None: + self._eviction_callback = eviction_callback + self._cache = LRU(cache_size, self._eviction_callback) + self._sentinel = object() + + def __enter__(self) -> Self: + return self + + def __exit__(self, exc_type, exc_val, exc_tb) -> None: # noqa: ANN001 + if self._eviction_callback is not None: + for k, v in self._cache.items(): + self._eviction_callback(k, v) + + def get_or_cache_resource( + self, key: str, default_factory: Callable[[], Any] + ) -> Any: + value = self._cache.get(key, self._sentinel) + if value is not self._sentinel: + return value + # Get the value first to be sure to return the right one in case of concurrent + # access. + # + # Additionally, the factory can be long to complete (that's one of the reason + # we want to cache its results). Because we don't lock the factory call, + # together with the cache update, in case of concurrent access, + # quicker factories will complete first and will stay longer in the cache. + # This is fine, the alternative is to lock factory + cache update to avoid + # concurrent access but this will mean waiting for the first factory call to + # complete, this can potentially imply longer waits than no getting the value + # from the cache + value = default_factory() + self._cache[key] = value + return value + + +def close_cm_callback(key: str, value: Any) -> None: # noqa: ARG001 + if hasattr(value, "__exit__"): + value.__exit__(*sys.exc_info()) + + +def enter_cm( + cm_factory: Callable[[], contextlib.AbstractContextManager], +) -> Callable[[], contextlib.AbstractContextManager]: # noqa: ARG001 + @wraps(cm_factory) + def factory() -> contextlib.AbstractContextManager: + cm = cm_factory() + cm.__enter__() + return cm + + return factory + + +def config_cache_key(config: BaseModel) -> str: + # Frozen model offer natural hash + if not config.__class__.model_config["frozen"]: + msg = f"{config.__class__} is not frozen, can't compute cache key" + raise ValueError(msg) + return config.__hash__() diff --git a/datashare-python/datashare_python/worker.py b/datashare-python/datashare_python/worker.py index 00bff7c9..5dec01b8 100644 --- a/datashare-python/datashare_python/worker.py +++ b/datashare-python/datashare_python/worker.py @@ -71,7 +71,7 @@ def datashare_worker( task_queue: str, # Scale horizontally be default for activities, each worker processes one activity # at a time - max_concurrent_io_activities: int = 10, + max_concurrent_activities: int = 1, sandboxed: bool = True, ) -> DatashareWorker: if workflows is None: @@ -91,7 +91,6 @@ def datashare_worker( ) logger.warning(_SEPARATE_IO_AND_CPU_ACTIVITIES) - max_concurrent_activities = max_concurrent_io_activities if isinstance(activity_executor, ThreadPoolExecutor): max_concurrent_activities = 1 if workflows: @@ -204,7 +203,7 @@ async def worker_context( workflows=workflows, activities=acts, task_queue=task_queue, - max_concurrent_io_activities=worker_config.max_concurrent_io_activities, + max_concurrent_activities=worker_config.max_concurrent_activities, sandboxed=sandboxed, ) async with worker: diff --git a/datashare-python/tests/test_dependencies.py b/datashare-python/tests/test_dependencies.py index 6c02b6e7..e8d16934 100644 --- a/datashare-python/tests/test_dependencies.py +++ b/datashare-python/tests/test_dependencies.py @@ -1,96 +1,7 @@ from typing import Any -from unittest.mock import MagicMock import pytest -from datashare_python.dependencies import ( - add_missing_args, - component_teardown, - lifespan_shared_resources, - set_shared_resources, -) -from datashare_python.exceptions import DependencyInjectionError -from datashare_python.objects import Shared - - -async def test_set_shared_resources_and_get() -> None: - shared = Shared() - shared.set_resource("k", "v") - returned = await set_shared_resources(shared) - assert returned is shared - assert lifespan_shared_resources() is shared - - -async def test_set_shared_resources_overwrites() -> None: - first = Shared() - second = Shared() - first.set_resource("k", "v1") - second.set_resource("k", "v2") - await set_shared_resources(first) - await set_shared_resources(second) - assert lifespan_shared_resources() is second - - -def test_shared_resources_raises_before_set() -> None: - with pytest.raises(DependencyInjectionError): - lifespan_shared_resources() - - -def test_shared_resources_field_is_frozen() -> None: - shared = Shared() - shared.set_resource("k", "v") - with pytest.raises(Exception): # noqa: B017, PT011 - shared._resources = {} - - -def test_get_resource_existing_key() -> None: - shared = Shared() - shared.set_resource("k", "v") - assert shared.get_resource("k") == "v" - - -def test_get_resource_missing_key_with_default() -> None: - shared = Shared() - assert shared.get_resource("missing", "default") == "default" - - -async def test_async_set_resource() -> None: - shared = Shared() - await shared.async_set_resource("k", "v") - assert shared.get_resource("k") == "v" - - -async def test_async_set_resource_overwrites() -> None: - shared = Shared() - shared.set_resource("k", "old") - await shared.async_set_resource("k", "new") - assert shared.get_resource("k") == "new" - - -async def test_async_pop_resource_missing_key_with_default() -> None: - shared = Shared() - assert await shared.async_pop_resource("missing", None) is None - - -def test_call_component_teardown_on_key_eviction() -> None: - shared = Shared(1, eviction_callback=component_teardown) - - first_resource = MagicMock() - second_resource = MagicMock() - - shared.set_resource("first", first_resource) - shared.set_resource("second", second_resource) - - assert shared._resources.keys() == ["second"] - - first_resource.__exit__.assert_called_once() - - -def test_dont_set_default_when_set_is_unavailable_is_false() -> None: - shared = Shared() - - shared.get_resource("first", default="isn't there", set_if_unavailable=False) - - assert "first" not in shared._resources +from datashare_python.dependencies import add_missing_args @pytest.mark.parametrize( diff --git a/datashare-python/tests/test_utils.py b/datashare-python/tests/test_utils.py index e8552feb..a3c9a265 100644 --- a/datashare-python/tests/test_utils.py +++ b/datashare-python/tests/test_utils.py @@ -7,6 +7,7 @@ from datetime import timedelta from pathlib import Path from typing import ClassVar +from unittest.mock import MagicMock import pytest from datashare_python.constants import MANIFEST_JSON @@ -20,6 +21,7 @@ from datashare_python.types_ import TemporalClient from datashare_python.utils import ( _LOCKED, + SharedResources, activity_defn, artifact_lock, positional_args_only, @@ -344,3 +346,46 @@ async def test_artifact_lock_in_process(tmp_path: Path) -> None: # Then completed_without_timeout = _LOCKED.wait(sleep_s / 2) assert not completed_without_timeout + + +def test_get_or_cache_resource_new_key() -> None: + # Given + shared = SharedResources() + new_key = "k" + + def factory() -> str: + return "cached" + + # When + from_cache = shared.get_or_cache_resource(new_key, factory) + # Then + assert from_cache == "cached" + + +def test_get_or_cache_resource_existing_key() -> None: + # Given + shared = SharedResources() + existing_key = "k" + + def factory() -> str: + return uuid.uuid4().hex + + # When + first = shared.get_or_cache_resource(existing_key, factory) + second = shared.get_or_cache_resource(existing_key, factory) + # Then + assert second == first + + +def test_get_or_cache_eviction_callback_is_called_on_exit() -> None: + # Given + mock = MagicMock() + eviction_callback = mock.evict + key = "k" + factory = lambda: "value" # noqa: E731 + shared = SharedResources(eviction_callback=eviction_callback) + # When + with shared: + shared.get_or_cache_resource(key, factory) + # Then + eviction_callback.assert_called_once_with(key, "value") diff --git a/workers/asr-worker/asr_worker/activities.py b/workers/asr-worker/asr_worker/activities.py index ba149bb0..0fc15887 100644 --- a/workers/asr-worker/asr_worker/activities.py +++ b/workers/asr-worker/asr_worker/activities.py @@ -7,13 +7,15 @@ from pathlib import Path from typing import Annotated, Any, Protocol, cast -from caul.objects import ASRResult, PreprocessedInput -from caul.tasks import ( +from caul_core import ( + ASRResult, InferenceRunner, - ParakeetPostprocessorConfig, - ParakeetPreprocessorConfig, + InferenceRunnerConfig, Postprocessor, + PostprocessorConfig, + PreprocessedInput, Preprocessor, + PreprocessorConfig, ) from datashare_python.dependencies import lifespan_worker_config from datashare_python.objects import ( @@ -30,7 +32,9 @@ ActivityWithProgress, activity_defn, activity_workdir, + config_cache_key, debuggable_name, + enter_cm, read_jsonl, safe_dir, symlink_embedded_document_to_workdir, @@ -63,10 +67,14 @@ SEARCH_AUDIOS_ACTIVITY, SUPPORTED_CONTENT_TYPES, ) -from .dependencies import lifespan_es_client +from .dependencies import ( + lifespan_es_client, + lifespan_inference_runner_cache, + lifespan_postprocessor_cache, + lifespan_preprocessor_cache, +) from .objects import ( ASRArgs, - InferenceRunnerConfig, Transcription, TranscriptionArtifact, TranscriptionManifestEntry, @@ -80,7 +88,6 @@ _INFERENCE_WEIGHT = 10 * _PREPROCESS_WEIGHT _LIST_OF_PATH_ADAPTER = TypeAdapter(list[Path]) -_INFERENCE_CONFIG_TYPE_ADAPTER = TypeAdapter(InferenceRunnerConfig) class ArtifactFactory(Protocol): @@ -122,28 +129,33 @@ def preprocess( self, audio_batch: Path, project: str, - config: ParakeetPreprocessorConfig, + config: PreprocessorConfig, *, progress: Annotated[ # noqa: ARG002 SyncProgressRateHandler | None, Weight(value=_PREPROCESS_WEIGHT) ] = None, ) -> list[Path]: - # TODO: this shouldn't be necessary, fix this bug + # Import caul.tasks to populate the Preprocessor registry + import caul.tasks # noqa: F401, PLC0415 + worker_config = cast(ASRWorkerConfig, lifespan_worker_config()) workdir = worker_config.workdir - # TODO: implement caching - preprocessor = Preprocessor.from_config(config) output_dir = activity_workdir(workdir, project) output_dir.mkdir(parents=True, exist_ok=True) audio_batch = workdir / audio_batch - with preprocessor: - batch_paths = preprocess_act( - preprocessor, - audio_batch, - worker_config=worker_config, - output_dir=output_dir, - ) - batches = [p.relative_to(workdir) for p in batch_paths] + preprocessor_factory = enter_cm(partial(Preprocessor.from_config, config)) + preprocessor_key = config_cache_key(config) + cache = lifespan_preprocessor_cache() + preprocessor = cache.get_or_cache_resource( + preprocessor_key, preprocessor_factory + ) + batch_paths = preprocess_act( + preprocessor, + audio_batch, + worker_config=worker_config, + output_dir=output_dir, + ) + batches = [p.relative_to(workdir) for p in batch_paths] return batches @activity_defn(name=RUN_INFERENCE_ACTIVITY) @@ -157,8 +169,9 @@ async def infer( AsyncProgressRateHandler | None, Weight(value=_INFERENCE_WEIGHT) ] = None, ) -> list[Path]: - # TODO: fix this temporal by, we shouldn't have to reload - config = _INFERENCE_CONFIG_TYPE_ADAPTER.validate_python(config) + # Import caul.tasks to populate the InferenceRunner registry + import caul.tasks # noqa: F401, PLC0415 + worker_config = cast(ASRWorkerConfig, lifespan_worker_config()) workdir = worker_config.workdir output_dir = activity_workdir(workdir, project) @@ -169,20 +182,22 @@ async def infer( progress = to_raw_async_progress( progress, max_progress=len(preprocessed_inputs) ) - inference_runner = InferenceRunner.from_config(config) logger.info("loading model %s", config.model) - with inference_runner: - logger.info( - "model loaded, starting inference on %s audio chunks !", - len(preprocessed_inputs), - ) - inference_res = infer_act( - inference_runner, - preprocessed_inputs, - output_dir=output_dir, - progress=progress, - ) - inference_res = [p.relative_to(workdir) async for p in inference_res] + runner_factory = enter_cm(partial(InferenceRunner.from_config, config)) + runner_key = config_cache_key(config) + cache = lifespan_inference_runner_cache() + inference_runner = cache.get_or_cache_resource(runner_key, runner_factory) + logger.info( + "model loaded, starting inference on %s audio chunks !", + len(preprocessed_inputs), + ) + inference_res = infer_act( + inference_runner, + preprocessed_inputs, + output_dir=output_dir, + progress=progress, + ) + inference_res = [p.relative_to(workdir) async for p in inference_res] return inference_res @activity_defn(name=POSTPROCESS_ACTIVITY) @@ -190,13 +205,16 @@ def postprocess( self, inference_results: list[Path], audio_batch: Path, - config: ParakeetPostprocessorConfig, + config: PostprocessorConfig, args: ASRArgs, *, progress: Annotated[ # noqa: ARG002 SyncProgressRateHandler | None, Weight(value=_BASE_WEIGHT) ] = None, ) -> int: + # Import caul.tasks to populate the Postprocessor‹ registry + import caul.tasks # noqa: F401, PLC0415 + worker_config = cast(ASRWorkerConfig, lifespan_worker_config()) workdir = worker_config.workdir audio_batch = workdir / audio_batch @@ -206,25 +224,29 @@ def postprocess( ASRResult.model_validate_json((workdir / p).read_text()) for p in inference_results ) - # TODO: implement caching - postprocessor = Postprocessor.from_config(config) - with postprocessor: - docs = ( - FilesystemDocument.model_validate(fs_doc) - for fs_doc in read_jsonl(audio_batch) - ) - doc_ids = [doc.id for doc in docs] - if progress is not None: - progress = to_raw_sync_progress(progress, max_progress=len(doc_ids)) - return postprocess_act( - postprocessor, - inference_results, - doc_ids, - args, - artifacts_root=artifacts_root, - event_loop=self._event_loop, - progress=progress, - ) + + docs = ( + FilesystemDocument.model_validate(fs_doc) + for fs_doc in read_jsonl(audio_batch) + ) + doc_ids = [doc.id for doc in docs] + if progress is not None: + progress = to_raw_sync_progress(progress, max_progress=len(doc_ids)) + postprocessor_factory = enter_cm(partial(Postprocessor.from_config, config)) + postprocessor_key = config_cache_key(config) + cache = lifespan_postprocessor_cache() + postprocessor = cache.get_or_cache_resource( + postprocessor_key, postprocessor_factory + ) + return postprocess_act( + postprocessor, + inference_results, + doc_ids, + args, + artifacts_root=artifacts_root, + event_loop=self._event_loop, + progress=progress, + ) async def search_audio_paths_act( diff --git a/workers/asr-worker/asr_worker/config.py b/workers/asr-worker/asr_worker/config.py index e717bb2c..191750b2 100644 --- a/workers/asr-worker/asr_worker/config.py +++ b/workers/asr-worker/asr_worker/config.py @@ -1,7 +1,13 @@ from pathlib import Path import datashare_python -from datashare_python.config import LogFormat, LoggingConfig, WorkerConfig +from datashare_python.config import ( + LogFormat, + LoggingConfig, + ResourceCacheConfig, + WorkerConfig, +) +from datashare_python.objects import BaseModel from pydantic import Field _DEFAULT_LOGGERS = {datashare_python.