fix incremental no-cluster graph updates (#1350)

Harden incremental no-cluster updates: fixes empty-write graph wipeout on no-op update --no-cluster (#1347) and git-hook subdir path resolution (#1348). Complementary to #1317. Validated: full suite 2107 passed, no-op re-run no longer wipes graph. Thanks @pkudinov.

graphify/__main__.py

@@ -4431,6 +4431,26 @@ def main() -> None:
             # anchors emitted per importing file, #1327).
             from graphify.build import dedupe_edges as _dedupe_edges, dedupe_nodes as _dedupe_nodes
             from graphify.export import backup_if_protected as _backup
+            if (
+                incremental_mode
+                and not code_files
+                and not semantic_files
+                and not deleted_files
+                and not pg_result.get("nodes")
+                and not pg_result.get("edges")
+                and not cargo_result.get("nodes")
+                and not cargo_result.get("edges")
+            ):
+                print(
+                    "[graphify extract] no incremental changes detected "
+                    "(--no-cluster); outputs left untouched."
+                )
+                try:
+                    _save_manifest(_manifest_files, manifest_path=str(manifest_path), kind="both", root=target)
+                except Exception as exc:
+                    print(f"[graphify extract] warning: could not write manifest: {exc}", file=sys.stderr)
+                sys.exit(0)
+
             merged["nodes"] = _dedupe_nodes(merged["nodes"])
             merged["edges"] = _dedupe_edges(merged["edges"])
             _backup(graphify_out)

graphify/watch.py

@@ -205,6 +205,36 @@ def _report_root_label(watch_path: Path) -> str:
     return Path.cwd().name if watch_path == Path(".") else str(watch_path)
 
 
+def _is_relative_to(path: Path, root: Path) -> bool:
+    try:
+        path.relative_to(root)
+        return True
+    except ValueError:
+        return False
+
+
+def _changed_path_candidates(raw: Path, *, change_root: Path, watch_root: Path) -> list[Path]:
+    """Return plausible absolute locations for a hook-provided changed path.
+
+    Git hooks pass paths relative to the repository root. Watch callers may
+    also pass paths relative to the watched root. Keep both interpretations so
+    a graph rooted at ``src`` accepts ``src/app.py`` and ``app.py``.
+    """
+    if raw.is_absolute():
+        return [raw.resolve()]
+
+    candidates: list[Path] = []
+    seen: set[str] = set()
+    for base in (change_root, watch_root):
+        cand = (base / raw).resolve()
+        key = os.fspath(cand)
+        if key in seen:
+            continue
+        seen.add(key)
+        candidates.append(cand)
+    return candidates
+
+
 def _relativize_source_files(payload: dict, root: Path) -> None:
     for bucket in ("nodes", "edges", "hyperedges"):
         for item in payload.get(bucket, []):
@@ -477,18 +507,47 @@ def _rebuild_code(
         # extract only changed-and-still-existing files. Deleted paths are
         # tracked separately so their stale nodes can be evicted below.
         deleted_paths: set[str] = set()
+        def _add_deleted_source(path: Path) -> None:
+            for root in (project_root, watch_root):
+                deleted_paths.add(_nsf(str(path), str(root)) or str(path))
+
         if changed_paths is not None:
             code_set = {p.resolve() for p in code_files}
             wanted: list[Path] = []
+            change_root = Path.cwd().resolve()
             for raw in changed_paths:
-                cand = (watch_root / raw).resolve() if not raw.is_absolute() else raw.resolve()
-                if cand.exists() and cand in code_set:
-                    wanted.append(cand)
-                else:
-                    # File was deleted, renamed away, or filtered out by detect
-                    # (e.g. .gitignore, vendored). Either way, evict any
-                    # preserved nodes that still claim this source path.
-                    deleted_paths.add(_nsf(str(cand), str(project_root)) or str(cand))
+                candidates = _changed_path_candidates(
+                    raw,
+                    change_root=change_root,
+                    watch_root=watch_root,
+                )
+                tracked = next((cand for cand in candidates if cand.exists() and cand in code_set), None)
+                if tracked is not None:
+                    if tracked not in wanted:
+                        wanted.append(tracked)
+                    continue
+
+                existing_in_root = next(
+                    (
+                        cand for cand in candidates
+                        if cand.exists() and _is_relative_to(cand, watch_root)
+                    ),
+                    None,
+                )
+                if existing_in_root is not None:
+                    # The path exists under the watched root but detect filtered
+                    # it out. Evict any stale nodes that still claim it.
+                    _add_deleted_source(existing_in_root)
+                    continue
+
+                deleted_in_root = next(
+                    (cand for cand in candidates if _is_relative_to(cand, watch_root)),
+                    None,
+                )
+                if deleted_in_root is not None:
+                    # File was deleted or renamed away inside the watched root.
+                    # Evict preserved nodes that still claim this source path.
+                    _add_deleted_source(deleted_in_root)
             if not wanted and not deleted_paths:
                 print("[graphify watch] No tracked code files in change set - skipping rebuild.")
                 return True
@@ -522,7 +581,8 @@ def _rebuild_code(
                 evict_sources: set[str] = set(deleted_paths)
                 if changed_paths is not None:
                     for p in extract_targets:
-                        evict_sources.add(_nsf(str(p), str(project_root)) or str(p))
+                        for root in (project_root, watch_root):
+                            evict_sources.add(_nsf(str(p), str(root)) or str(p))
                 else:
                     # Full re-extraction: reconcile against current code files to
                     # evict nodes from files deleted since the last run (#1007).

