@optave/codegraph 3.8.0 → 3.9.0

package/src/domain/graph/cycles.ts

@@ -1,9 +1,14 @@
-import { tarjan } from '../../graph/algorithms/tarjan.js';
-import { buildDependencyGraph } from '../../graph/builders/dependency.js';
-import { CodeGraph } from '../../graph/model.js';
+import { getCallableNodes, getCallEdges, getFileNodesAll, getImportEdges } from '../../db/index.js';
 import { loadNative } from '../../infrastructure/native.js';
+import { isTestFile } from '../../infrastructure/test-filter.js';
 import type { BetterSqlite3Database } from '../../types.js';
 
+/**
+ * Find cycles using Tarjan's SCC algorithm.
+ *
+ * Builds a label-based adjacency list directly from DB rows — no intermediate
+ * CodeGraph construction. This is O(V + E) with minimal memory overhead.
+ */
 export function findCycles(
   db: BetterSqlite3Database,
   opts: { fileLevel?: boolean; noTests?: boolean } = {},
@@ -11,40 +16,122 @@ export function findCycles(
   const fileLevel = opts.fileLevel !== false;
   const noTests = opts.noTests || false;
 
-  const graph = buildDependencyGraph(db, { fileLevel, noTests });
+  const edges: Array<{ source: string; target: string }> = [];
+  const seen = new Set<string>();
 
-  const idToLabel = new Map<string, string>();
-  for (const [id, attrs] of graph.nodes()) {
-    if (fileLevel) {
-      idToLabel.set(id, attrs.file as string);
-    } else {
-      idToLabel.set(id, `${attrs.label}|${attrs.file}`);
+  if (fileLevel) {
+    let nodes = getFileNodesAll(db);
+    if (noTests) nodes = nodes.filter((n) => !isTestFile(n.file));
+    const nodeIds = new Set<number>();
+    const idToFile = new Map<number, string>();
+    for (const n of nodes) {
+      nodeIds.add(n.id);
+      idToFile.set(n.id, n.file);
+    }
+    for (const e of getImportEdges(db)) {
+      if (!nodeIds.has(e.source_id) || !nodeIds.has(e.target_id)) continue;
+      if (e.source_id === e.target_id) continue;
+      const src = idToFile.get(e.source_id)!;
+      const tgt = idToFile.get(e.target_id)!;
+      const key = `${src}\0${tgt}`;
+      if (seen.has(key)) continue;
+      seen.add(key);
+      edges.push({ source: src, target: tgt });
+    }
+  } else {
+    let nodes = getCallableNodes(db);
+    if (noTests) nodes = nodes.filter((n) => !isTestFile(n.file));
+    const nodeIds = new Set<number>();
+    const idToLabel = new Map<number, string>();
+    for (const n of nodes) {
+      nodeIds.add(n.id);
+      idToLabel.set(n.id, `${n.name}|${n.file}`);
+    }
+    for (const e of getCallEdges(db)) {
+      if (!nodeIds.has(e.source_id) || !nodeIds.has(e.target_id)) continue;
+      if (e.source_id === e.target_id) continue;
+      const src = idToLabel.get(e.source_id)!;
+      const tgt = idToLabel.get(e.target_id)!;
+      const key = `${src}\0${tgt}`;
+      if (seen.has(key)) continue;
+      seen.add(key);
+      edges.push({ source: src, target: tgt });
     }
   }
 
-  const edges = graph.toEdgeArray().map((e) => ({
-    source: idToLabel.get(e.source) ?? e.source,
-    target: idToLabel.get(e.target) ?? e.target,
-  }));
-
   const native = loadNative();
   if (native) {
     return native.detectCycles(edges) as string[][];
   }
 
