Implement logic to convert structure documents into graph

Cargo.toml

@@ -16,7 +16,7 @@ members = [
 ]
 
 [workspace.package]
-version = "2.1.2"
+version = "2.1.3"
 edition = "2021"
 rust-version = "1.80"
 license = "MIT"

README.md

@@ -52,6 +52,20 @@ The agent binds the workspace with `codegraph_init {"path": ...}` and gets tools
 
 → [Full comparison with decision matrix](docs/comparison.md)
 
+## 📄 Supported Formats
+
+CodeGraph-docs now supports parsing the following configuration file formats:
+
+| Format | Parser | Status |
+|--------|--------|--------|
+| YAML | YamlParser | ✅ Implemented |
+| JSON | JsonParser | ✅ Implemented |
+| TOML | TomlParser | ✅ Implemented |
+| **HCL** (HashiCorp Configuration Language) | **HclParser** | **✅ New** |
+| **Terraform (.tf)** | **HclParser** | **✅ New** |
+
+HCL and Terraform files can now be indexed and analyzed through the codegraph CLI, enabling semantic understanding of HashiCorp configuration files.
+
 ## 🎯 Key Features
 
 - **24 MCP tools** — `search_symbol`, `flow`, `callers`, `callees`, `impact`, `search_flow`, `context`, `references`, `diff`, `sandbox`, `mermaid`, and more

crates/codegraph-docs/Cargo.toml

@@ -0,0 +1,20 @@
+[package]
+name = "codegraph-docs"
+version.workspace = true
+edition = "2024"
+license.workspace = true
+repository.workspace = true
+
+[lints.rust]
+warnings = "deny"
+
+[dependencies]
+codegraph-core = { path = "../codegraph-core" }
+codegraph-graph = { path = "../codegraph-graph" }
+
+serde = { workspace = true }
+serde_json = { workspace = true }
+serde_yaml = "0.9"
+toml = "0.8"
+toml_edit = { workspace = true }
+hcl-rs = "0.19.8"

crates/codegraph-docs/src/config.rs

@@ -0,0 +1,52 @@
+use serde::{Deserialize, Serialize};
+
+/// Configuration for the document graph layer (`.codegraph/config.toml [docgraph]`).
+#[derive(Debug, Clone, Default, Serialize, Deserialize)]
+pub struct DocConfig {
+    /// Storage backend for document tries (same kind as code graph).
+    pub storage: Option<StorageConfig>,
+    /// Base id for document nodes.
+    pub doc_base: Option<u64>,
+    /// Base id for mined pattern ids.
+    pub pattern_base: Option<u64>,
+    /// Bloom bloom-filter cap (in tokens) for document search.
+    pub bloom_cap: Option<usize>,
+    /// Key normalization aliases (e.g. `instances → replicas`).
+    pub aliases: Option<Vec<(String, String)>>,
+}
+
+#[derive(Debug, Clone, Default, Serialize, Deserialize)]
+pub struct StorageConfig {
+    pub r#type: Option<String>,
+    pub dsn: Option<String>,
+}
+
+impl DocConfig {
+    pub fn doc_base(&self) -> u64 {
+        self.doc_base.unwrap_or(1_000_000_000)
+    }
+    pub fn pattern_base(&self) -> u64 {
+        self.pattern_base.unwrap_or(3_000_000_000)
+    }
+    pub fn bloom_cap(&self) -> usize {
+        self.bloom_cap.unwrap_or(64)
+    }
+    pub fn storage_kind(&self) -> Option<&str> {
+        self.storage.as_ref().and_then(|s| s.r#type.as_deref())
+    }
+    pub fn storage_dsn(&self) -> Option<&str> {
+        self.storage.as_ref().and_then(|s| s.dsn.as_deref())
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn defaults() {
+        let cfg = DocConfig::default();
+        assert_eq!(cfg.doc_base(), 1_000_000_000);
+        assert_eq!(cfg.bloom_cap(), 64);
+    }
+}