tests/test_incremental.py

@@ -72,3 +72,27 @@ def test_no_incremental_without_manifest(tmp_path):
     # which pytest derives from the test name and contains "incremental".
     assert "incremental update" not in r.stdout.lower()
     assert "incremental scan" not in r.stdout.lower()
+
+
+def test_extract_no_cluster_incremental_noop_preserves_existing_graph(tmp_path):
+    """#1347: no-op incremental no-cluster extract must not overwrite graph.json."""
+    project = tmp_path / "project"
+    project.mkdir()
+    (project / "app.py").write_text(
+        "def alpha():\n    return 1\n", encoding="utf-8"
+    )
+
+    first = _run(["extract", str(project), "--no-cluster"], tmp_path)
+    assert first.returncode == 0, first.stderr
+    graph_path = project / "graphify-out" / "graph.json"
+    before_text = graph_path.read_text(encoding="utf-8")
+    before = json.loads(before_text)
+    assert before.get("nodes"), "first run should produce a non-empty code graph"
+
+    second = _run(["extract", str(project), "--no-cluster"], tmp_path)
+    assert second.returncode == 0, second.stderr
+
+    after_text = graph_path.read_text(encoding="utf-8")
+    after = json.loads(after_text)
+    assert after.get("nodes"), "no-op incremental run must not empty the graph"
+    assert after_text == before_text

tests/test_watch.py

@@ -654,6 +654,40 @@ def test_rebuild_code_prunes_deleted_file_nodes(tmp_path):
         os.chdir(cwd)
 
 
+def test_rebuild_code_accepts_repo_relative_changed_path_for_subdir_root(tmp_path):
+    """#1348: git-hook paths are repo-root-relative even when the graph root is a subdir."""
+    from graphify.watch import _rebuild_code
+
+    src = tmp_path / "src"
+    src.mkdir()
+    app = src / "app.py"
+    app.write_text("def old_name():\n    return 1\n", encoding="utf-8")
+
+    cwd = os.getcwd()
+    try:
+        os.chdir(tmp_path)
+        assert _rebuild_code(Path("src"), no_cluster=True, acquire_lock=False) is True
+        graph_path = src / "graphify-out" / "graph.json"
+        before = json.loads(graph_path.read_text(encoding="utf-8"))
+        assert "old_name()" in {n.get("label") for n in before.get("nodes", [])}
+
+        app.write_text("def new_name():\n    return 2\n", encoding="utf-8")
+        assert _rebuild_code(
+            Path("src"),
+            changed_paths=[Path("src/app.py")],
+            no_cluster=True,
+            acquire_lock=False,
+            force=True,
+        ) is True
+
+        after = json.loads(graph_path.read_text(encoding="utf-8"))
+        labels = {n.get("label") for n in after.get("nodes", [])}
+        assert "old_name()" not in labels
+        assert "new_name()" in labels
+    finally:
+        os.chdir(cwd)
+
+
 # --- #1059: pending-changes queue prevents commit drops under lock contention ---