__name__: "INFO", __name__: "INFO"} @@ -10,6 +16,18 @@ ) +class ASRCache(BaseModel): + preprocessor: ResourceCacheConfig = ResourceCacheConfig( + size=1, exit_context_managers=True + ) + inference_runner: ResourceCacheConfig = ResourceCacheConfig( + size=1, exit_context_managers=True + ) + postprocessor: ResourceCacheConfig = ResourceCacheConfig( + size=1, exit_context_managers=True + ) + + class ASRWorkerConfig(WorkerConfig): logging: LoggingConfig = _DEFAULT_LOGGING_CONFIG @@ -17,6 +35,12 @@ class ASRWorkerConfig(WorkerConfig): artifacts_root: Path workdir: Path + # Set max concurrent activity to 1 to avoid parallel inference in practice, + # this must be set to 1 for the inference worker and > 1 for the io worker + max_concurrent_activities: int = Field(frozen=True, default=1) + + cache: ASRCache = Field(default_factory=ASRCache) + def audios_root(self) -> Path: return self.docs_root diff --git a/workers/asr-worker/asr_worker/dependencies.py b/workers/asr-worker/asr_worker/dependencies.py index 7139818b..5720cc4e 100644 --- a/workers/asr-worker/asr_worker/dependencies.py +++ b/workers/asr-worker/asr_worker/dependencies.py @@ -3,6 +3,7 @@ import platform from collections.abc import Generator from contextlib import contextmanager +from contextvars import ContextVar from datashare_python.dependencies import ( # noqa: F401 lifespan_es_client, @@ -10,9 +11,56 @@ set_es_client, set_worker_config, ) +from datashare_python.exceptions import DependencyInjectionError +from datashare_python.utils import SharedResources + +from asr_worker.config import ASRWorkerConfig logger = logging.getLogger(__name__) +_PREPROCESSORS: ContextVar[SharedResources] = ContextVar("preprocessor_cache") +_INFERENCE_RUNNERS: ContextVar[SharedResources] = ContextVar("inference_runner_cache") +_POSTPROCESSORS: ContextVar[SharedResources] = ContextVar("postprocessor_cache") + + +def set_preprocessor_cache(worker_config: ASRWorkerConfig) -> SharedResources: + cache = worker_config.cache.preprocessor.to_resource_cache() + _PREPROCESSORS.set(cache) + return cache + + +def lifespan_preprocessor_cache() -> SharedResources: + try: + return _PREPROCESSORS.get() + except LookupError as e: + raise DependencyInjectionError("preprocessor cache") from e + + +def set_inference_runner_cache(worker_config: ASRWorkerConfig) -> SharedResources: + cache = worker_config.cache.inference_runner.to_resource_cache() + _INFERENCE_RUNNERS.set(cache) + return cache + + +def lifespan_inference_runner_cache() -> SharedResources: + try: + return _INFERENCE_RUNNERS.get() + except LookupError as e: + raise DependencyInjectionError("inference runner cache") from e + + +def set_postprocessor_cache(worker_config: ASRWorkerConfig) -> SharedResources: + cache = worker_config.cache.postprocessor.to_resource_cache() + _POSTPROCESSORS.set(cache) + return cache + + +def lifespan_postprocessor_cache() -> SharedResources: + try: + return _POSTPROCESSORS.get() + except LookupError as e: + raise DependencyInjectionError("postprocessor cache") from e + @contextmanager def set_multiprocessing_start_method(**_) -> Generator[None, None, None]: @@ -31,7 +79,11 @@ def set_multiprocessing_start_method(**_) -> Generator[None, None, None]: REGISTRY = { - "asr.inference": [set_worker_config, set_multiprocessing_start_method], + "asr.inference": [ + set_worker_config, + set_multiprocessing_start_method, + set_inference_runner_cache, + ], "asr.io": [set_worker_config, set_es_client], - "asr.cpu": [set_worker_config], + "asr.cpu": [set_worker_config, set_preprocessor_cache, set_postprocessor_cache], } diff --git a/workers/asr-worker/asr_worker/objects.py b/workers/asr-worker/asr_worker/objects.py index b12a8cf0..3027a350 100644 --- a/workers/asr-worker/asr_worker/objects.py +++ b/workers/asr-worker/asr_worker/objects.py @@ -1,11 +1,10 @@ import math from collections import defaultdict from functools import cache -from typing import Annotated, Any, ClassVar, Self +from typing import Any, ClassVar, Self -from caul.asr_pipeline import ASRPipelineConfig -from caul.config import InferenceRunnerConfig as CaulInferenceRunnerConfig -from caul.objects import ASRLanguage, ASRModel, ASRResult +from caul_core import ASRPipelineConfig, ASRResult +from caul_core.objects import ASRLanguage, ASRModel from datashare_python.objects import ( ArtifactType, DatashareModel, @@ -13,17 +12,7 @@ ManifestEntry, TaskArgs, ) -from icij_common.pydantic_utils import make_enum_discriminator, tagged_union -from pydantic import Discriminator, Field, RootModel - -model_discriminator = make_enum_discriminator("model", ASRModel) -InferenceRunnerConfig = Annotated[ - tagged_union( - CaulInferenceRunnerConfig.__subclasses__(), lambda t: t.model.default.value - ), - Discriminator(model_discriminator), -] - +from pydantic import Field, RootModel DocumentSearchQuery = dict[str, Any] DocId = str diff --git a/workers/asr-worker/asr_worker/workflows.py b/workers/asr-worker/asr_worker/workflows.py index aa572c89..77e50df3 100644 --- a/workers/asr-worker/asr_worker/workflows.py +++ b/workers/asr-worker/asr_worker/workflows.py @@ -9,11 +9,11 @@ from pydantic import TypeAdapter from temporalio import workflow -from asr_worker.constants import ASR_WORKFLOW, TEN_MINUTES -from asr_worker.objects import ASRArgs, ASRResponse +from .constants import ASR_WORKFLOW, TEN_MINUTES +from .objects import ASRArgs, ASRResponse with workflow.unsafe.imports_passed_through(): - from asr_worker.activities import ASRActivities + from .activities import ASRActivities _ASR_INPUTS_TYPE_ADAPTER = TypeAdapter(ASRArgs) diff --git a/workers/asr-worker/pyproject.toml b/workers/asr-worker/pyproject.toml index 2cdc87a8..e164ade2 100644 --- a/workers/asr-worker/pyproject.toml +++ b/workers/asr-worker/pyproject.toml @@ -11,15 +11,13 @@ readme = "README.md" requires-python = ">=3.11.0, <3.13" dependencies = [ "datashare-python~=0.9.0", - "caul==0.5.8", + "caul-core==0.3.1", ] [project.scripts] asr-worker = "asr_worker.__main__:main" [project.optional-dependencies] -cli = [ -] cpu = [ "torch==2.10.0", "torchaudio==2.10.0", @@ -32,7 +30,7 @@ gpu = [ "torchcodec==0.10.0+cu129; sys_platform == 'linux'", ] inference = [ - "caul[nemo]==0.5.8", + "caul[nemo]==0.10.2", "kaldialign==0.9.3", "ml-dtypes==0.5.4", "numpy==2.3.0", diff --git a/workers/asr-worker/tests/test_activities.py b/workers/asr-worker/tests/test_activities.py index 97d8d930..c1cb6ebc 100644 --- a/workers/asr-worker/tests/test_activities.py +++ b/workers/asr-worker/tests/test_activities.py @@ -15,8 +15,15 @@ ) from asr_worker.config import ASRWorkerConfig from asr_worker.objects import ASRArgs, DocId, Transcription, TranscriptionManifestEntry -from caul.objects import ASRResult, InputMetadata, PreprocessedInput, PreprocessorOutput -from caul.tasks import InferenceRunner, Postprocessor, Preprocessor +from caul_core import ( + ASRResult, + InferenceRunner, + InputMetadata, + Postprocessor, + PreprocessedInput, + Preprocessor, + PreprocessorOutput, +) from datashare_python.conftest import TEST_PROJECT from datashare_python.objects import DocumentLocation, FilesystemDocument from datashare_python.utils import read_jsonl diff --git a/workers/asr-worker/tests/test_workflows.py b/workers/asr-worker/tests/test_workflows.py index 4cb2c0f5..1f91a9a0 100644 --- a/workers/asr-worker/tests/test_workflows.py +++ b/workers/asr-worker/tests/test_workflows.py @@ -17,7 +17,7 @@ TranscriptionManifestEntry, ) from asr_worker.workflows import ASRWorkflow, TaskQueues -from caul.objects import ASRResult +from caul_core import ASRResult from datashare_python.conftest import TEST_PROJECT from datashare_python.objects import FilesystemDocument, ManifestEntryStatus from datashare_python.types_ import TemporalClient diff --git a/workers/asr-worker/uv.dist.lock b/workers/asr-worker/uv.dist.lock index e34fee6b..3cc99b20 100644 --- a/workers/asr-worker/uv.dist.lock +++ b/workers/asr-worker/uv.dist.lock @@ -269,21 +269,22 @@ wheels = [ [[package]] name = "caul" -version = "0.5.8" +version = "0.10.2" source = { registry = "https://pypi.org/simple" } dependencies = [ + { name = "caul-core", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "kaldialign", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "langcodes", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "librosa", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "tiktoken", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "typer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/fe/26/dd58fe1d199abbbe352d87ec3018b358f5061ada2a89f87e88f37303977a/caul-0.5.8.tar.gz", hash = "sha256:6b28b8f1a6a7372c6106183ee430c7d438d0866ab301c324eeaa6406083db286", size = 826594, upload-time = "2026-05-18T10:47:10.049Z" } +sdist = { url = "https://files.pythonhosted.org/packages/7d/bb/1808df4f36fa29a6f7dff855557ec3db89eb694397de4986576141f435ad/caul-0.10.2.tar.gz", hash = "sha256:50581b9a97f2e96cb395dd03dc7cc47cf3f1b57c06bae7cc9a3a96b5790ddb72", size = 248621, upload-time = "2026-07-16T10:08:52.018Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/a8/c8/aaa4ed76b05d311f1bf4ae5c1cb5ed8092ded768fb0c6cd611ec6087b0cc/caul-0.5.8-py3-none-any.whl", hash = "sha256:9d99d713a410194848f56227bb752f03fe4fc42b2acf3828783b5e2c1ab9c415", size = 37465, upload-time = "2026-05-18T10:47:11.016Z" }, + { url = "https://files.pythonhosted.org/packages/3a/8b/54b18a5185e4b895321e62e28d710472b429bdee574a4b84be3fbd62a8c5/caul-0.10.2-py3-none-any.whl", hash = "sha256:21b9f24f6b743956e14ac2eec1925a4e22602d5e0b9447e439c8a370009184c8", size = 39296, upload-time = "2026-07-16T10:08:51.056Z" }, ] [package.optional-dependencies] @@ -295,6 +296,19 @@ nemo = [ { name = "torch", version = "2.10.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-20-datashare-asr-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] +[[package]] +name = "caul-core" +version = "0.3.1" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/25/c1/165544ebbc1c4fdb8ec863a09e9d026cc5d77883f5ef3ec3912037514b3c/caul_core-0.3.1.tar.gz", hash = "sha256:30797791726c3c4a36e403f0a1cf18aecce42620fcfed1fe7fd72d00dae26476", size = 21880, upload-time = "2026-07-16T09:17:34.959Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/41/df/53a0fc06ccf7dbd5a5ce4514ac3ce553b7ec56b3c078e668ea5ea0f5feef/caul_core-0.3.1-py3-none-any.whl", hash = "sha256:97c5f46ea2f2cd245aa9efce53c0e6c1e0af6006e2031a8150ecb867d7075743", size = 12676, upload-time = "2026-07-16T09:17:34.137Z" }, +] + [[package]] name = "certifi" version = "2026.4.22" @@ -486,7 +500,7 @@ wheels = [ name = "datashare-asr-worker" source = { editable = "." } dependencies = [ - { name = "caul", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "caul-core", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "datashare-python", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] @@ -529,8 +543,8 @@ dev = [ [package.metadata] requires-dist = [ - { name = "caul", specifier = "==0.5.8" }, - { name = "caul", extras = ["nemo"], marker = "extra == 'inference'", specifier = "==0.5.8" }, + { name = "caul", extras = ["nemo"], marker = "extra == 'inference'", specifier = "==0.10.2" }, + { name = "caul-core", specifier = "==0.3.1" }, { name = "cuda-bindings", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==12.9.4" }, { name = "datashare-python", specifier = "~=0.9.0" }, { name = "kaldialign", marker = "extra == 'inference'", specifier = "==0.9.3" }, @@ -544,7 +558,7 @@ requires-dist = [ { name = "torchcodec", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==0.10.0+cu129", index = "https://download.pytorch.org/whl/cu129", conflict = { package = "datashare-asr-worker", extra = "gpu" } }, { name = "torchcodec", marker = "extra == 'cpu'", specifier = "==0.10.0", index = "https://download.pytorch.org/whl/cpu", conflict = { package = "datashare-asr-worker", extra = "cpu" } }, ] -provides-extras = ["cli", "cpu", "gpu", "inference", "preprocessing"] +provides-extras = ["cpu", "gpu", "inference", "preprocessing"] [package.metadata.requires-dev] dev = [ @@ -563,7 +577,7 @@ dev = [ [[package]] name = "datashare-python" -version = "0.9.5" +version = "0.9.9" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "aiohttp", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, @@ -572,8 +586,6 @@ dependencies = [ { name = "icij-common", extra = ["elasticsearch"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "langcodes", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "lru-dict", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, - { name = "nest-asyncio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, - { name = "orjson", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "python-json-logger", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pyyaml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, @@ -581,9 +593,9 @@ dependencies = [ { name = "tomlkit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "typer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/cc/6a/cb3c7c3073573c6ad0cf56cd6c6b6c97dbd1dbd6392ec0314e78865c9dee/datashare_python-0.9.5.tar.gz", hash = "sha256:9ff757584b46bf95af74a1c890734d3233f02f54724806ec9b346a04c39e311a", size = 320645, upload-time = "2026-06-23T12:11:01.145Z" } +sdist = { url = "https://files.pythonhosted.org/packages/1e/92/d39036df8af1ab2195abffc1b110b3ae73b7492b4817dec6e2d78a171a0c/datashare_python-0.9.9.tar.gz", hash = "sha256:e4f78f38bdb84703d6869bd13a767bd6a3e329409999a096aac8eb5119bf44a4", size = 315745, upload-time = "2026-07-15T15:35:15.519Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/1b/27/826ac48804b0bda7a35260fdc925ff25fda3fcc527da32c4c484461ed8d2/datashare_python-0.9.5-py3-none-any.whl", hash = "sha256:c1dce1b981342dc9a3e80d3bfd3246d4e774d55c116a37321e31ac2b39d4ae08", size = 326631, upload-time = "2026-06-23T12:10:59.749Z" }, + { url = "https://files.pythonhosted.org/packages/2d/73/68356d617d630af7154b986b528c531295e34fc096d5a3aef663b8cba78f/datashare_python-0.9.9-py3-none-any.whl", hash = "sha256:ff99059a87c6c0da9c0098dc3eaf782b54db22b5e2d8317ec534023b39a130a9", size = 322639, upload-time = "2026-07-15T15:35:14.434Z" }, ] [[package]] @@ -843,16 +855,16 @@ wheels = [ [[package]] name = "icij-common" -version = "0.8.2" +version = "0.8.3" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "aiostream", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/e0/51/91d85cc23e275bb51a5cae47873af9a7160f62797b25663d902cf38e6ab7/icij_common-0.8.2.tar.gz", hash = "sha256:7db68266d8facb43142131d81e998cd74c6ae73508456743e00df43a15cc2995", size = 15937, upload-time = "2026-04-07T12:13:39.774Z" } +sdist = { url = "https://files.pythonhosted.org/packages/6a/8d/4b2b9bb7dcac24d8bf514ccf2da6465ca3cd9b4c4f4bde0a1c97e681d595/icij_common-0.8.3.tar.gz", hash = "sha256:009dda7c1d688ecf7705cf88517deb3d0abe02300b42c8c433eb36516462fe75", size = 16161, upload-time = "2026-07-03T09:43:50.837Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/15/5a/99d123230023124ba4ec67765feb2a04b656ae1f473a14f6b29cc54b02bf/icij_common-0.8.2-py3-none-any.whl", hash = "sha256:c5a1a3ccd54e7bf53acad293a0af50ab15156b225791b43383197b896bae955f", size = 20565, upload-time = "2026-04-07T12:13:38.709Z" }, + { url = "https://files.pythonhosted.org/packages/1c/8e/99879d623fdf4e2371e44135dcf690d7f43b921616e4c918633fe4cece48/icij_common-0.8.3-py3-none-any.whl", hash = "sha256:5214cbb73dca364cca16005f5fac3c2f3944acab6dea269a68d26f631e39a348", size = 21030, upload-time = "2026-07-03T09:43:49.934Z" }, ] [package.optional-dependencies] @@ -1985,22 +1997,6 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/ac/24/7c731839566d30dc70556d9824ef17692d896c15e3df627bce8c16f753e1/optuna-4.8.0-py3-none-any.whl", hash = "sha256:c57a7682679c36bfc9bca0da430698179e513874074b71bebedb0334964ab930", size = 419456, upload-time = "2026-03-16T04:59:56.977Z" }, ] -[[package]] -name = "orjson" -version = "3.11.9" -source = { registry = "https://pypi.org/simple" } -sdist = { url = "https://files.pythonhosted.org/packages/7e/0c/964746fcafbd16f8ff53219ad9f6b412b34f345c75f384ad434ceaadb538/orjson-3.11.9.tar.gz", hash = "sha256:4fef17e1f8722c11587a6ef18e35902450221da0028e65dbaaa543619e68e48f", size = 5599163, upload-time = "2026-05-06T15:11:08.309Z" } -wheels = [ - { url = "https://files.pythonhosted.org/packages/1e/51/3fb9e65ae76ee97bd611869a503fa3fc0a6e81dd8b737cf3003f682df7ff/orjson-3.11.9-cp311-cp311-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:f01c4818b3fc9b0da8e096722a84318071eaa118df35f6ed2344da0e73a5444f", size = 228522, upload-time = "2026-05-06T15:09:35.362Z" }, - { url = "https://files.pythonhosted.org/packages/16/fa/9d54b07cb3f3b0bfd57841478e42d7a0ece4a9f49f9907eecf5a45461687/orjson-3.11.9-cp311-cp311-macosx_15_0_arm64.whl", hash = "sha256:3ebca4179031ee716ed076ffadc29428e900512f6fccee8614c9983157fcf19c", size = 128463, upload-time = "2026-05-06T15:09:37.063Z" }, - { url = "https://files.pythonhosted.org/packages/49/bd/360686f39348aa88827cb6fbf7dc606fd41c831a35235e1abf1db8e3a9e6/orjson-3.11.9-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:147302878da387104b66bb4a8b0227d1d487e976ce41a8501916161072ed87b1", size = 133971, upload-time = "2026-05-06T15:09:45.239Z" }, - { url = "https://files.pythonhosted.org/packages/b6/8a/4081492586d75b073d60c5271a8d0f05a0955cabf1e34c8473f6fcd84235/orjson-3.11.9-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:63e0efbc991250c0b3143488fa57d95affcabbfc63c99c48d625dd37779aafe2", size = 136959, upload-time = "2026-05-06T15:09:51.311Z" }, - { url = "https://files.pythonhosted.org/packages/16/6d/11867a3ffa3a3608d84a4de51ef4dd0896d6b5cc9132fbe1daf593e677bc/orjson-3.11.9-cp312-cp312-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:9ef6fe90aadef185c7b128859f40beb24720b4ecea95379fc9000931179c3a49", size = 228515, upload-time = "2026-05-06T15:09:57.265Z" }, - { url = "https://files.pythonhosted.org/packages/24/75/05912954c8b288f34fcf5cd4b9b071cb4f6e77b9961e175e56ebb258089f/orjson-3.11.9-cp312-cp312-macosx_15_0_arm64.whl", hash = "sha256:e5c9b8f28e726e97d97696c826bc7bea5d71cecd63576dba92924a32c1961291", size = 128409, upload-time = "2026-05-06T15:09:59.063Z" }, - { url = "https://files.pythonhosted.org/packages/0e/a4/82b7a2fe5d8a67a59ed831b24d59a3d46ea7d207b66e1602d376541d94a6/orjson-3.11.9-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:be4fa4f0af7fa18951f7ab3fc2148e223af211bf03f59e1c6034ec3f97f21d61", size = 134014, upload-time = "2026-05-06T15:10:08.213Z" }, - { url = "https://files.pythonhosted.org/packages/df/e5/4d2d8af06f788329b4f78f8cc3679bb395392fcaa1e4d8d3c33e85308fa4/orjson-3.11.9-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:71e63adb0e1f1ed5d9e168f50a91ceb93ae6420731d222dc7da5c69409aa47aa", size = 136943, upload-time = "2026-05-06T15:10:14.405Z" }, -] - [[package]] name = "overrides" version = "7.7.0" @@ -2998,6 +2994,26 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/32/d5/f9a850d79b0851d1d4ef6456097579a9005b31fea68726a4ae5f2d82ddd9/threadpoolctl-3.6.0-py3-none-any.whl", hash = "sha256:43a0b8fd5a2928500110039e43a5eed8480b918967083ea48dc3ab9f13c4a7fb", size = 18638, upload-time = "2025-03-13T13:49:21.846Z" }, ] +[[package]] +name = "tiktoken" +version = "0.13.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/e4/e5/5f3cb2159769d0f4324c0e9e87f9de3c4b1cd45848a96b2eb3566ad5ca77/tiktoken-0.13.0.tar.gz", hash = "sha256:c9435714c3a84c2319499de9a300c0e604449dd0799ff246458b3bb6a7f433c1", size = 38986, upload-time = "2026-05-15T04:51:27.153Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/1a/4c/1bc81f4cd53e827c4ee67ca951b5935724716049452d8dfa09b8b82372bb/tiktoken-0.13.0-cp311-cp311-macosx_10_12_x86_64.whl", hash = "sha256:7bfe1849caa65d1e1d9871817170ec497bbb7984e182012e1bdce72f66608cdb", size = 1036353, upload-time = "2026-05-15T04:50:21.757Z" }, + { url = "https://files.pythonhosted.org/packages/75/91/10b9c7076bc02c246c853201fdbbe300a4b8c5ed7b84c25f7403f4e32655/tiktoken-0.13.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:91c180fe255bd5a86d8316210d2833a1d4d33d026cd86a67812f4773743c8d26", size = 984644, upload-time = "2026-05-15T04:50:23.256Z" }, + { url = "https://files.pythonhosted.org/packages/f9/39/fe42ad00de01a8c4a49ad8649a2c8a316835a9cad5961b11d21eac0020a5/tiktoken-0.13.0-cp311-cp311-manylinux_2_28_x86_64.whl", hash = "sha256:36217497eaffc158607a3b26f065300db2aefd43b115263f3b9688ce38146173", size = 1138253, upload-time = "2026-05-15T04:50:25.505Z" }, + { url = "https://files.pythonhosted.org/packages/9d/03/cd0cba295522b91eb55c6b2704f1df895f8226cfe60ab10d4d51d0cc9e69/tiktoken-0.13.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:5d48843bee149630eb735a99e1f4a85b47308d21868ea63163f6e87768d3cfed", size = 1241265, upload-time = "2026-05-15T04:50:28.815Z" }, + { url = "https://files.pythonhosted.org/packages/85/8e/144bde4e01df66b34bb865557c7cd754ed08b036217ebd79c9db5e9048a9/tiktoken-0.13.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:32ac870a806cfb260a02d0cb70426aef02e038297f8ad50df5040bb5af360791", size = 1034888, upload-time = "2026-05-15T04:50:31.579Z" }, + { url = "https://files.pythonhosted.org/packages/36/18/d4ac9d20956cdebca04841316660ed584c2fecdc2b81722a28bc7ad3b1e4/tiktoken-0.13.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:4d9980f11429ed2d737c463bb1fb78cf330caa026adf002f714aced7849a687b", size = 982970, upload-time = "2026-05-15T04:50:32.961Z" }, + { url = "https://files.pythonhosted.org/packages/34/de/2ca96b07a82d972b74fe4b46de055b79c904e45c7eab699354a0bfa697dc/tiktoken-0.13.0-cp312-cp312-manylinux_2_28_x86_64.whl", hash = "sha256:a116178fa7e1b4065bff05214360373a65cac22f965be7b3f73d00a0dbfe7649", size = 1136523, upload-time = "2026-05-15T04:50:35.782Z" }, + { url = "https://files.pythonhosted.org/packages/a1/d0/1f8578c45b2f24759b46f0b50d31878c63c73e6bf0f2227e10ec5c5408dc/tiktoken-0.13.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:95097e4f89b06403976e498abf61a0ee73a7497e73fb599cb211d8197a054d91", size = 1240069, upload-time = "2026-05-15T04:50:38.221Z" }, +] + [[package]] name = "tokenizers" version = "0.22.2" diff --git a/workers/asr-worker/uv.lock b/workers/asr-worker/uv.lock index e49ee42e..ed43a2b1 100644 --- a/workers/asr-worker/uv.lock +++ b/workers/asr-worker/uv.lock @@ -269,21 +269,22 @@ wheels = [ [[package]] name = "caul" -version = "0.5.8" +version = "0.10.2" source = { registry = "https://pypi.org/simple" } dependencies = [ + { name = "caul-core", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "kaldialign", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "langcodes", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "librosa", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "tiktoken", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "typer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/fe/26/dd58fe1d199abbbe352d87ec3018b358f5061ada2a89f87e88f37303977a/caul-0.5.8.tar.gz", hash = "sha256:6b28b8f1a6a7372c6106183ee430c7d438d0866ab301c324eeaa6406083db286", size = 826594, upload-time = "2026-05-18T10:47:10.049Z" } +sdist = { url = "https://files.pythonhosted.org/packages/7d/bb/1808df4f36fa29a6f7dff855557ec3db89eb694397de4986576141f435ad/caul-0.10.2.tar.gz", hash = "sha256:50581b9a97f2e96cb395dd03dc7cc47cf3f1b57c06bae7cc9a3a96b5790ddb72", size = 248621, upload-time = "2026-07-16T10:08:52.018Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/a8/c8/aaa4ed76b05d311f1bf4ae5c1cb5ed8092ded768fb0c6cd611ec6087b0cc/caul-0.5.8-py3-none-any.whl", hash = "sha256:9d99d713a410194848f56227bb752f03fe4fc42b2acf3828783b5e2c1ab9c415", size = 37465, upload-time = "2026-05-18T10:47:11.016Z" }, + { url = "https://files.pythonhosted.org/packages/3a/8b/54b18a5185e4b895321e62e28d710472b429bdee574a4b84be3fbd62a8c5/caul-0.10.2-py3-none-any.whl", hash = "sha256:21b9f24f6b743956e14ac2eec1925a4e22602d5e0b9447e439c8a370009184c8", size = 39296, upload-time = "2026-07-16T10:08:51.056Z" }, ] [package.optional-dependencies] @@ -295,6 +296,19 @@ nemo = [ { name = "torch", version = "2.10.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-20-datashare-asr-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] +[[package]] +name = "caul-core" +version = "0.3.1" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/25/c1/165544ebbc1c4fdb8ec863a09e9d026cc5d77883f5ef3ec3912037514b3c/caul_core-0.3.1.tar.gz", hash = "sha256:30797791726c3c4a36e403f0a1cf18aecce42620fcfed1fe7fd72d00dae26476", size = 21880, upload-time = "2026-07-16T09:17:34.959Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/41/df/53a0fc06ccf7dbd5a5ce4514ac3ce553b7ec56b3c078e668ea5ea0f5feef/caul_core-0.3.1-py3-none-any.whl", hash = "sha256:97c5f46ea2f2cd245aa9efce53c0e6c1e0af6006e2031a8150ecb867d7075743", size = 12676, upload-time = "2026-07-16T09:17:34.137Z" }, +] + [[package]] name = "certifi" version = "2026.4.22" @@ -486,7 +500,7 @@ wheels = [ name = "datashare-asr-worker" source = { editable = "." } dependencies = [ - { name = "caul", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "caul-core", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, { name = "datashare-python", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, ] @@ -529,8 +543,8 @@ dev = [ [package.metadata] requires-dist = [ - { name = "caul", specifier = "==0.5.8" }, - { name = "caul", extras = ["nemo"], marker = "extra == 'inference'", specifier = "==0.5.8" }, + { name = "caul", extras = ["nemo"], marker = "extra == 'inference'", specifier = "==0.10.2" }, + { name = "caul-core", specifier = "==0.3.1" }, { name = "cuda-bindings", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==12.9.4" }, { name = "datashare-python", editable = "../../datashare-python" }, { name = "kaldialign", marker = "extra == 'inference'", specifier = "==0.9.3" }, @@ -544,7 +558,7 @@ requires-dist = [ { name = "torchcodec", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==0.10.0+cu129", index = "https://download.pytorch.org/whl/cu129", conflict = { package = "datashare-asr-worker", extra = "gpu" } }, { name = "torchcodec", marker = "extra == 'cpu'", specifier = "==0.10.0", index = "https://download.pytorch.org/whl/cpu", conflict = { package = "datashare-asr-worker", extra = "cpu" } }, ] -provides-extras = ["cli", "cpu", "gpu", "inference", "preprocessing"] +provides-extras = ["cpu", "gpu", "inference", "preprocessing"] [package.metadata.requires-dev] dev = [ @@ -563,7 +577,7 @@ dev = [ [[package]] name = "datashare-python" -version = "0.9.8" +version = "0.9.9" source = { editable = "../../datashare-python" } dependencies = [ { name = "aiohttp", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, @@ -3007,6 +3021,26 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/32/d5/f9a850d79b0851d1d4ef6456097579a9005b31fea68726a4ae5f2d82ddd9/threadpoolctl-3.6.0-py3-none-any.whl", hash = "sha256:43a0b8fd5a2928500110039e43a5eed8480b918967083ea48dc3ab9f13c4a7fb", size = 18638, upload-time = "2025-03-13T13:49:21.846Z" }, ] +[[package]] +name = "tiktoken" +version = "0.13.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-20-datashare-asr-worker-cpu' and extra == 'extra-20-datashare-asr-worker-gpu')" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/e4/e5/5f3cb2159769d0f4324c0e9e87f9de3c4b1cd45848a96b2eb3566ad5ca77/tiktoken-0.13.0.tar.gz", hash = "sha256:c9435714c3a84c2319499de9a300c0e604449dd0799ff246458b3bb6a7f433c1", size = 38986, upload-time = "2026-05-15T04:51:27.153Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/1a/4c/1bc81f4cd53e827c4ee67ca951b5935724716049452d8dfa09b8b82372bb/tiktoken-0.13.0-cp311-cp311-macosx_10_12_x86_64.whl", hash = "sha256:7bfe1849caa65d1e1d9871817170ec497bbb7984e182012e1bdce72f66608cdb", size = 1036353, upload-time = "2026-05-15T04:50:21.757Z" }, + { url = "https://files.pythonhosted.org/packages/75/91/10b9c7076bc02c246c853201fdbbe300a4b8c5ed7b84c25f7403f4e32655/tiktoken-0.13.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:91c180fe255bd5a86d8316210d2833a1d4d33d026cd86a67812f4773743c8d26", size = 984644, upload-time = "2026-05-15T04:50:23.256Z" }, + { url = "https://files.pythonhosted.org/packages/f9/39/fe42ad00de01a8c4a49ad8649a2c8a316835a9cad5961b11d21eac0020a5/tiktoken-0.13.0-cp311-cp311-manylinux_2_28_x86_64.whl", hash = "sha256:36217497eaffc158607a3b26f065300db2aefd43b115263f3b9688ce38146173", size = 1138253, upload-time = "2026-05-15T04:50:25.505Z" }, + { url = "https://files.pythonhosted.org/packages/9d/03/cd0cba295522b91eb55c6b2704f1df895f8226cfe60ab10d4d51d0cc9e69/tiktoken-0.13.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:5d48843bee149630eb735a99e1f4a85b47308d21868ea63163f6e87768d3cfed", size = 1241265, upload-time = "2026-05-15T04:50:28.815Z" }, + { url = "https://files.pythonhosted.org/packages/85/8e/144bde4e01df66b34bb865557c7cd754ed08b036217ebd79c9db5e9048a9/tiktoken-0.13.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:32ac870a806cfb260a02d0cb70426aef02e038297f8ad50df5040bb5af360791", size = 1034888, upload-time = "2026-05-15T04:50:31.579Z" }, + { url = "https://files.pythonhosted.org/packages/36/18/d4ac9d20956cdebca04841316660ed584c2fecdc2b81722a28bc7ad3b1e4/tiktoken-0.13.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:4d9980f11429ed2d737c463bb1fb78cf330caa026adf002f714aced7849a687b", size = 982970, upload-time = "2026-05-15T04:50:32.961Z" }, + { url = "https://files.pythonhosted.org/packages/34/de/2ca96b07a82d972b74fe4b46de055b79c904e45c7eab699354a0bfa697dc/tiktoken-0.13.0-cp312-cp312-manylinux_2_28_x86_64.whl", hash = "sha256:a116178fa7e1b4065bff05214360373a65cac22f965be7b3f73d00a0dbfe7649", size = 1136523, upload-time = "2026-05-15T04:50:35.782Z" }, + { url = "https://files.pythonhosted.org/packages/a1/d0/1f8578c45b2f24759b46f0b50d31878c63c73e6bf0f2227e10ec5c5408dc/tiktoken-0.13.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:95097e4f89b06403976e498abf61a0ee73a7497e73fb599cb211d8197a054d91", size = 1240069, upload-time = "2026-05-15T04:50:38.221Z" }, +] + [[package]] name = "tokenizers" version = "0.22.2" diff --git a/workers/extract-worker/pyproject.toml b/workers/extract-worker/pyproject.toml index 733b1541..11b860ce 100644 --- a/workers/extract-worker/pyproject.toml +++ b/workers/extract-worker/pyproject.toml @@ -10,7 +10,7 @@ readme = "README.md" requires-python = ">=3.13,<3.15" dependencies = [ "datashare-python~=0.9.0", - "extract-core==0.7.0", + "extract-core==0.7.1", "temporalio==1.23.0", ] diff --git a/workers/extract-worker/uv.dist.lock b/workers/extract-worker/uv.dist.lock index d9f3fd1e..bd9d8dbe 100644 --- a/workers/extract-worker/uv.dist.lock +++ b/workers/extract-worker/uv.dist.lock @@ -172,13 +172,13 @@ name = "anthropic" version = "0.46.