-  const labelGraph = new CodeGraph();
-  for (const { source, target } of edges) {
-    labelGraph.addEdge(source, target);
-  }
-  return tarjan(labelGraph);
+  return tarjanFromEdges(edges);
 }
 
 export function findCyclesJS(edges: Array<{ source: string; target: string }>): string[][] {
-  const graph = new CodeGraph();
+  return tarjanFromEdges(edges);
+}
+
+/**
+ * Run Tarjan's SCC on a flat edge list. Returns SCCs with length > 1 (cycles).
+ * Uses a simple adjacency-list Map instead of a full CodeGraph.
+ */
+function tarjanFromEdges(edges: Array<{ source: string; target: string }>): string[][] {
+  const adj = new Map<string, string[]>();
+  const allNodes = new Set<string>();
   for (const { source, target } of edges) {
-    graph.addEdge(source, target);
+    allNodes.add(source);
+    allNodes.add(target);
+    let list = adj.get(source);
+    if (!list) {
+      list = [];
+      adj.set(source, list);
+    }
+    list.push(target);
   }
-  return tarjan(graph);
+
+  let index = 0;
+  const stack: string[] = [];
+  const onStack = new Set<string>();
+  const indices = new Map<string, number>();
+  const lowlinks = new Map<string, number>();
+  const sccs: string[][] = [];
+
+  function strongconnect(v: string): void {
+    indices.set(v, index);
+    lowlinks.set(v, index);
+    index++;
+    stack.push(v);
+    onStack.add(v);
+
+    const successors = adj.get(v);
+    if (successors) {
+      for (const w of successors) {
+        if (!indices.has(w)) {
+          strongconnect(w);
+          lowlinks.set(v, Math.min(lowlinks.get(v)!, lowlinks.get(w)!));
+        } else if (onStack.has(w)) {
+          lowlinks.set(v, Math.min(lowlinks.get(v)!, indices.get(w)!));
+        }
+      }
+    }
+
+    if (lowlinks.get(v) === indices.get(v)) {
+      const scc: string[] = [];
+      let w: string | undefined;
+      do {
+        w = stack.pop()!;
+        onStack.delete(w);
+        scc.push(w);
+      } while (w !== v);
+      if (scc.length > 1) sccs.push(scc);
+    }
+  }
+
+  for (const id of allNodes) {
+    if (!indices.has(id)) strongconnect(id);
+  }
+
+  return sccs;
 }
 
 export function formatCycles(cycles: string[][]): string {

package/src/domain/graph/resolve.ts

@@ -565,7 +565,7 @@ export function resolveImportsBatch(
       // Native resolver's .js → .ts remap fails on unnormalized paths —
       // apply JS-side fallback (same fix as resolveImportPath).
       const resolved = remapJsToTs(normalized, rootDir);
-      map.set(`${r.fromFile}|${r.importSource}`, resolved);
+      map.set(`${normalizePath(r.fromFile)}|${r.importSource}`, resolved);
     }
     return map;
   } catch (e) {

package/src/domain/graph/watcher.ts

@@ -18,29 +18,8 @@ function isTrackedExt(filePath: string): boolean {
   return EXTENSIONS.has(path.extname(filePath));
 }
 
-export async function watchProject(rootDir: string, opts: { engine?: string } = {}): Promise<void> {
-  const dbPath = path.join(rootDir, '.codegraph', 'graph.db');
-  if (!fs.existsSync(dbPath)) {
-    throw new DbError('No graph.db found. Run `codegraph build` first.', { file: dbPath });
-  }
-
-  const db = openDb(dbPath);
-  initSchema(db);
-  const engineOpts: import('../../types.js').EngineOpts = {
-    engine: (opts.engine || 'auto') as import('../../types.js').EngineMode,
-    dataflow: false,
-    ast: false,
-  };
-  const { name: engineName, version: engineVersion } = getActiveEngine(engineOpts);
-  info(`Watch mode using ${engineName} engine${engineVersion ? ` (v${engineVersion})` : ''}`);
-
-  const cache = createParseTreeCache();
-  info(
-    cache
-      ? 'Incremental parsing enabled (native tree cache)'
-      : 'Incremental parsing unavailable (full re-parse)',
-  );
-
+/** Prepare all SQL statements needed by the watcher's incremental rebuild. */
+function prepareWatcherStatements(db: ReturnType<typeof openDb>): IncrementalStmts {
   const stmts = {
     insertNode: db.prepare(
       'INSERT OR IGNORE INTO nodes (name, kind, file, line, end_line) VALUES (?, ?, ?, ?, ?)',
@@ -67,7 +46,6 @@ export async function watchProject(rootDir: string, opts: { engine?: string } =
     listSymbols: db.prepare("SELECT name, kind, line FROM nodes WHERE file = ? AND kind != 'file'"),
   };
 
