test(yaml): add structural extraction coverage and update document tests (#2214)

- Add tests/test_yaml_extraction.py (classification, key nodes, contains edges,
  stable ids, scalar/empty documents).
- Drop .yaml/.yml from the oversized-document slicing parametrization: they are
  code now, not documents sent to the semantic pass.
- Point the language-override lifecycle test at .md so it still exercises a
  document override (yaml is no longer a document).

tests/test_language_override_lifecycle.py

@@ -186,11 +186,11 @@ def test_watch_declared_code_updates_graph_on_observer_thread(tmp_path, monkeypa
 
 def test_watch_document_override_marks_semantic_work_instead_of_rebuilding(tmp_path, monkeypatch):
     events = pytest.importorskip("watchdog.events")
-    (tmp_path / ".graphifyrc").write_text("language.tpl=yaml\n")
+    (tmp_path / ".graphifyrc").write_text("language.tpl=md\n")
     source = tmp_path / "notes.tpl"
 
     def change():
-        source.write_text("service: database\n")
+        source.write_text("# notes\n\nservice: database\n")
         return events.FileModifiedEvent(str(source))
 
     drive_watch(tmp_path, monkeypatch, [change])

tests/test_oversized_document_slicing.py

@@ -104,7 +104,7 @@ def test_slices_stay_within_the_cap(tmp_path):
 # End to end through the prompt builder
 # ---------------------------------------------------------------------------
 
-@pytest.mark.parametrize("ext", [".qmd", ".html", ".yaml", ".yml", ".skill"])
+@pytest.mark.parametrize("ext", [".qmd", ".html", ".skill"])
 def test_the_tail_of_a_big_document_reaches_the_prompt(tmp_path, ext):
     """The symptom a user would notice: content past 20k was invisible to the
     semantic pass, so nothing in the tail could ever become a node."""

tests/test_yaml_extraction.py

@@ -0,0 +1,63 @@
+"""Structural YAML extraction (#2214).
+
+YAML was previously classified as a document, so AST-only / code-only runs
+(``graphify update``) produced zero nodes from config-heavy repositories
+(Helm charts, Kubernetes manifests, CI pipelines). These tests pin the new
+structural path: ``.yaml``/``.yml`` classify as code and ``extract_yaml``
+emits deterministic file/key nodes.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+from graphify.detect import FileType, classify_file
+from graphify.extractors.yaml_config import extract_yaml
+
+
+def _write(path: Path, text: str) -> Path:
+    path.write_text(text, encoding="utf-8")
+    return path
+
+
+def test_yaml_and_yml_classify_as_code(tmp_path):
+    assert classify_file(_write(tmp_path / "values.yaml", "a: 1\n")) is FileType.CODE
+    assert classify_file(_write(tmp_path / "deploy.yml", "a: 1\n")) is FileType.CODE
+
+
+def test_extract_yaml_top_level_keys(tmp_path):
+    path = _write(
+        tmp_path / "values.yaml",
+        "replicaCount: 2\nimage:\n  tag: latest\nservice:\n  port: 80\n",
+    )
+    result = extract_yaml(path)
+
+    assert "error" not in result
+    labels = {n["label"] for n in result["nodes"]}
+    assert {"values.yaml", "replicaCount", "image", "service"} <= labels
+
+
+def test_extract_yaml_nested_keys_are_contains_edges(tmp_path):
+    path = _write(tmp_path / "values.yaml", "image:\n  tag: latest\n  pullPolicy: Always\n")
+    result = extract_yaml(path)
+
+    contains = [(e["source"], e["target"]) for e in result["edges"] if e["relation"] == "contains"]
+    assert any("image" in source and "tag" in target for source, target in contains)
+    assert any("image" in source and "pullpolicy" in target for source, target in contains)
+
+
+def test_extract_yaml_ids_are_stable_across_runs(tmp_path):
+    path = _write(tmp_path / "values.yaml", "a: 1\nb:\n  c: 2\n")
+    first = extract_yaml(path)
+    second = extract_yaml(path)
+    assert [n["id"] for n in first["nodes"]] == [n["id"] for n in second["nodes"]]
+
+
+def test_extract_yaml_scalar_document_is_skipped(tmp_path):
+    result = extract_yaml(_write(tmp_path / "scalar.yaml", "just-a-string\n"))
+    assert result.get("skipped") == "scalar yaml"
+
+
+def test_extract_yaml_empty_document(tmp_path):
+    result = extract_yaml(_write(tmp_path / "empty.yaml", ""))
+    assert result.get("skipped") == "scalar yaml"