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/d4/68/3b4c045edf6dc6933895e8f279cc77c7684874c8aba46a4e6241c8b147cf/anthropic-0.46.0.tar.gz", hash = "sha256:eac3d43271d02321a57c3ca68aca84c3d58873e8e72d1433288adee2d46b745b", size = 202191, upload-time = "2025-02-18T20:35:33.314Z" } wheels = [ @@ -271,7 +271,7 @@ name = "cffi" version = "2.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "pycparser", marker = "(implementation_name != 'PyPy' and platform_machine == 'x86_64' and sys_platform == 'linux') or (implementation_name != 'PyPy' and sys_platform == 'darwin') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pycparser", marker = "(implementation_name != 'PyPy' and platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/eb/56/b1ba7935a17738ae8453301356628e8147c79dbb825bcbc73dc7401f9846/cffi-2.0.0.tar.gz", hash = "sha256:44d1b5909021139fe36001ae048dbdde8214afa20200eda0f64c068cac5d5529", size = 523588, upload-time = "2025-09-08T23:24:04.541Z" } wheels = [ @@ -348,7 +348,7 @@ name = "cryptography" version = "49.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cffi", marker = "(platform_machine == 'x86_64' and platform_python_implementation != 'PyPy' and sys_platform == 'linux') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_python_implementation != 'PyPy' and sys_platform == 'darwin') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cffi", marker = "(platform_machine == 'x86_64' and platform_python_implementation != 'PyPy' and sys_platform == 'linux') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_python_implementation != 'PyPy' and sys_platform == 'darwin') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/1f/99/d1c90d6041656cc6ee229dc99cd67fd0cd5aec3c5f7d72fffc27cc750054/cryptography-49.0.0.tar.gz", hash = "sha256:f89660a348f4f78a92366240a61404e337586ef7f5909a2fef59ca88ef505493", size = 854345, upload-time = "2026-06-12T20:02:30.512Z" } wheels = [ @@ -597,7 +597,7 @@ dev = [ requires-dist = [ { name = "cuda-bindings", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==12.9.4" }, { name = "datashare-python", specifier = "~=0.9.0" }, - { name = "extract-core", specifier = "==0.7.0" }, + { name = "extract-core", specifier = "==0.7.1" }, { name = "extract-python", extras = ["docling", "marker"], marker = "extra == 'base'", specifier = "==0.7.2" }, { name = "extract-python", extras = ["mineru"], marker = "extra == 'mineru'", specifier = "==0.7.2" }, { name = "flash-attn", marker = "extra == 'flash-attn'", specifier = "==2.8.3.post1" }, @@ -628,7 +628,7 @@ dev = [ [[package]] name = "datashare-python" -version = "0.9.8" +version = "0.9.9" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "aiohttp", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, @@ -644,9 +644,9 @@ dependencies = [ { name = "tomlkit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "typer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/9f/3c/a8d28a86627a8967d65070500ed1999434aa556b918c32c0029e9d3a998a/datashare_python-0.9.8.tar.gz", hash = "sha256:8b8a56780446827b77032d4eed402c9733a8399df6f301322d2577fdb19255c9", size = 320930, upload-time = "2026-07-02T08:08:16.289Z" } +sdist = { url = "https://files.pythonhosted.org/packages/1e/92/d39036df8af1ab2195abffc1b110b3ae73b7492b4817dec6e2d78a171a0c/datashare_python-0.9.9.tar.gz", hash = "sha256:e4f78f38bdb84703d6869bd13a767bd6a3e329409999a096aac8eb5119bf44a4", size = 315745, upload-time = "2026-07-15T15:35:15.519Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/71/2f/7f7f6aa1f670b3ab7d6cd6bc862ced5a2e499a14fe35a8a6629f6e0b2c98/datashare_python-0.9.8-py3-none-any.whl", hash = "sha256:ec3595ffb45e7eb45f0556813be0f278a7c779584579dab435eb88ff05703734", size = 327002, upload-time = "2026-07-02T08:08:14.688Z" }, + { url = "https://files.pythonhosted.org/packages/2d/73/68356d617d630af7154b986b528c531295e34fc096d5a3aef663b8cba78f/datashare_python-0.9.9-py3-none-any.whl", hash = "sha256:ff99059a87c6c0da9c0098dc3eaf782b54db22b5e2d8317ec534023b39a130a9", size = 322639, upload-time = "2026-07-15T15:35:14.434Z" }, ] [[package]] @@ -900,19 +900,17 @@ wheels = [ [[package]] name = "extract-core" -version = "0.7.0" +version = "0.7.1" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "docling-slim", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "marker-pdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mineru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/e6/ea/1bbc3485d0570455ca874aa84b4f02c4b4d509123710fe1832c02acca08c/extract_core-0.7.0.tar.gz", hash = "sha256:09bfef043c84afb11dd0d50582044cee7c6a5d3cdf95788067d90904b737dd80", size = 138372, upload-time = "2026-06-25T09:25:17.672Z" } +sdist = { url = "https://files.pythonhosted.org/packages/4c/a3/7b4a72cb925ee84da756520675b428ac872dd9fc9962de2a0e80703ac7ea/extract_core-0.7.1.tar.gz", hash = "sha256:bd8eec934bf73749e912455f9d7818320ef73382e6f8382a93d502c9466fba53", size = 109907, upload-time = "2026-07-02T08:02:52.016Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/d4/41/2eb44fcd9b33d9a4da7d61691f2eea3ebbad565288601d1b8169d2f169f7/extract_core-0.7.0-py3-none-any.whl", hash = "sha256:51d7cb28fcc8173a34e20862a236354f5da35e2bd0156d0001650e22d6e655d3", size = 10414, upload-time = "2026-06-25T09:25:18.496Z" }, + { url = "https://files.pythonhosted.org/packages/ac/14/584c2c62830871464c475a7ea4117ce0df6fefb6fb31ac4df7838e6d0b13/extract_core-0.7.1-py3-none-any.whl", hash = "sha256:b360c958d4375c263e391bd1d7328d6786ec20d095d8758e86329548333a110d", size = 10382, upload-time = "2026-07-02T08:02:53.029Z" }, ] [[package]] @@ -957,9 +955,9 @@ name = "fast-langdetect" version = "0.2.5" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "fasttext-predict", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "robust-downloader", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fasttext-predict", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "robust-downloader", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/8d/17/2753b3d280e16f594e7a5d735568c021a355cd1edd3827f26cda5c9fd460/fast_langdetect-0.2.5.tar.gz", hash = "sha256:e5fe65973f5737107bb8314f3829280d196c87d4da0b4a6e95000175512cf2c3", size = 788620, upload-time = "2025-01-28T02:15:24.637Z" } wheels = [ @@ -971,11 +969,11 @@ name = "fastapi" version = "0.138.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "annotated-doc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "starlette", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-inspection", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "annotated-doc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "starlette", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-inspection", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/5b/58/ff455d9fe47c60abadb34b9e05a304b1f05f5ab8000ac01565156b6f5e43/fastapi-0.138.0.tar.gz", hash = "sha256:d445a4877636ad191e7053e08c9bf98cb921a6756776848400bb773d1740c061", size = 419240, upload-time = "2026-06-20T01:18:05.259Z" } wheels = [ @@ -1085,7 +1083,7 @@ name = "ftfy" version = "6.3.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "wcwidth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "wcwidth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/a5/d3/8650919bc3c7c6e90ee3fa7fd618bf373cbbe55dff043bd67353dbb20cd8/ftfy-6.3.1.tar.gz", hash = "sha256:9b3c3d90f84fb267fe64d375a07b7f8912d817cf86009ae134aa03e1819506ec", size = 308927, upload-time = "2024-10-26T00:50:35.149Z" } wheels = [ @@ -1097,8 +1095,8 @@ name = "google-auth" version = "2.55.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cryptography", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pyasn1-modules", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cryptography", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pyasn1-modules", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/81/1c/70b23fc52b2bb3c70b379f3bd05c4a60ab3a873e30c6bd21c57e0154848a/google_auth-2.55.0.tar.gz", hash = "sha256:fcd3a130f575fa36403d38774af1c64a4fbfbca09215f0589d2372b5119697cb", size = 349379, upload-time = "2026-06-15T22:33:16.466Z" } wheels = [ @@ -1107,7 +1105,7 @@ wheels = [ [package.optional-dependencies] requests = [ - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] [[package]] @@ -1115,16 +1113,16 @@ name = "google-genai" version = "1.75.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "google-auth", extra = ["requests"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tenacity", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "websockets", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "google-auth", extra = ["requests"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tenacity", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "websockets", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/9d/59/3ed61240ef20b3ae6ed54e82c6f8b6d1f194947bc6679679dd6cdb037594/google_genai-1.75.0.tar.gz", hash = "sha256:56bac3991b311c93f980c0a2abcd287b672146905df1fbd71c92ed633d5a07cf", size = 539039, upload-time = "2026-05-04T22:48:54.857Z" } wheels = [ @@ -1217,7 +1215,7 @@ name = "httpx-retries" version = "0.5.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fb/f5/046cac13877ce9b55aebdbb3999e0e45b19b989a95c5fd1040fa04bd1f92/httpx_retries-0.5.0.tar.gz", hash = "sha256:d8c8e1e0852d84be3837aba0bcf78aeb89a4b77db95e8cc988c8c058830b3044", size = 15647, upload-time = "2026-04-20T01:21:47.154Z" } wheels = [ @@ -1245,16 +1243,16 @@ wheels = [ [[package]] name = "icij-common" -version = "0.8.2" +version = "0.8.3" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "aiostream", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/e0/51/91d85cc23e275bb51a5cae47873af9a7160f62797b25663d902cf38e6ab7/icij_common-0.8.2.tar.gz", hash = "sha256:7db68266d8facb43142131d81e998cd74c6ae73508456743e00df43a15cc2995", size = 15937, upload-time = "2026-04-07T12:13:39.774Z" } +sdist = { url = "https://files.pythonhosted.org/packages/6a/8d/4b2b9bb7dcac24d8bf514ccf2da6465ca3cd9b4c4f4bde0a1c97e681d595/icij_common-0.8.3.tar.gz", hash = "sha256:009dda7c1d688ecf7705cf88517deb3d0abe02300b42c8c433eb36516462fe75", size = 16161, upload-time = "2026-07-03T09:43:50.837Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/15/5a/99d123230023124ba4ec67765feb2a04b656ae1f473a14f6b29cc54b02bf/icij_common-0.8.2-py3-none-any.whl", hash = "sha256:c5a1a3ccd54e7bf53acad293a0af50ab15156b225791b43383197b896bae955f", size = 20565, upload-time = "2026-04-07T12:13:38.709Z" }, + { url = "https://files.pythonhosted.org/packages/1c/8e/99879d623fdf4e2371e44135dcf690d7f43b921616e4c918633fe4cece48/icij_common-0.8.3-py3-none-any.whl", hash = "sha256:5214cbb73dca364cca16005f5fac3c2f3944acab6dea269a68d26f631e39a348", size = 21030, upload-time = "2026-07-03T09:43:49.934Z" }, ] [package.optional-dependencies] @@ -1507,8 +1505,8 @@ name = "magika" version = "1.0.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "onnxruntime", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "onnxruntime", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/00/be/fa7d512ec15763ad7d8ba083e39bb445fa969ce5159361fa449a02569cf9/magika-1.0.3.tar.gz", hash = "sha256:ad3216012f6dd337be34c23ae3dfab36f0623bc62fbfb329f9a1852c9ad40304", size = 3041970, upload-time = "2026-05-04T08:41:53.124Z" } wheels = [ @@ -1531,7 +1529,7 @@ name = "mammoth" version = "1.12.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cobble", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cobble", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/af/0c/b8d04b142c28f705ac434aedfb492f62e3fa9082421b6aa0ec7be9202dc7/mammoth-1.12.0.tar.gz", hash = "sha256:10955a55d9173167b550de3aeb8f2ed48b420756fd66378156b2f78661a33dd5", size = 53388, upload-time = "2026-03-12T21:42:37.289Z" } wheels = [ @@ -1564,8 +1562,8 @@ name = "markdownify" version = "1.2.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "six", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "six", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/3f/bc/c8c8eea5335341306b0fa7e1cb33c5e1c8d24ef70ddd684da65f41c49c92/markdownify-1.2.2.tar.gz", hash = "sha256:b274f1b5943180b031b699b199cbaeb1e2ac938b75851849a31fd0c3d6603d09", size = 18816, upload-time = "2025-11-16T19:21:18.565Z" } wheels = [ @@ -1577,30 +1575,30 @@ name = "marker-pdf" version = "1.10.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anthropic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "ftfy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "google-genai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "markdown2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "markdownify", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "rapidfuzz", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "scikit-learn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "surya-ocr", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anthropic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "ftfy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "google-genai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "markdown2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "markdownify", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "rapidfuzz", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "scikit-learn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "surya-ocr", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/04/bc/a9e72fd015f562b1f86c9dbc201c80365e7baf43d5aa616ba1c89453537f/marker_pdf-1.10.2.tar.gz", hash = "sha256:ce0fc839e11ad7519a576d254ca9d51a0f9454b9d7da02211f722b141317f9f1", size = 140056, upload-time = "2026-01-31T00:04:55.179Z" } wheels = [ @@ -1663,34 +1661,34 @@ name = "mineru" version = "3.4.