Implement repository size preflight checks

docs/CLI.md

@@ -2,6 +2,29 @@
 
 GasGuard supports command-line autocompletion for `bash`, `zsh`, and `fish`.
 
+## Scan resource limits
+
+Local scans check the number and combined size of supported source files before
+starting analysis. The defaults are 10,000 files and 100 MiB of source. Files
+under `node_modules`, `.git`, `target`, `dist`, and `build` are excluded by
+default. Configure limits in `gasguard.config.json` or override them per run:
+
+```json
+{
+	"scan": {
+		"maxFiles": 10000,
+		"maxBytes": 104857600
+	}
+}
+```
+
+```sh
+gasguard scan ./contracts --max-files 5000 --max-bytes 52428800
+```
+
+Discovery stops as soon as a configured limit is exceeded. The CLI reports
+`FILE_LIMIT_EXCEEDED` or `SIZE_LIMIT_EXCEEDED` and does not start analysis.
+
 ## Installation
 
 To load completions in your current shell session, add the following to your profile (`~/.bashrc`, `~/.zshrc`, or your fish config):

packages/cli/src/commands/scan.spec.ts

@@ -114,4 +114,42 @@ describe("runScan", () => {
       code: "SIZE_LIMIT_EXCEEDED",
     });
   });
+
+  it("accepts repositories exactly at the file and source-byte limits", async () => {
+    const sourceFiles = ["contract A {}", "contract B {}"];
+    await fs.writeFile(path.join(directory, "A.sol"), sourceFiles[0]);
+    await fs.writeFile(path.join(directory, "B.sol"), sourceFiles[1]);
+    const totalBytes = sourceFiles.reduce(
+      (total, source) => total + Buffer.byteLength(source),
+      0,
+    );
+
+    await expect(
+      collectScannableFiles(directory, {
+        maxFiles: sourceFiles.length,
+        maxBytes: totalBytes,
+      }),
+    ).resolves.toHaveLength(sourceFiles.length);
+  });
+
+  it("rejects the first file or byte beyond the configured preflight limit", async () => {
+    const sourceFiles = ["contract A {}", "contract B {}"];
+    await fs.writeFile(path.join(directory, "A.sol"), sourceFiles[0]);
+    await fs.writeFile(path.join(directory, "B.sol"), sourceFiles[1]);
+    const totalBytes = sourceFiles.reduce(
+      (total, source) => total + Buffer.byteLength(source),
+      0,
+    );
+
+    await expect(
+      collectScannableFiles(directory, { maxFiles: sourceFiles.length - 1 }),
+    ).rejects.toMatchObject<Partial<RepositoryAnalysisError>>({
+      code: "FILE_LIMIT_EXCEEDED",
+    });
+    await expect(
+      collectScannableFiles(directory, { maxBytes: totalBytes - 1 }),
+    ).rejects.toMatchObject<Partial<RepositoryAnalysisError>>({
+      code: "SIZE_LIMIT_EXCEEDED",
+    });
+  });
 });

packages/cli/src/commands/scan.ts

@@ -18,7 +18,7 @@ import type {
   Finding as EngineFinding,
 } from "../../../../libs/engine/core/analyzer-interface";
 import { ScanWatcher } from "../../../../src/analysis/watch/watcher";
-import { glob } from "glob";
+import { globIterate } from "glob";
 import { loadCliConfig } from "../config/config-loader";
 import {
   finishProgress,
@@ -132,7 +132,11 @@ export const scanCommand = new Command("scan")
         process.exitCode = 130;
         return;
       }
-      console.error(chalk.red(`Error during scan: ${error}`));
+      const message =
+        error instanceof RepositoryAnalysisError
+          ? `${error.code}: ${error.message}`
+          : String(error);
+      console.error(chalk.red(`Error during scan: ${message}`));
       process.exitCode = 1;
     } finally {
       process.removeListener("SIGINT", abort);
@@ -298,16 +302,16 @@ export async function collectScannableFiles(
   const patterns = limits.include?.length
     ? limits.include
     : ["**/*.sol", "**/*.vy", "**/*.rs"];
-  const discovered = await glob(patterns, {
+  const discovered = globIterate(patterns, {
     cwd: path.resolve(dirPath),
     absolute: true,
     nodir: true,
     ignore: [...defaultExcludes, ...(limits.exclude ?? [])],
   });
-  for (const file of discovered
-    .filter((candidate) => extensions.includes(path.extname(candidate)))
-    .sort()) {
+  const seen = new Set<string>();
+  for await (const file of discovered) {
     checkCancelled();
+    if (!extensions.includes(path.extname(file)) || seen.has(file)) continue;
     if (files.length >= maxFiles) {
       throw new RepositoryAnalysisError(
         `Repository exceeds the ${maxFiles} file limit.`,
@@ -323,8 +327,9 @@ export async function collectScannableFiles(
     }
     totalBytes += fileStats.size;
     files.push(file);
+    seen.add(file);
   }
-  return files;
+  return files.sort();
 }
 
 function asConfigObject(value: unknown): Record<string, unknown> {