-  // Use named params for statements needing the same value twice
   const origDeleteEdges = db.prepare(
     `DELETE FROM edges WHERE source_id IN (SELECT id FROM nodes WHERE file = @f) OR target_id IN (SELECT id FROM nodes WHERE file = @f)`,
   );
@@ -79,96 +57,224 @@ export async function watchProject(rootDir: string, opts: { engine?: string } =
     get: (f: string) => origCountEdges.get({ f }) as { c: number } | undefined,
   };
 
+  return stmts as IncrementalStmts;
+}
+
+/** Rebuild result shape from rebuildFile. */
+interface RebuildResult {
+  file: string;
+  deleted?: boolean;
+  event: string;
+  symbolDiff: unknown;
+  nodesBefore: number;
+  nodesAfter: number;
+  nodesAdded: number;
+  nodesRemoved: number;
+  edgesAdded: number;
+}
+
+/** Process a batch of pending file changes: rebuild, journal, and log. */
+async function processPendingFiles(
+  files: string[],
+  db: ReturnType<typeof openDb>,
+  rootDir: string,
+  stmts: IncrementalStmts,
+  engineOpts: import('../../types.js').EngineOpts,
+  cache: ReturnType<typeof createParseTreeCache>,
+): Promise<void> {
+  const results: RebuildResult[] = [];
+  for (const filePath of files) {
+    const result = (await rebuildFile(db, rootDir, filePath, stmts, engineOpts, cache, {
+      diffSymbols: diffSymbols as (old: unknown[], new_: unknown[]) => unknown,
+    })) as RebuildResult | null;
+    if (result) results.push(result);
+  }
+
+  if (results.length > 0) {
+    writeJournalAndChangeEvents(rootDir, results);
+  }
+
+  logRebuildResults(results);
+}
+
+/** Write journal entries and change events for processed files. */
+function writeJournalAndChangeEvents(rootDir: string, updates: RebuildResult[]): void {
+  const entries = updates.map((r) => ({
+    file: r.file,
+    deleted: r.deleted || false,
+  }));
+  try {
+    appendJournalEntries(rootDir, entries);
+  } catch (e: unknown) {
+    debug(`Journal write failed (non-fatal): ${(e as Error).message}`);
+  }
+
+  const changeEvents = updates.map((r) =>
+    buildChangeEvent(r.file, r.event, r.symbolDiff, {
+      nodesBefore: r.nodesBefore,
+      nodesAfter: r.nodesAfter,
+      edgesAdded: r.edgesAdded,
+    }),
+  );
+  try {
+    appendChangeEvents(rootDir, changeEvents);
+  } catch (e: unknown) {
+    debug(`Change event write failed (non-fatal): ${(e as Error).message}`);
+  }
+}
+
+/** Log rebuild results to the user. */
+function logRebuildResults(updates: RebuildResult[]): void {
+  for (const r of updates) {
+    const nodeDelta = r.nodesAdded - r.nodesRemoved;
+    const nodeStr = nodeDelta >= 0 ? `+${nodeDelta}` : `${nodeDelta}`;
+    if (r.deleted) {
+      info(`Removed: ${r.file} (-${r.nodesRemoved} nodes)`);
+    } else {
+      info(`Updated: ${r.file} (${nodeStr} nodes, +${r.edgesAdded} edges)`);
+    }
+  }
+}
+
+/** Recursively collect tracked source files for stat-based polling. */
+function collectTrackedFiles(dir: string, result: string[]): void {
+  let entries: fs.Dirent[];
+  try {
+    entries = fs.readdirSync(dir, { withFileTypes: true });
+  } catch {
+    return;
+  }
+  for (const entry of entries) {
+    if (IGNORE_DIRS.has(entry.name) || entry.name.startsWith('.')) continue;
+    const full = path.join(dir, entry.name);
+    if (entry.isDirectory()) {
+      collectTrackedFiles(full, result);
+    } else if (EXTENSIONS.has(path.extname(entry.name))) {
+      result.push(full);
+    }
+  }
+}
+
+export async function watchProject(
+  rootDir: string,
+  opts: { engine?: string; poll?: boolean; pollInterval?: number } = {},
+): Promise<void> {
+  const dbPath = path.join(rootDir, '.codegraph', 'graph.db');
+  if (!fs.existsSync(dbPath)) {
+    throw new DbError('No graph.db found. Run `codegraph build` first.', { file: dbPath });
+  }
+
+  const db = openDb(dbPath);
+  initSchema(db);
+  const engineOpts: import('../../types.js').EngineOpts = {
+    engine: (opts.engine || 'auto') as import('../../types.js').EngineMode,
+    dataflow: false,
+    ast: false,
+  };
+  const { name: engineName, version: engineVersion } = getActiveEngine(engineOpts);
+  info(`Watch mode using ${engineName} engine${engineVersion ? ` (v${engineVersion})` : ''}`);
+
+  const cache = createParseTreeCache();
+  info(
+    cache
+      ? 'Incremental parsing enabled (native tree cache)'
+      : 'Incremental parsing unavailable (full re-parse)',
+  );
+
+  const stmts = prepareWatcherStatements(db);
+
   const pending = new Set<string>();
   let timer: ReturnType<typeof setTimeout> | null = null;
   const DEBOUNCE_MS = 300;
 