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "fast-langdetect", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "fastapi", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "json-repair", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "magika", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mammoth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mineru-vl-utils", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "modelscope", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openpyxl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pylatexenc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypptx-with-oxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-docx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-multipart", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "reportlab", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "uvicorn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fast-langdetect", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fastapi", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "json-repair", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "magika", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "mammoth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "mineru-vl-utils", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "modelscope", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openpyxl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pylatexenc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypptx-with-oxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-docx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-multipart", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "reportlab", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "uvicorn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/2e/34/d506545a25aa2a84ee9b6d9b7f7ec355e9de8723c08f00679152d2499aee/mineru-3.4.0-py3-none-any.whl", hash = "sha256:c7dd1c6970c00bff01465210217c1107819072a1eeff02bc4a8f764b0672e10f", size = 1535099, upload-time = "2026-06-18T07:23:46.14Z" }, @@ -1732,12 +1730,12 @@ name = "mineru-vl-utils" version = "1.0.5" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "aiofiles", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx-retries", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "aiofiles", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx-retries", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/a0/e6/71556a212ab86c8845f5bec220952e08ba288342593161da41076c678627/mineru_vl_utils-1.0.5-py3-none-any.whl", hash = "sha256:cf910e68f0607634e61b613b7f5992daf604bf80b400d81e7b9f0f117b7c3c15", size = 94450, upload-time = "2026-06-11T13:49:32.395Z" }, @@ -1816,12 +1814,12 @@ name = "modelscope" version = "1.37.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "filelock", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "setuptools", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "urllib3", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filelock", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "setuptools", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "urllib3", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fa/9e/9fd75db9797b9543ce78a646538a3c5537cb6adb015d128b4cbc2b4169e2/modelscope-1.37.1.tar.gz", hash = "sha256:7d9124970b4e53639bebe89f60473dee7d5a80589bf35bfbef4c839933cdaf93", size = 4594233, upload-time = "2026-05-22T05:57:09.621Z" } wheels = [ @@ -2297,10 +2295,10 @@ name = "onnxruntime" version = "1.27.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "flatbuffers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "protobuf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "flatbuffers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "protobuf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/fb/2b/54208fd03ad410480bc17edf4869376362da8bbf46fe186ddf4cb5cc20fe/onnxruntime-1.27.0-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:b3e5b58b8c89c2b20e086e890aa9527377e5c240dc3ecc1640d18e07705eeb1c", size = 18432958, upload-time = "2026-06-15T22:42:53.105Z" }, @@ -2316,14 +2314,14 @@ name = "openai" version = "1.109.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/c6/a1/a303104dc55fc546a3f6914c842d3da471c64eec92043aef8f652eb6c524/openai-1.109.1.tar.gz", hash = "sha256:d173ed8dbca665892a6db099b4a2dfac624f94d20a93f46eb0b56aae940ed869", size = 564133, upload-time = "2025-09-24T13:00:53.075Z" } wheels = [ @@ -2407,10 +2405,10 @@ name = "pdftext" version = "0.6.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/a9/7b/fe3205d44d6058932bbc785f0b9da2ed35b62e17479a8a7d2baca9df1cc6/pdftext-0.6.3.tar.gz", hash = "sha256:ab5c5dfe0f1fb78de1db837ccadac1ea41b07ce1890fead973c9a84cdaf54dec", size = 21968, upload-time = "2025-06-11T14:42:09.492Z" } wheels = [ @@ -2591,7 +2589,7 @@ name = "pyasn1-modules" version = "0.4.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "pyasn1", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pyasn1", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/e9/e6/78ebbb10a8c8e4b61a59249394a4a594c1a7af95593dc933a349c8d00964/pyasn1_modules-0.4.2.tar.gz", hash = "sha256:677091de870a80aae844b1ca6134f54652fa2c8c5a52aa396440ac3106e941e6", size = 307892, upload-time = "2025-03-28T02:41:22.17Z" } wheels = [ @@ -2837,10 +2835,10 @@ name = "pypptx-with-oxml" version = "1.0.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "xlsxwriter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "xlsxwriter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/43/17/7a499cea9e03b37ef5a502a7792431b5e609328908e2e11b02453071367b/pypptx_with_oxml-1.0.3.tar.gz", hash = "sha256:a093d8c8e99dfcd0131ea98b4dbc0a0a40adf44869edbb74c57d0cb87b8e4208", size = 9767175, upload-time = "2026-01-30T08:51:32.511Z" } wheels = [ @@ -3103,8 +3101,8 @@ name = "reportlab" version = "5.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "charset-normalizer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "charset-normalizer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/41/d6/4b7b0cf56880eb96533e607967be6a939e344675601e033d113a0bfa1f4e/reportlab-5.0.0.tar.gz", hash = "sha256:e4494a0c6623ae213bb856fba523171b2b54a7bf629fda02d5e525a7b899a784", size = 3701928, upload-time = "2026-06-18T11:34:31.145Z" } wheels = [ @@ -3144,9 +3142,9 @@ name = "robust-downloader" version = "0.0.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "colorlog", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "colorlog", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/63/20/8d28efa080f58fa06f6378875ac482ee511c076369e5293a2e65128cf9a0/robust-downloader-0.0.2.tar.gz", hash = "sha256:08c938b96e317abe6b037e34230a91bda9b5d613f009bca4a47664997c61de90", size = 15785, upload-time = "2023-11-13T03:00:20.637Z" } wheels = [ @@ -3261,11 +3259,11 @@ name = "scikit-learn" version = "1.9.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "joblib", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "narwhals", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "scipy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "threadpoolctl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "joblib", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "narwhals", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "scipy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "threadpoolctl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fa/6f/37092bdb25f712817231799fc5674d8e704066a8a70c1d2d40517e18b4ab/scikit_learn-1.9.0.tar.gz", hash = "sha256:8833266989d3a5110178a9fae30783675460724d0e1efb13b14901d2c660c557", size = 7750767, upload-time = "2026-06-02T11:54:32.706Z" } wheels = [ @@ -3440,7 +3438,7 @@ name = "starlette" version = "1.3.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/eb/e3/7c1dc7381d9f8ab7d854328ebfa884e62cb3f3d8549ddfd37c7814f42afa/starlette-1.3.1.tar.gz", hash = "sha256:05d0213193f2fbaae60e2ecb593b4add4262ad4e46536b54abe36f11a71724e0", size = 2703240, upload-time = "2026-06-12T09:23:11.602Z" } wheels = [ @@ -3452,22 +3450,22 @@ name = "surya-ocr" version = "0.17.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "einops", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "opencv-python-headless", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "platformdirs", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "einops", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "opencv-python-headless", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "platformdirs", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, + { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/b3/b4/b0f3afd024b4e6fcb4db6fdd17ee963d081e960c907aebd0556ff948abb6/surya_ocr-0.17.1.tar.gz", hash = "sha256:349d78d854c1ed5f816e583545ed6451aa0bc6992e283a805034799aacee8c24", size = 161854, upload-time = "2026-01-30T21:52:59.361Z" } wheels = [ @@ -3979,8 +3977,8 @@ name = "uvicorn" version = "0.49.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "h11", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "h11", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/c4/1f/fa18009dea8469069cca78a4e877a008ab78f08b064bfc9ab891579077ff/uvicorn-0.49.0.tar.gz", hash = "sha256:ebf4271aa580d9de97f93192d4595176df6e91f9aae919ca73e4fc07df1e66a3", size = 91284, upload-time = "2026-06-03T22:01:30.448Z" } wheels = [ diff --git a/workers/extract-worker/uv.lock b/workers/extract-worker/uv.lock index a4891f6f..4098f0e9 100644 --- a/workers/extract-worker/uv.lock +++ b/workers/extract-worker/uv.lock @@ -172,13 +172,13 @@ name = "anthropic" version = "0.46.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/d4/68/3b4c045edf6dc6933895e8f279cc77c7684874c8aba46a4e6241c8b147cf/anthropic-0.46.0.tar.gz", hash = "sha256:eac3d43271d02321a57c3ca68aca84c3d58873e8e72d1433288adee2d46b745b", size = 202191, upload-time = "2025-02-18T20:35:33.314Z" } wheels = [ @@ -271,7 +271,7 @@ name = "cffi" version = "2.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "pycparser", marker = "(implementation_name != 'PyPy' and platform_machine == 'x86_64' and sys_platform == 'linux') or (implementation_name != 'PyPy' and sys_platform == 'darwin') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pycparser", marker = "(implementation_name != 'PyPy' and platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name != 'PyPy' and sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (implementation_name == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/eb/56/b1ba7935a17738ae8453301356628e8147c79dbb825bcbc73dc7401f9846/cffi-2.0.0.tar.gz", hash = "sha256:44d1b5909021139fe36001ae048dbdde8214afa20200eda0f64c068cac5d5529", size = 523588, upload-time = "2025-09-08T23:24:04.541Z" } wheels = [ @@ -348,7 +348,7 @@ name = "cryptography" version = "49.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cffi", marker = "(platform_machine == 'x86_64' and platform_python_implementation != 'PyPy' and sys_platform == 'linux') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_python_implementation != 'PyPy' and sys_platform == 'darwin') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cffi", marker = "(platform_machine == 'x86_64' and platform_python_implementation != 'PyPy' and sys_platform == 'linux') or (platform_machine != 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_python_implementation != 'PyPy' and sys_platform == 'darwin') or (platform_python_implementation == 'PyPy' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'darwin' and sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/1f/99/d1c90d6041656cc6ee229dc99cd67fd0cd5aec3c5f7d72fffc27cc750054/cryptography-49.0.0.tar.gz", hash = "sha256:f89660a348f4f78a92366240a61404e337586ef7f5909a2fef59ca88ef505493", size = 854345, upload-time = "2026-06-12T20:02:30.512Z" } wheels = [ @@ -597,7 +597,7 @@ dev = [ requires-dist = [ { name = "cuda-bindings", marker = "sys_platform == 'linux' and extra == 'gpu'", specifier = "==12.9.4" }, { name = "datashare-python", editable = "../../datashare-python" }, - { name = "extract-core", specifier = "==0.7.0" }, + { name = "extract-core", specifier = "==0.7.1" }, { name = "extract-python", extras = ["docling", "marker"], marker = "extra == 'base'", specifier = "==0.7.2" }, { name = "extract-python", extras = ["mineru"], marker = "extra == 'mineru'", specifier = "==0.7.2" }, { name = "flash-attn", marker = "extra == 'flash-attn'", specifier = "==2.8.3.post1" }, @@ -628,7 +628,7 @@ dev = [ [[package]] name = "datashare-python" -version = "0.9.8" +version = "0.9.9" source = { editable = "../../datashare-python" } dependencies = [ { name = "aiohttp", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, @@ -927,19 +927,17 @@ wheels = [ [[package]] name = "extract-core" -version = "0.7.0" +version = "0.7.1" source = { registry = "https://pypi.org/simple" } dependencies = [ { name = "docling-slim", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "icij-common", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "marker-pdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mineru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, { name = "pydantic-extra-types", extra = ["pycountry"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] -sdist = { url = "https://files.pythonhosted.org/packages/e6/ea/1bbc3485d0570455ca874aa84b4f02c4b4d509123710fe1832c02acca08c/extract_core-0.7.0.tar.gz", hash = "sha256:09bfef043c84afb11dd0d50582044cee7c6a5d3cdf95788067d90904b737dd80", size = 138372, upload-time = "2026-06-25T09:25:17.672Z" } +sdist = { url = "https://files.pythonhosted.org/packages/4c/a3/7b4a72cb925ee84da756520675b428ac872dd9fc9962de2a0e80703ac7ea/extract_core-0.7.1.tar.gz", hash = "sha256:bd8eec934bf73749e912455f9d7818320ef73382e6f8382a93d502c9466fba53", size = 109907, upload-time = "2026-07-02T08:02:52.016Z" } wheels = [ - { url = "https://files.pythonhosted.org/packages/d4/41/2eb44fcd9b33d9a4da7d61691f2eea3ebbad565288601d1b8169d2f169f7/extract_core-0.7.0-py3-none-any.whl", hash = "sha256:51d7cb28fcc8173a34e20862a236354f5da35e2bd0156d0001650e22d6e655d3", size = 10414, upload-time = "2026-06-25T09:25:18.496Z" }, + { url = "https://files.pythonhosted.org/packages/ac/14/584c2c62830871464c475a7ea4117ce0df6fefb6fb31ac4df7838e6d0b13/extract_core-0.7.1-py3-none-any.whl", hash = "sha256:b360c958d4375c263e391bd1d7328d6786ec20d095d8758e86329548333a110d", size = 10382, upload-time = "2026-07-02T08:02:53.029Z" }, ] [[package]] @@ -984,9 +982,9 @@ name = "fast-langdetect" version = "0.2.5" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "fasttext-predict", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "robust-downloader", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fasttext-predict", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "robust-downloader", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/8d/17/2753b3d280e16f594e7a5d735568c021a355cd1edd3827f26cda5c9fd460/fast_langdetect-0.2.5.tar.gz", hash = "sha256:e5fe65973f5737107bb8314f3829280d196c87d4da0b4a6e95000175512cf2c3", size = 788620, upload-time = "2025-01-28T02:15:24.637Z" } wheels = [ @@ -998,11 +996,11 @@ name = "fastapi" version = "0.138.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "annotated-doc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "starlette", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-inspection", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "annotated-doc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "starlette", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-inspection", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/5b/58/ff455d9fe47c60abadb34b9e05a304b1f05f5ab8000ac01565156b6f5e43/fastapi-0.138.0.tar.gz", hash = "sha256:d445a4877636ad191e7053e08c9bf98cb921a6756776848400bb773d1740c061", size = 419240, upload-time = "2026-06-20T01:18:05.259Z" } wheels = [ @@ -1112,7 +1110,7 @@ name = "ftfy" version = "6.3.