feat(extraction): make MAX_FILE_SIZE configurable via CODEGRAPH_MAX_FILE_SIZE (#1016)

README.md

@@ -599,6 +599,9 @@ What it skips out of the box:
 - **Anything in your `.gitignore`** — honored in git repos via git, and in
   non-git projects by reading `.gitignore` directly (root and nested).
 - **Files larger than 1 MB** — generated bundles, minified JS, vendored blobs.
+  Override the threshold (in bytes) with `CODEGRAPH_MAX_FILE_SIZE` for repos with
+  legitimately large hand-written sources — e.g. `CODEGRAPH_MAX_FILE_SIZE=5242880`
+  for 5 MB. Invalid or non-positive values fall back to the 1 MB default.
 
 To keep something else out, add it to `.gitignore`. To pull a default-excluded
 directory back **in** (say you really do want a vendored dependency indexed),

__tests__/resolve-max-file-size.test.ts

@@ -0,0 +1,34 @@
+import { afterEach, describe, expect, it } from 'vitest';
+import { resolveMaxFileSize } from '../src/extraction';
+
+const DEFAULT = 1024 * 1024;
+const ENV = 'CODEGRAPH_MAX_FILE_SIZE';
+
+describe('resolveMaxFileSize', () => {
+  afterEach(() => {
+    delete process.env[ENV];
+  });
+
+  it('falls back to the 1 MB default when the env var is unset', () => {
+    delete process.env[ENV];
+    expect(resolveMaxFileSize()).toBe(DEFAULT);
+  });
+
+  it('honours a valid positive override', () => {
+    process.env[ENV] = String(5 * 1024 * 1024);
+    expect(resolveMaxFileSize()).toBe(5 * 1024 * 1024);
+  });
+
+  it('floors fractional byte counts', () => {
+    process.env[ENV] = '2097152.9';
+    expect(resolveMaxFileSize()).toBe(2 * 1024 * 1024);
+  });
+
+  it.each(['', 'not-a-number', '0', '-1', 'NaN', 'Infinity'])(
+    'falls back to the default for invalid value %j',
+    (raw) => {
+      process.env[ENV] = raw;
+      expect(resolveMaxFileSize()).toBe(DEFAULT);
+    },
+  );
+});

src/extraction/index.ts

@@ -114,7 +114,23 @@ export function hashContent(content: string): string {
  * vendored blobs blow the WASM heap and the worker-recycle budget for no useful
  * symbols. 1 MB covers essentially all hand-written source.
  */
-const MAX_FILE_SIZE = 1024 * 1024;
+const DEFAULT_MAX_FILE_SIZE = 1024 * 1024;
+
+/**
+ * Resolve the max-file-size threshold (bytes), allowing an override via the
+ * `CODEGRAPH_MAX_FILE_SIZE` environment variable for repos with legitimately
+ * large hand-written sources. Falls back to the 1 MB default when the value is
+ * unset, non-numeric, or not a positive number.
+ */
+export function resolveMaxFileSize(): number {
+  const raw = process.env.CODEGRAPH_MAX_FILE_SIZE;
+  if (raw === undefined || raw === '') return DEFAULT_MAX_FILE_SIZE;
+  const parsed = Number(raw);
+  if (!Number.isFinite(parsed) || parsed <= 0) return DEFAULT_MAX_FILE_SIZE;
+  return Math.floor(parsed);
+}
+
+const MAX_FILE_SIZE = resolveMaxFileSize();
 
 /**
  * Directory names that are dependency, build, cache, or tooling output across the