-  async function processPending(): Promise<void> {
-    const files = [...pending];
-    pending.clear();
-
-    const results: Array<{
-      file: string;
-      deleted?: boolean;
-      event: string;
-      symbolDiff: unknown;
-      nodesBefore: number;
-      nodesAfter: number;
-      nodesAdded: number;
-      nodesRemoved: number;
-      edgesAdded: number;
-    }> = [];
-    for (const filePath of files) {
-      const result = (await rebuildFile(
-        db,
-        rootDir,
-        filePath,
-        stmts as IncrementalStmts,
-        engineOpts,
-        cache,
-        {
-          diffSymbols: diffSymbols as (old: unknown[], new_: unknown[]) => unknown,
-        },
-      )) as (typeof results)[number] | null;
-      if (result) results.push(result);
-    }
-    const updates = results;
+  const usePoll = opts.poll ?? process.platform === 'win32';
+  const POLL_INTERVAL_MS = opts.pollInterval ?? 2000;
 
-    // Append processed files to journal for Tier 0 detection on next build
-    if (updates.length > 0) {
-      const entries = updates.map((r) => ({
-        file: r.file,
-        deleted: r.deleted || false,
-      }));
-      try {
-        appendJournalEntries(rootDir, entries);
-      } catch (e: unknown) {
-        debug(`Journal write failed (non-fatal): ${(e as Error).message}`);
-      }
+  info(`Watching ${rootDir} for changes${usePoll ? ' (polling mode)' : ''}...`);
+  info('Press Ctrl+C to stop.');
+
+  let cleanup: () => void;
+
+  if (usePoll) {
+    // Polling mode: avoids native OS file watchers (NtNotifyChangeDirectoryFileEx)
+    // which can crash ReFS drivers on Windows Dev Drives.
+    const mtimeMap = new Map<string, number>();
 