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "wcwidth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "wcwidth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/a5/d3/8650919bc3c7c6e90ee3fa7fd618bf373cbbe55dff043bd67353dbb20cd8/ftfy-6.3.1.tar.gz", hash = "sha256:9b3c3d90f84fb267fe64d375a07b7f8912d817cf86009ae134aa03e1819506ec", size = 308927, upload-time = "2024-10-26T00:50:35.149Z" } wheels = [ @@ -1124,8 +1122,8 @@ name = "google-auth" version = "2.55.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cryptography", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pyasn1-modules", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cryptography", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pyasn1-modules", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/81/1c/70b23fc52b2bb3c70b379f3bd05c4a60ab3a873e30c6bd21c57e0154848a/google_auth-2.55.0.tar.gz", hash = "sha256:fcd3a130f575fa36403d38774af1c64a4fbfbca09215f0589d2372b5119697cb", size = 349379, upload-time = "2026-06-15T22:33:16.466Z" } wheels = [ @@ -1134,7 +1132,7 @@ wheels = [ [package.optional-dependencies] requests = [ - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] [[package]] @@ -1142,16 +1140,16 @@ name = "google-genai" version = "1.75.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "google-auth", extra = ["requests"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tenacity", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "websockets", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "google-auth", extra = ["requests"], marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tenacity", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "websockets", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/9d/59/3ed61240ef20b3ae6ed54e82c6f8b6d1f194947bc6679679dd6cdb037594/google_genai-1.75.0.tar.gz", hash = "sha256:56bac3991b311c93f980c0a2abcd287b672146905df1fbd71c92ed633d5a07cf", size = 539039, upload-time = "2026-05-04T22:48:54.857Z" } wheels = [ @@ -1244,7 +1242,7 @@ name = "httpx-retries" version = "0.5.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fb/f5/046cac13877ce9b55aebdbb3999e0e45b19b989a95c5fd1040fa04bd1f92/httpx_retries-0.5.0.tar.gz", hash = "sha256:d8c8e1e0852d84be3837aba0bcf78aeb89a4b77db95e8cc988c8c058830b3044", size = 15647, upload-time = "2026-04-20T01:21:47.154Z" } wheels = [ @@ -1534,8 +1532,8 @@ name = "magika" version = "1.0.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "onnxruntime", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "onnxruntime", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/00/be/fa7d512ec15763ad7d8ba083e39bb445fa969ce5159361fa449a02569cf9/magika-1.0.3.tar.gz", hash = "sha256:ad3216012f6dd337be34c23ae3dfab36f0623bc62fbfb329f9a1852c9ad40304", size = 3041970, upload-time = "2026-05-04T08:41:53.124Z" } wheels = [ @@ -1558,7 +1556,7 @@ name = "mammoth" version = "1.12.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "cobble", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "cobble", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/af/0c/b8d04b142c28f705ac434aedfb492f62e3fa9082421b6aa0ec7be9202dc7/mammoth-1.12.0.tar.gz", hash = "sha256:10955a55d9173167b550de3aeb8f2ed48b420756fd66378156b2f78661a33dd5", size = 53388, upload-time = "2026-03-12T21:42:37.289Z" } wheels = [ @@ -1591,8 +1589,8 @@ name = "markdownify" version = "1.2.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "six", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "six", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/3f/bc/c8c8eea5335341306b0fa7e1cb33c5e1c8d24ef70ddd684da65f41c49c92/markdownify-1.2.2.tar.gz", hash = "sha256:b274f1b5943180b031b699b199cbaeb1e2ac938b75851849a31fd0c3d6603d09", size = 18816, upload-time = "2025-11-16T19:21:18.565Z" } wheels = [ @@ -1604,30 +1602,30 @@ name = "marker-pdf" version = "1.10.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anthropic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "ftfy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "google-genai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "markdown2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "markdownify", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "rapidfuzz", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "scikit-learn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "surya-ocr", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anthropic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "ftfy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "google-genai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "markdown2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "markdownify", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "rapidfuzz", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "regex", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "scikit-learn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "surya-ocr", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/04/bc/a9e72fd015f562b1f86c9dbc201c80365e7baf43d5aa616ba1c89453537f/marker_pdf-1.10.2.tar.gz", hash = "sha256:ce0fc839e11ad7519a576d254ca9d51a0f9454b9d7da02211f722b141317f9f1", size = 140056, upload-time = "2026-01-31T00:04:55.179Z" } wheels = [ @@ -1690,34 +1688,34 @@ name = "mineru" version = "3.4.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "fast-langdetect", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "fastapi", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "json-repair", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "magika", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mammoth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "mineru-vl-utils", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "modelscope", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "openpyxl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pylatexenc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypptx-with-oxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-docx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-multipart", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "reportlab", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "uvicorn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "beautifulsoup4", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fast-langdetect", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "fastapi", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "huggingface-hub", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "json-repair", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "magika", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "mammoth", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "mineru-vl-utils", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "modelscope", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openai", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "openpyxl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pdftext", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pylatexenc", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypptx-with-oxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-docx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-multipart", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "reportlab", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "uvicorn", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/2e/34/d506545a25aa2a84ee9b6d9b7f7ec355e9de8723c08f00679152d2499aee/mineru-3.4.0-py3-none-any.whl", hash = "sha256:c7dd1c6970c00bff01465210217c1107819072a1eeff02bc4a8f764b0672e10f", size = 1535099, upload-time = "2026-06-18T07:23:46.14Z" }, @@ -1759,12 +1757,12 @@ name = "mineru-vl-utils" version = "1.0.5" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "aiofiles", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx-retries", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "aiofiles", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx-retries", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "loguru", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/a0/e6/71556a212ab86c8845f5bec220952e08ba288342593161da41076c678627/mineru_vl_utils-1.0.5-py3-none-any.whl", hash = "sha256:cf910e68f0607634e61b613b7f5992daf604bf80b400d81e7b9f0f117b7c3c15", size = 94450, upload-time = "2026-06-11T13:49:32.395Z" }, @@ -1843,12 +1841,12 @@ name = "modelscope" version = "1.37.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "filelock", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "setuptools", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "urllib3", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filelock", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "setuptools", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "urllib3", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fa/9e/9fd75db9797b9543ce78a646538a3c5537cb6adb015d128b4cbc2b4169e2/modelscope-1.37.1.tar.gz", hash = "sha256:7d9124970b4e53639bebe89f60473dee7d5a80589bf35bfbef4c839933cdaf93", size = 4594233, upload-time = "2026-05-22T05:57:09.621Z" } wheels = [ @@ -2324,10 +2322,10 @@ name = "onnxruntime" version = "1.27.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "flatbuffers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "protobuf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "flatbuffers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "packaging", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "protobuf", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] wheels = [ { url = "https://files.pythonhosted.org/packages/fb/2b/54208fd03ad410480bc17edf4869376362da8bbf46fe186ddf4cb5cc20fe/onnxruntime-1.27.0-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:b3e5b58b8c89c2b20e086e890aa9527377e5c240dc3ecc1640d18e07705eeb1c", size = 18432958, upload-time = "2026-06-15T22:42:53.105Z" }, @@ -2343,14 +2341,14 @@ name = "openai" version = "1.109.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "distro", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "httpx", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "jiter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "sniffio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/c6/a1/a303104dc55fc546a3f6914c842d3da471c64eec92043aef8f652eb6c524/openai-1.109.1.tar.gz", hash = "sha256:d173ed8dbca665892a6db099b4a2dfac624f94d20a93f46eb0b56aae940ed869", size = 564133, upload-time = "2025-09-24T13:00:53.075Z" } wheels = [ @@ -2434,10 +2432,10 @@ name = "pdftext" version = "0.6.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/a9/7b/fe3205d44d6058932bbc785f0b9da2ed35b62e17479a8a7d2baca9df1cc6/pdftext-0.6.3.tar.gz", hash = "sha256:ab5c5dfe0f1fb78de1db837ccadac1ea41b07ce1890fead973c9a84cdaf54dec", size = 21968, upload-time = "2025-06-11T14:42:09.492Z" } wheels = [ @@ -2618,7 +2616,7 @@ name = "pyasn1-modules" version = "0.4.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "pyasn1", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pyasn1", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/e9/e6/78ebbb10a8c8e4b61a59249394a4a594c1a7af95593dc933a349c8d00964/pyasn1_modules-0.4.2.tar.gz", hash = "sha256:677091de870a80aae844b1ca6134f54652fa2c8c5a52aa396440ac3106e941e6", size = 307892, upload-time = "2025-03-28T02:41:22.17Z" } wheels = [ @@ -2864,10 +2862,10 @@ name = "pypptx-with-oxml" version = "1.0.3" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "xlsxwriter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "lxml", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "typing-extensions", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "xlsxwriter", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/43/17/7a499cea9e03b37ef5a502a7792431b5e609328908e2e11b02453071367b/pypptx_with_oxml-1.0.3.tar.gz", hash = "sha256:a093d8c8e99dfcd0131ea98b4dbc0a0a40adf44869edbb74c57d0cb87b8e4208", size = 9767175, upload-time = "2026-01-30T08:51:32.511Z" } wheels = [ @@ -3130,8 +3128,8 @@ name = "reportlab" version = "5.0.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "charset-normalizer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "charset-normalizer", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/41/d6/4b7b0cf56880eb96533e607967be6a939e344675601e033d113a0bfa1f4e/reportlab-5.0.0.tar.gz", hash = "sha256:e4494a0c6623ae213bb856fba523171b2b54a7bf629fda02d5e525a7b899a784", size = 3701928, upload-time = "2026-06-18T11:34:31.145Z" } wheels = [ @@ -3171,9 +3169,9 @@ name = "robust-downloader" version = "0.0.2" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "colorlog", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "colorlog", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "requests", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "tqdm", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/63/20/8d28efa080f58fa06f6378875ac482ee511c076369e5293a2e65128cf9a0/robust-downloader-0.0.2.tar.gz", hash = "sha256:08c938b96e317abe6b037e34230a91bda9b5d613f009bca4a47664997c61de90", size = 15785, upload-time = "2023-11-13T03:00:20.637Z" } wheels = [ @@ -3288,11 +3286,11 @@ name = "scikit-learn" version = "1.9.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "joblib", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "narwhals", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "scipy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "threadpoolctl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "joblib", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "narwhals", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "numpy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "scipy", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "threadpoolctl", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/fa/6f/37092bdb25f712817231799fc5674d8e704066a8a70c1d2d40517e18b4ab/scikit_learn-1.9.0.tar.gz", hash = "sha256:8833266989d3a5110178a9fae30783675460724d0e1efb13b14901d2c660c557", size = 7750767, upload-time = "2026-06-02T11:54:32.706Z" } wheels = [ @@ -3467,7 +3465,7 @@ name = "starlette" version = "1.3.