-      const changeEvents = updates.map((r) =>
-        buildChangeEvent(r.file, r.event, r.symbolDiff, {
-          nodesBefore: r.nodesBefore,
-          nodesAfter: r.nodesAfter,
-          edgesAdded: r.edgesAdded,
-        }),
-      );
+    // Seed initial mtimes
+    const initial: string[] = [];
+    collectTrackedFiles(rootDir, initial);
+    for (const f of initial) {
       try {
-        appendChangeEvents(rootDir, changeEvents);
-      } catch (e: unknown) {
-        debug(`Change event write failed (non-fatal): ${(e as Error).message}`);
+        mtimeMap.set(f, fs.statSync(f).mtimeMs);
+      } catch {
+        /* deleted between collect and stat */
       }
     }
+    info(`Polling ${initial.length} tracked files every ${POLL_INTERVAL_MS}ms`);
+
+    const pollTimer = setInterval(() => {
+      const current: string[] = [];
+      collectTrackedFiles(rootDir, current);
+      const currentSet = new Set(current);
 
-    for (const r of updates) {
-      const nodeDelta = r.nodesAdded - r.nodesRemoved;
-      const nodeStr = nodeDelta >= 0 ? `+${nodeDelta}` : `${nodeDelta}`;
-      if (r.deleted) {
-        info(`Removed: ${r.file} (-${r.nodesRemoved} nodes)`);
-      } else {
-        info(`Updated: ${r.file} (${nodeStr} nodes, +${r.edgesAdded} edges)`);
+      // Detect modified or new files
+      for (const f of current) {
+        try {
+          const mtime = fs.statSync(f).mtimeMs;
+          const prev = mtimeMap.get(f);
+          if (prev === undefined || mtime !== prev) {
+            mtimeMap.set(f, mtime);
+            pending.add(f);
+          }
+        } catch {
+          /* deleted between collect and stat */
+        }
       }
-    }
-  }
 
-  info(`Watching ${rootDir} for changes...`);
-  info('Press Ctrl+C to stop.');
+      // Detect deleted files
+      for (const f of mtimeMap.keys()) {
+        if (!currentSet.has(f)) {
+          mtimeMap.delete(f);
+          pending.add(f);
+        }
+      }
 
-  const watcher = fs.watch(rootDir, { recursive: true }, (_eventType, filename) => {
-    if (!filename) return;
-    if (shouldIgnore(filename)) return;
-    if (!isTrackedExt(filename)) return;
+      if (pending.size > 0) {
+        if (timer) clearTimeout(timer);
+        timer = setTimeout(async () => {
+          const files = [...pending];
+          pending.clear();
+          await processPendingFiles(files, db, rootDir, stmts, engineOpts, cache);
+        }, DEBOUNCE_MS);
+      }
+    }, POLL_INTERVAL_MS);
 
-    const fullPath = path.join(rootDir, filename);
-    pending.add(fullPath);
+    cleanup = () => clearInterval(pollTimer);
+  } else {
+    // Native OS watcher — efficient but can trigger ReFS crashes on Windows Dev Drives.
+    // Use --poll if you experience BSOD/HYPERVISOR_ERROR on ReFS volumes.
+    const watcher = fs.watch(rootDir, { recursive: true }, (_eventType, filename) => {
+      if (!filename) return;
+      if (shouldIgnore(filename)) return;
+      if (!isTrackedExt(filename)) return;
 
-    if (timer) clearTimeout(timer);
-    timer = setTimeout(processPending, DEBOUNCE_MS);
-  });
+      const fullPath = path.join(rootDir, filename);
+      pending.add(fullPath);
+
+      if (timer) clearTimeout(timer);
+      timer = setTimeout(async () => {
+        const files = [...pending];
+        pending.clear();
+        await processPendingFiles(files, db, rootDir, stmts, engineOpts, cache);
+      }, DEBOUNCE_MS);
+    });
+
+    cleanup = () => watcher.close();
+  }
 
   process.on('SIGINT', () => {
     info('Stopping watcher...');
-    watcher.close();
+    cleanup();
     // Flush any pending file paths to journal before exit
     if (pending.size > 0) {
       const entries = [...pending].map((filePath) => ({