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "anyio", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/eb/e3/7c1dc7381d9f8ab7d854328ebfa884e62cb3f3d8549ddfd37c7814f42afa/starlette-1.3.1.tar.gz", hash = "sha256:05d0213193f2fbaae60e2ecb593b4add4262ad4e46536b54abe36f11a71724e0", size = 2703240, upload-time = "2026-06-12T09:23:11.602Z" } wheels = [ @@ -3479,22 +3477,22 @@ name = "surya-ocr" version = "0.17.1" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "einops", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "opencv-python-headless", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "platformdirs", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform == 'darwin' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, - { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "einops", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "filetype", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "opencv-python-headless", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pillow", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "platformdirs", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pre-commit", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pydantic-settings", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "pypdfium2", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "python-dotenv", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.11.0+cu129", source = { registry = "https://download.pytorch.org/whl/cu129" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra != 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "torch", version = "2.12.1", source = { registry = "https://pypi.org/simple" }, marker = "(platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu' and extra != 'extra-24-datashare-extract-worker-gpu') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-cpu') or (sys_platform == 'darwin' and extra == 'extra-24-datashare-extract-worker-base' and extra != 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-mineru') or (extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-gpu' and extra == 'extra-24-datashare-extract-worker-mineru')" }, + { name = "transformers", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/b3/b4/b0f3afd024b4e6fcb4db6fdd17ee963d081e960c907aebd0556ff948abb6/surya_ocr-0.17.1.tar.gz", hash = "sha256:349d78d854c1ed5f816e583545ed6451aa0bc6992e283a805034799aacee8c24", size = 161854, upload-time = "2026-01-30T21:52:59.361Z" } wheels = [ @@ -4006,8 +4004,8 @@ name = "uvicorn" version = "0.49.0" source = { registry = "https://pypi.org/simple" } dependencies = [ - { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, - { name = "h11", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-base' and extra == 'extra-24-datashare-extract-worker-mineru') or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "click", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, + { name = "h11", marker = "(platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'x86_64' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu') or sys_platform == 'darwin' or (sys_platform != 'linux' and extra == 'extra-24-datashare-extract-worker-cpu' and extra == 'extra-24-datashare-extract-worker-gpu')" }, ] sdist = { url = "https://files.pythonhosted.org/packages/c4/1f/fa18009dea8469069cca78a4e877a008ab78f08b064bfc9ab891579077ff/uvicorn-0.49.0.tar.gz", hash = "sha256:ebf4271aa580d9de97f93192d4595176df6e91f9aae919ca73e4fc07df1e66a3", size = 91284, upload-time = "2026-06-03T22:01:30.448Z" } wheels = [ diff --git a/workers/translation-worker/tests/conftest.py b/workers/translation-worker/tests/conftest.py index 1f2a409c..c7d089d4 100644 --- a/workers/translation-worker/tests/conftest.py +++ b/workers/translation-worker/tests/conftest.py @@ -24,29 +24,24 @@ test_worker_config, worker_lifetime_deps, ) -from datashare_python.dependencies import SHARED, lifespan_shared_resources -from datashare_python.objects import DatashareLanguage, Document, Shared +from datashare_python.objects import DatashareLanguage, Document from datashare_python.types_ import ContextManagerFactory, TemporalClient from datashare_python.worker import worker_context from icij_common.es import ESClient from translation_worker.activities import TranslationActivities from translation_worker.config import TranslationWorkerConfig -from translation_worker.dependencies import ( - set_es_client, - set_hunyuan_translator, - set_worker_config, -) +from translation_worker.dependencies import REGISTRY from translation_worker.workflows import TaskQueue, TranslationWorkflow @pytest.fixture(scope="session") def test_io_deps() -> list[ContextManagerFactory]: - return [set_worker_config, set_es_client] + return REGISTRY["translation.io"] @pytest.fixture(scope="session") def test_inference_deps() -> list[ContextManagerFactory]: - return [set_worker_config, set_es_client, set_hunyuan_translator] + return REGISTRY["translation.inference"] @pytest.fixture(scope="session") @@ -210,9 +205,3 @@ async def translation_inference_worker( ) async with worker_ctx: yield - - -@pytest.fixture(scope="session") -def test_shared_resources() -> Shared: - SHARED.set(Shared()) - return lifespan_shared_resources() diff --git a/workers/translation-worker/tests/test_activities.py b/workers/translation-worker/tests/test_activities.py index 8ee38ff4..1c811520 100644 --- a/workers/translation-worker/tests/test_activities.py +++ b/workers/translation-worker/tests/test_activities.py @@ -6,7 +6,7 @@ from unittest.mock import patch import pytest -from datashare_python.objects import DatashareLanguage, Document, Translation +from datashare_python.objects import DatashareLanguage, Document, Language, Translation from icij_common.es import ( DOC_CONTENT, DOC_LANGUAGE, @@ -73,6 +73,8 @@ class MockSentenceSplitter(SentenceSplitter): def __init__(self, splits: list[list[str]]): self._splits = iter(splits) + def load(self, language: Language) -> Self: ... + def split_sentences(self, text: str) -> list[str]: # noqa: ARG002 return next(self._splits) @@ -353,13 +355,16 @@ async def test_translate_docs_act__returns_count_of_unique_docs_translated( sentence_splitter = MockSentenceSplitter(sentences) monkeypatch.setattr(activities, "_update_docs_translation", _do_nothing_es_update) es_client = MockESClient([FR_DOC_1, FR_DOC_2]) + worker_config = TranslationWorkerConfig() # When - with translator.load(source=DS_ENGLISH, target=DS_ENGLISH): + with translator.load_cm( + source=DS_ENGLISH, target=DS_ENGLISH, worker_config=worker_config + ): n_translated = await translate_docs_act( batches, project=TEST_PROJECT, es_client=es_client, - worker_config=TranslationWorkerConfig(), + worker_config=worker_config, translator=translator, sentence_splitter=sentence_splitter, ) @@ -378,13 +383,16 @@ async def test_translate_docs_act__sentences_from_same_doc_count_as_one( sentence_splitter = MockSentenceSplitter(sentences) monkeypatch.setattr(activities, "_update_docs_translation", _do_nothing_es_update) es_client = MockESClient([FR_DOC_1]) + worker_config = TranslationWorkerConfig() # When - with translator.load(source=DS_ENGLISH, target=DS_ENGLISH): + with translator.load_cm( + source=DS_ENGLISH, target=DS_ENGLISH, worker_config=worker_config + ): n_translated = await translate_docs_act( batches, project=TEST_PROJECT, es_client=es_client, - worker_config=TranslationWorkerConfig(), + worker_config=worker_config, translator=translator, sentence_splitter=sentence_splitter, ) @@ -403,13 +411,16 @@ async def test_translate_docs_act__es_update(monkeypatch) -> None: update_doc = partial(_capturing_es_update, captured=captured) monkeypatch.setattr(activities, "_update_docs_translation", update_doc) es_client = MockESClient([FR_DOC_1]) + worker_config = TranslationWorkerConfig() # When - with translator.load(source=DS_FRENCH, target=DS_ENGLISH): + with translator.load_cm( + source=DS_FRENCH, target=DS_ENGLISH, worker_config=worker_config + ): n_translated = await translate_docs_act( batches, project=TEST_PROJECT, es_client=es_client, - worker_config=TranslationWorkerConfig(), + worker_config=worker_config, translator=translator, sentence_splitter=sentence_splitter, ) diff --git a/workers/translation-worker/tests/test_hunyuan_translator.py b/workers/translation-worker/tests/test_hunyuan_translator.py index 0bd45d32..1fcbce33 100644 --- a/workers/translation-worker/tests/test_hunyuan_translator.py +++ b/workers/translation-worker/tests/test_hunyuan_translator.py @@ -2,8 +2,7 @@ import sys from unittest.mock import MagicMock, patch -from datashare_python.objects import Shared -from translation_worker.constants import DEFAULT_HUNYUAN_MODEL_REF, HUNYUAN_SHARED_KEY +from translation_worker.config import DEFAULT_HUNYUAN_MODEL_REF from translation_worker.translators.hunyuan import HunyuanMtTranslator from tests.conftest import DS_CHINESE, DS_ENGLISH @@ -112,7 +111,7 @@ def test__load__initialises_tokenizer_and_model_from_config_model_ref_with_devic with ( patch.dict(sys.modules, {"transformers": dummy_tf}), ): - translator._load(MagicMock(), target=MagicMock(), worker_config=MagicMock()) + translator.load(MagicMock(), target=MagicMock(), worker_config=MagicMock()) dummy_tf.AutoTokenizer.from_pretrained.assert_called_once_with( DEFAULT_HUNYUAN_MODEL_REF @@ -123,69 +122,3 @@ def test__load__initialises_tokenizer_and_model_from_config_model_ref_with_devic device_map="cuda:0", torch_dtype="bfloat16", ) - - -def test__load__assigns_loaded_tokenizer_and_model_to_instance_without_shared_resources() -> ( # noqa: E501 - None -): - translator = _translator() - dummy_tf = _dummy_transformers() - mock_tokenizer = MagicMock() - mock_model = MagicMock() - mock_model.to.return_value = mock_model - dummy_tf.AutoTokenizer.from_pretrained.return_value = mock_tokenizer - dummy_tf.AutoModelForCausalLM.from_pretrained.return_value = mock_model - - with ( - patch.dict(sys.modules, {"transformers": dummy_tf}), - ): - translator._load(MagicMock(), target=MagicMock(), worker_config=MagicMock()) - - assert translator._tokenizer is mock_tokenizer - assert translator._translator is mock_model - - -def test__load__assigns_loaded_tokenizer_and_model_to_instance_with_shared_resources( - test_shared_resources: Shared, -) -> None: - translator = _translator() - dummy_tf = _dummy_transformers() - mock_tokenizer = MagicMock() - mock_model = MagicMock() - mock_model.to.return_value = mock_model - dummy_tf.AutoTokenizer.from_pretrained.return_value = mock_tokenizer - dummy_tf.AutoModelForCausalLM.from_pretrained.return_value = mock_model - - test_shared_resources.set_resource(HUNYUAN_SHARED_KEY, None) - - with ( - patch.dict(sys.modules, {"transformers": dummy_tf}), - ): - translator._load(MagicMock(), target=MagicMock(), worker_config=MagicMock()) - - assert translator._tokenizer is mock_tokenizer - assert translator._translator is mock_model - assert test_shared_resources.get_resource(HUNYUAN_SHARED_KEY) is mock_model - - -# @pytest.mark.parametrize( -# ("source", "initial_translator", "expected_translator_type"), -# [ -# # should route to hunyuan -# (DS_CHINESE, ArgosTranslatorConfig(), HunyuanMtTranslatorConfig), -# # should remain with argos -# (DS_FRENCH, ArgosTranslatorConfig(), ArgosTranslatorConfig), -# # should remain with hunyuan -# (DS_CHINESE, HunyuanMtTranslatorConfig(), HunyuanMtTranslatorConfig), -# # should route to argos -# (DS_FRENCH, HunyuanMtTranslatorConfig(), ArgosTranslatorConfig), -# ], -# ) -# def test__set_config_translator( -# source: DatashareLanguage, -# initial_translator: ArgosTranslatorConfig | HunyuanMtTranslatorConfig, -# expected_translator_type: type, -# ) -> None: -# config = TranslationConfig(translator=initial_translator) -# result = _set_config_translator(source, config) -# assert isinstance(result.translator, expected_translator_type) diff --git a/workers/translation-worker/translation_worker/activities.py b/workers/translation-worker/translation_worker/activities.py index a8a42229..1d681197 100644 --- a/workers/translation-worker/translation_worker/activities.py +++ b/workers/translation-worker/translation_worker/activities.py @@ -11,6 +11,7 @@ from datashare_python.utils import ( ActivityWithProgress, activity_defn, + config_cache_key, to_raw_async_progress, ) from elasticsearch._async.helpers import async_bulk @@ -31,9 +32,15 @@ from translation_worker.constants import DOC_CONTENT_TEXT_LENGTH -from .config import TranslationConfig, TranslationWorkerConfig +from .config import ( + SentenceSplitterConfig, + TranslationConfig, + TranslationWorkerConfig, + TranslatorConfig, +) from .constants import BATCHING_DOC_SOURCES, TRANSLATION_DOC_SOURCES -from .objects import _LANGUAGE_TYPE_ADAPTER, Language +from .dependencies import lifespan_sentence_splitter_cache, lifespan_translator_cache +from .objects import Language from .processors import SentenceSplitter, Translator logger = logging.getLogger(__name__) @@ -82,35 +89,63 @@ async def translate_docs( ) -> int: es_client = lifespan_es_client() worker_config = cast(TranslationWorkerConfig, lifespan_worker_config()) - # TODO: make a generic fix using interceptors, - # see https://github.com/temporalio/sdk-python/issues/360 - source = _LANGUAGE_TYPE_ADAPTER.validate_python(source) - target = _LANGUAGE_TYPE_ADAPTER.validate_python(target) - if isinstance(config, dict): - config = TranslationConfig.model_validate(config) - # TODO: perform some caching here to avoid reloading - translator = config.to_translator() - sentence_splitter = config.to_sentence_splitter() # Load the translator first to install the - # SBD, then load the splitter logger.debug("loading %s -> %s translator...", source, target) - with translator.load(source, target=target, worker_config=worker_config): - logger.debug("loading %s sentence splitter...", source) - with sentence_splitter.load(source): - logger.info("translating %s batches...", len(batches)) - n_translated = await translate_docs_act( - batches, - project=project, - es_client=es_client, - progress=progress, - worker_config=worker_config, - translator=translator, - sentence_splitter=sentence_splitter, - ) - logger.info("done translating !") + translator_factory = partial( + _load_translator_from_config, + config=config.translator, + source=source, + target=target, + worker_config=worker_config, + ) + translator_key = config_cache_key(config.translator) + translator_cache = lifespan_translator_cache() + translator = translator_cache.get_or_cache_resource( + translator_key, translator_factory + ) + # SBD, then load the splitter + logger.debug("loading %s sentence splitter...", source) + splitter_factory = partial( + _load_splitter_from_config, config=config.sentence_splitter, language=source + ) + splitter_key = config_cache_key(config.sentence_splitter) + splitter_cache = lifespan_sentence_splitter_cache() + sentence_splitter = splitter_cache.get_or_cache_resource( + splitter_key, splitter_factory + ) + logger.info("translating %s batches...", len(batches)) + n_translated = await translate_docs_act( + batches, + project=project, + es_client=es_client, + progress=progress, + worker_config=worker_config, + translator=translator, + sentence_splitter=sentence_splitter, + ) + logger.info("done translating !") return n_translated +def _load_translator_from_config( + config: TranslatorConfig, + source: DatashareLanguage, + target: Language, + worker_config: TranslationWorkerConfig, +) -> Translator: + translator = Translator.from_config(config) + translator.load(source, target=target, worker_config=worker_config) + return translator + + +def _load_splitter_from_config( + config: SentenceSplitterConfig, language: DatashareLanguage +) -> SentenceSplitter: + splitter = SentenceSplitter.from_config(config) + splitter.load(language) + return splitter + + async def create_translation_batches_act( project: str, query: dict[str, Any], diff --git a/workers/translation-worker/translation_worker/config.py b/workers/translation-worker/translation_worker/config.py index 9c38e2ba..6db86977 100644 --- a/workers/translation-worker/translation_worker/config.py +++ b/workers/translation-worker/translation_worker/config.py @@ -1,7 +1,7 @@ from abc import ABC from typing import TYPE_CHECKING, ClassVar -from datashare_python.config import WorkerConfig +from datashare_python.config import ResourceCacheConfig, WorkerConfig from datashare_python.objects import BaseModel, DatashareModel from icij_common.pydantic_utils import make_enum_discriminator, tagged_union from icij_common.registrable import RegistrableConfig @@ -13,15 +13,26 @@ TranslationModel, ) -from .constants import DEFAULT_HUNYUAN_MODEL_REF, TorchDevice +from .constants import TorchDevice if TYPE_CHECKING: from translation_worker.processors import SentenceSplitter, Translator +DEFAULT_HUNYUAN_MODEL_REF = "tencent/Hunyuan-MT-Chimera-7B" + class _BaseProcessorConfig(BaseModel, RegistrableConfig, ABC): ... +class TranslationCache(BaseModel): + sentence_splitter: ResourceCacheConfig = ResourceCacheConfig( + size=1, exit_context_managers=True + ) + translator: ResourceCacheConfig = ResourceCacheConfig( + size=1, exit_context_managers=True + ) + + class C2TranslateConfig(DatashareModel): beam_size: int = 4 inter_threads: int = 1 @@ -37,6 +48,8 @@ class TranslationWorkerConfig(WorkerConfig): batches_per_worker: int = 10 es_buffer_size: int = 10 + cache: TranslationCache = Field(default_factory=TranslationCache) + c2_translate: C2TranslateConfig = Field(default_factory=C2TranslateConfig) @@ -46,11 +59,15 @@ class SentenceSplitterConfig(_BaseProcessorConfig): class DefaultSentenceSplitterConfig(SentenceSplitterConfig): - model: ClassVar[SentenceSplitterModel] = SentenceSplitterModel.DEFAULT + model: ClassVar[SentenceSplitterModel] = Field( + frozen=True, default=SentenceSplitterModel.DEFAULT + ) class ArgosSentenceSplitterConfig(SentenceSplitterConfig): - model: ClassVar[SentenceSplitterModel] = SentenceSplitterModel.ARGOS + model: ClassVar[SentenceSplitterModel] = Field( + frozen=True, default=SentenceSplitterModel.ARGOS + ) sentencizer: ArgosSentencizer = ArgosSentencizer.MINI_SBD @@ -61,13 +78,15 @@ class TranslatorConfig(_BaseProcessorConfig): _SentenceSplitterConfig = tagged_union( - SentenceSplitterConfig.__subclasses__(), lambda t: t.model + SentenceSplitterConfig.__subclasses__(), lambda t: t.model.default ) splitter_discriminator = make_enum_discriminator("model", SentenceSplitterModel) class ArgosTranslatorConfig(TranslatorConfig): - model: ClassVar[TranslationModel] = TranslationModel.ARGOS + model: ClassVar[TranslationModel] = Field( + frozen=True, default=TranslationModel.ARGOS + ) beam_size: int = 2 length_penalty: float = 0.2 @@ -75,7 +94,9 @@ class ArgosTranslatorConfig(TranslatorConfig): class HunyuanMtTranslatorConfig(TranslatorConfig): model_config = {"arbitrary_types_allowed": True} - model: ClassVar[TranslationModel] = TranslationModel.HUNYUAN + model: ClassVar[TranslationModel] = Field( + frozen=True, default=TranslationModel.HUNYUAN + ) model_ref: str = DEFAULT_HUNYUAN_MODEL_REF max_new_tokens: int = 2048 @@ -87,7 +108,9 @@ class HunyuanMtTranslatorConfig(TranslatorConfig): device_map: str = "auto" -_TranslatorConfig = tagged_union(TranslatorConfig.__subclasses__(), lambda t: t.model) +_TranslatorConfig = tagged_union( + TranslatorConfig.__subclasses__(), lambda t: t.model.default +) translator_discriminator = make_enum_discriminator("model", TranslationModel) diff --git a/workers/translation-worker/translation_worker/constants.py b/workers/translation-worker/translation_worker/constants.py index e4f1a588..79e88738 100644 --- a/workers/translation-worker/translation_worker/constants.py +++ b/workers/translation-worker/translation_worker/constants.py @@ -22,45 +22,3 @@ class TorchDevice(StrEnum): DOC_CONTENT_TEXT_LENGTH = "contentTextLength" TRANSLATION_DOC_SOURCES = [DOC_CONTENT, DOC_ROOT_ID, DOC_LANGUAGE] BATCHING_DOC_SOURCES = TRANSLATION_DOC_SOURCES[1:] + [DOC_CONTENT_TEXT_LENGTH] - -DEFAULT_HUNYUAN_MODEL_REF = "tencent/Hunyuan-MT-Chimera-7B" -HUNYUAN_SHARED_KEY = "hunyuan_translator" - -HUNYUAN_LANGUAGES = [ - # "en", - "zh", - # "fr", - # "pt", - # "es", - # "ja", - # "tr", - # "ru", - # "ar", - # "ko", - # "th", - # "it", - # "de", - # "vi", - # "ms", - # "id", - # "tl", - # "hi", - # "pl", - # "cs", - # "nl", - # "km", - # "my", - # "fa", - # "gu", - # "ur", - # "te", - # "mr", - # "he", - # "bn", - # "ta", - # "uk", - # "bo", - # "kk", - # "mn", - # "ug", -] diff --git a/workers/translation-worker/translation_worker/dependencies.py b/workers/translation-worker/translation_worker/dependencies.py index 2a95837b..61b00e1e 100644 --- a/workers/translation-worker/translation_worker/dependencies.py +++ b/workers/translation-worker/translation_worker/dependencies.py @@ -1,49 +1,60 @@ -import gc import logging -from contextlib import asynccontextmanager -from typing import Any +from _contextvars import ContextVar from datashare_python.dependencies import ( # noqa: F401 lifespan_es_client, - lifespan_shared_resources, lifespan_worker_config, set_es_client, - set_shared_resources, set_worker_config, ) from datashare_python.exceptions import DependencyInjectionError -from datashare_python.objects import Shared +from datashare_python.utils import SharedResources -from .constants import HUNYUAN_SHARED_KEY +from .config import TranslationWorkerConfig logger = logging.getLogger(__name__) -@asynccontextmanager -async def set_hunyuan_translator(**kwargs) -> None: # noqa: ARG001 - shared = Shared() - await set_shared_resources(shared) +_SENTENCE_SPLITTERS: ContextVar[SharedResources] = ContextVar("sentence_splitter_cache") +_TRANSLATORS: ContextVar[SharedResources] = ContextVar("translator_cache") + + +def set_sentence_splitter_cache( + worker_config: TranslationWorkerConfig, +) -> SharedResources: + cache = worker_config.cache.sentence_splitter.to_resource_cache() + _SENTENCE_SPLITTERS.set(cache) + return cache + + +def lifespan_sentence_splitter_cache() -> SharedResources: try: - yield - finally: - import torch # noqa: PLC0415 + return _SENTENCE_SPLITTERS.get() + except LookupError as e: + raise DependencyInjectionError("sentence preprocessors cache") from e + - model = await shared.async_pop_resource(HUNYUAN_SHARED_KEY, None) - if model is not None: - del model - if torch.cuda.is_available(): - torch.cuda.empty_cache() - gc.collect() +def set_translator_cache( + worker_config: TranslationWorkerConfig, +) -> SharedResources: + cache = worker_config.cache.translator.to_resource_cache() + _TRANSLATORS.set(cache) + return cache -def lifespan_hunyuan_translator() -> Any: +def lifespan_translator_cache() -> SharedResources: try: - return lifespan_shared_resources().get_resource(HUNYUAN_SHARED_KEY) - except (LookupError, KeyError) as e: - raise DependencyInjectionError("hunyuan translator") from e + return _TRANSLATORS.get() + except LookupError as e: + raise DependencyInjectionError("translators cache") from e REGISTRY = { - "translation.inference": [set_worker_config, set_es_client, set_hunyuan_translator], + "translation.inference": [ + set_worker_config, + set_es_client, + set_sentence_splitter_cache, + set_translator_cache, + ], "translation.io": [set_worker_config, set_es_client], } diff --git a/workers/translation-worker/translation_worker/objects.py b/workers/translation-worker/translation_worker/objects.py index 72b6eced..3e891163 100644 --- a/workers/translation-worker/translation_worker/objects.py +++ b/workers/translation-worker/translation_worker/objects.py @@ -16,8 +16,6 @@ ) from pydantic import TypeAdapter -from .constants import HUNYUAN_LANGUAGES - if TYPE_CHECKING: from argostranslate.sbd import ISentenceBoundaryDetectionModel @@ -56,10 +54,6 @@ def sentencizer_cls(self) -> type["ISentenceBoundaryDetectionModel"]: _LANGUAGE_TYPE_ADAPTER = TypeAdapter(Language) -_VALIDATED_HUNYUAN_LANGUAGES = { - _LANGUAGE_TYPE_ADAPTER.validate_python(lang) for lang in HUNYUAN_LANGUAGES -} - DocumentSearchQuery = dict[str, Any] DocId = str diff --git a/workers/translation-worker/translation_worker/processors.py b/workers/translation-worker/translation_worker/processors.py index 2b791606..1f6b5e8c 100644 --- a/workers/translation-worker/translation_worker/processors.py +++ b/workers/translation-worker/translation_worker/processors.py @@ -1,4 +1,5 @@ import logging +import sys from abc import abstractmethod from collections.abc import Iterable from contextlib import contextmanager @@ -24,12 +25,15 @@ def split_sentences_batch(self, batch: list[str]) -> list[list[str]]: @final @contextmanager - def load(self, language: "Language") -> Self: - with self: - self._load(language) + def load_cm(self, language: "Language") -> Self: + self.load(language) + try: yield self + finally: + self.__exit__(*sys.exc_info()) - def _load(self, language: "Language") -> Self: ... + @abstractmethod + def load(self, language: "Language") -> Self: ... @final def __enter__(self): @@ -62,7 +66,7 @@ def target(self) -> "Language": @contextmanager @final - def load( + def load_cm( self, source: "Language", *, @@ -71,11 +75,13 @@ def load( ) -> Self: if worker_config is None: worker_config = TranslationWorkerConfig() - with self: - self._load(source, target=target, worker_config=worker_config) + self.load(source, target=target, worker_config=worker_config) + try: yield self + finally: + self.__exit__(*sys.exc_info()) - def _load( + def load( self, source: "Language", *, diff --git a/workers/translation-worker/translation_worker/sentence_splitters.py b/workers/translation-worker/translation_worker/sentence_splitters.py index ac661152..a7ec7caa 100644 --- a/workers/translation-worker/translation_worker/sentence_splitters.py +++ b/workers/translation-worker/translation_worker/sentence_splitters.py @@ -14,6 +14,8 @@ class DefaultSentenceSplitter(SentenceSplitter): def __init__(self, config: DefaultSentenceSplitterConfig): self._config = config + def load(self, language: "Language") -> Self: ... + @classmethod def _from_config(cls, config: DefaultSentenceSplitterConfig) -> Self: return cls(config) @@ -28,7 +30,7 @@ def __init__(self, config: ArgosSentenceSplitterConfig): self._config = config self._inner = None - def _load(self, language: Language) -> Self: + def load(self, language: Language) -> Self: from argostranslate.package import get_installed_packages # noqa: PLC0415 if isinstance(language, DatashareLanguage): diff --git a/workers/translation-worker/translation_worker/translators/argos.py b/workers/translation-worker/translation_worker/translators/argos.py index e9a71285..d43e6af9 100644 --- a/workers/translation-worker/translation_worker/translators/argos.py +++ b/workers/translation-worker/translation_worker/translators/argos.py @@ -34,7 +34,7 @@ def __init__(self, config: "ArgosTranslatorConfig"): def _from_config(cls, config: "ArgosTranslatorConfig", **extras) -> Self: # noqa: ARG003 return cls(config) - def _load( + def load( self, source: Language, *, @@ -43,7 +43,7 @@ def _load( ) -> None: import ctranslate2 # noqa: PLC0415 - super()._load(source, target=target, worker_config=worker_config) + super().load(source, target=target, worker_config=worker_config) if isinstance(source, DatashareLanguage): source = LanguageAlpha2(source.alpha2) diff --git a/workers/translation-worker/translation_worker/translators/hunyuan.py b/workers/translation-worker/translation_worker/translators/hunyuan.py index b36197fc..0b933fc7 100644 --- a/workers/translation-worker/translation_worker/translators/hunyuan.py +++ b/workers/translation-worker/translation_worker/translators/hunyuan.py @@ -5,7 +5,6 @@ from datashare_python.objects import Language from ..config import HunyuanMtTranslatorConfig, TranslationWorkerConfig -from ..constants import HUNYUAN_SHARED_KEY from ..objects import TranslationModel from ..processors import Translator @@ -38,41 +37,21 @@ def _from_config( ) -> Self: return cls(config, device=device) - def _load( + def load( self, source: Language, *, target: Language, worker_config: TranslationWorkerConfig, ) -> None: - from datashare_python.dependencies import ( # noqa: PLC0415 - lifespan_shared_resources, - ) - from datashare_python.exceptions import ( # noqa: PLC0415 - DependencyInjectionError, - ) from transformers import AutoModelForCausalLM, AutoTokenizer # noqa: PLC0415 - super()._load(source, target=target, worker_config=worker_config) - - try: - shared = lifespan_shared_resources() - except DependencyInjectionError: - shared = None - - translator = ( - shared.get_resource(HUNYUAN_SHARED_KEY) if shared is not None else None + super().load(source, target=target, worker_config=worker_config) + translator = AutoModelForCausalLM.from_pretrained( + self._config.model_ref, + device_map=self._config.device_map, + torch_dtype=self._config.torch_dtype, ) - - if translator is None: - translator = AutoModelForCausalLM.from_pretrained( - self._config.model_ref, - device_map=self._config.device_map, - torch_dtype=self._config.torch_dtype, - ) - if shared is not None: - shared.set_resource(HUNYUAN_SHARED_KEY, translator) - self._translator = translator self._tokenizer = AutoTokenizer.from_pretrained(self._config.model_ref) self._device = next(self._translator.parameters()).device diff --git a/workers/translation-worker/translation_worker/workflows.py b/workers/translation-worker/translation_worker/workflows.py index 64fb8daf..d97acb76 100644 --- a/workers/translation-worker/translation_worker/workflows.py +++ b/workers/translation-worker/translation_worker/workflows.py @@ -1,7 +1,6 @@ import asyncio from datetime import timedelta -from datashare_python.objects import DatashareLanguage from icij_common.iter_utils import batches from temporalio import workflow @@ -10,29 +9,10 @@ from .activities import TranslationActivities from .config import ( - ArgosTranslatorConfig, - HunyuanMtTranslatorConfig, - TranslationConfig, TranslationWorkerConfig, ) from .constants import TRANSLATION_WORKFLOW_NAME, TaskQueue - from .objects import ( - _VALIDATED_HUNYUAN_LANGUAGES, - TranslationArgs, - TranslationResponse, - ) - - -def _set_config_translator( - source: DatashareLanguage, config: "TranslationConfig" -) -> "TranslationConfig": - source_code = source.alpha2 if hasattr(source, "alpha2") else str(source) - if source_code in _VALIDATED_HUNYUAN_LANGUAGES: - if not isinstance(config.translator, HunyuanMtTranslatorConfig): - return config.model_copy(update={"translator": HunyuanMtTranslatorConfig()}) - elif isinstance(config.translator, HunyuanMtTranslatorConfig): - return config.model_copy(update={"translator": ArgosTranslatorConfig()}) - return config + from .objects import TranslationArgs, TranslationResponse @workflow.defn(name=TRANSLATION_WORKFLOW_NAME) diff --git a/workers/translation-worker/uv.lock b/workers/translation-worker/uv.lock index 6db81e81..a476e38b 100644 --- a/workers/translation-worker/uv.lock +++ b/workers/translation-worker/uv.lock @@ -437,7 +437,7 @@ wheels = [ [[package]] name = "datashare-python" -version = "0.9.8" +version = "0.9.9" source = { editable = "../../datashare-python" } dependencies = [ { name = "aiohttp" },