Carry GroundRecall query bundles through doclift demo

src/didactopus/doclift_bundle_demo.py

@@ -18,9 +18,11 @@ def run_doclift_bundle_demo(
     pack_dir: str | Path,
     author: str = "doclift bundle import",
     license_name: str = "See source bundle metadata",
+    groundrecall_query_bundle_path: str | Path | None = None,
 ) -> dict:
     bundle_dir = Path(bundle_dir)
     pack_dir = Path(pack_dir)
+    explicit_groundrecall_path = Path(groundrecall_query_bundle_path) if groundrecall_query_bundle_path else None
 
     docs = adapt_documents(bundle_dir)
     if not docs:
@@ -33,6 +35,7 @@ def run_doclift_bundle_demo(
     concepts = [concept for concept in concepts if concept.id in lesson_concept_ids]
     ctx = RuleContext(course=merged, concepts=concepts)
     run_rules(ctx, build_default_rules(enable_projects=False, enable_review=False))
+    groundrecall_bundle = _load_groundrecall_query_bundle(bundle_dir, explicit_groundrecall_path)
 
     draft = build_draft_pack(
         merged,
@@ -41,6 +44,7 @@ def run_doclift_bundle_demo(
         license_name=license_name,
         review_flags=ctx.review_flags,
         conflicts=[],
+        groundrecall_query_bundle=groundrecall_bundle,
     )
     write_draft_pack(draft, pack_dir)
     write_source_corpus(merged, pack_dir)
@@ -58,11 +62,28 @@ def run_doclift_bundle_demo(
         "module_count": len(merged.modules),
         "concept_count": len(ctx.concepts),
         "review_flags": list(ctx.review_flags),
+        "groundrecall_bundle_included": bool(groundrecall_bundle),
     }
     (pack_dir / "doclift_bundle_summary.json").write_text(json.dumps(summary, indent=2), encoding="utf-8")
     return summary
 
 
+def _load_groundrecall_query_bundle(bundle_dir: Path, explicit_path: Path | None) -> dict | None:
+    candidates = []
+    if explicit_path is not None:
+        candidates.append(explicit_path)
+    candidates.extend(
+        [
+            bundle_dir / "groundrecall_query_bundle.json",
+            bundle_dir / "manifest.groundrecall_query_bundle.json",
+        ]
+    )
+    for path in candidates:
+        if path.exists():
+            return json.loads(path.read_text(encoding="utf-8"))
+    return None
+
+
 def main() -> None:
     parser = argparse.ArgumentParser(description="Generate a Didactopus draft pack from a doclift bundle.")
     parser.add_argument("bundle_dir")
@@ -70,6 +91,7 @@ def main() -> None:
     parser.add_argument("--course-title", required=True)
     parser.add_argument("--author", default="doclift bundle import")
     parser.add_argument("--license-name", default="See source bundle metadata")
+    parser.add_argument("--groundrecall-query-bundle", default=None)
     args = parser.parse_args()
 
     summary = run_doclift_bundle_demo(
@@ -78,6 +100,7 @@ def main() -> None:
         pack_dir=args.pack_dir,
         author=args.author,
         license_name=args.license_name,
+        groundrecall_query_bundle_path=args.groundrecall_query_bundle,
     )
     print(json.dumps(summary, indent=2))
 

src/didactopus/pack_emitter.py

@@ -71,8 +71,12 @@ def build_draft_pack(
     license_name: str,
     review_flags: list[str],
     conflicts: list[str] | None = None,
+    groundrecall_query_bundle: dict | None = None,
 ) -> DraftPack:
     pack_name = course.title.lower().replace(" ", "-")
+    supporting_artifacts = ["source_corpus.json", "knowledge_graph.json"]
+    if groundrecall_query_bundle is not None:
+        supporting_artifacts.append("groundrecall_query_bundle.json")
     pack = {
         "name": pack_name,
         "display_name": course.title,
@@ -87,7 +91,7 @@ def build_draft_pack(
         "overrides": [],
         "profile_templates": {},
         "cross_pack_links": [],
-        "supporting_artifacts": ["source_corpus.json", "knowledge_graph.json"],
+        "supporting_artifacts": supporting_artifacts,
     }
     concepts_yaml = {
         "concepts": [
@@ -134,6 +138,8 @@ def build_draft_pack(
             for src in course.source_records
         ],
     }
+    if groundrecall_query_bundle is not None:
+        attribution["groundrecall_query_bundle"] = groundrecall_query_bundle
     return DraftPack(
         pack=pack,
         concepts=concepts_yaml,
@@ -159,6 +165,11 @@ def write_draft_pack(pack: DraftPack, outdir: str | Path) -> None:
     conflict_lines = ["# Conflict Report", ""] + [f"- {flag}" for flag in pack.conflicts] if pack.conflicts else ["# Conflict Report", "", "- none"]
     (out / "conflict_report.md").write_text("\n".join(conflict_lines), encoding="utf-8")
     (out / "license_attribution.json").write_text(json.dumps(pack.attribution, indent=2), encoding="utf-8")
+    if isinstance(pack.attribution.get("groundrecall_query_bundle"), dict):
+        (out / "groundrecall_query_bundle.json").write_text(
+            json.dumps(pack.attribution["groundrecall_query_bundle"], indent=2),
+            encoding="utf-8",
+        )
 
 
 def write_source_corpus(course: NormalizedCourse, outdir: str | Path) -> None:

tests/test_doclift_bundle_demo.py

@@ -1,6 +1,8 @@
 from __future__ import annotations
 
+import json
 from pathlib import Path
+import shutil
 
 from didactopus.doclift_bundle_demo import run_doclift_bundle_demo
 
@@ -17,3 +19,23 @@ def test_doclift_bundle_demo_generates_pack(tmp_path: Path) -> None:
     assert (tmp_path / "pack" / "source_corpus.json").exists()
     assert (tmp_path / "pack" / "knowledge_graph.json").exists()
     assert (tmp_path / "pack" / "doclift_bundle_summary.json").exists()
+
+
+def test_doclift_bundle_demo_carries_groundrecall_query_bundle(tmp_path: Path) -> None:
+    bundle_dir = tmp_path / "bundle"
+    shutil.copytree(_fixture_bundle(), bundle_dir)
+    (bundle_dir / "groundrecall_query_bundle.json").write_text(
+        json.dumps(
+            {
+                "bundle_kind": "groundrecall_query_bundle",
+                "concept": {"concept_id": "concept::lecture-1", "title": "Lecture 1"},
+                "review_candidates": [{"candidate_id": "concept::lecture-1", "rationale": "Lecture 1 | lane=knowledge_capture | priority=20"}],
+            }
+        ),
+        encoding="utf-8",
+    )
+
+    summary = run_doclift_bundle_demo(bundle_dir, "Example Course", tmp_path / "pack")
+
+    assert summary["groundrecall_bundle_included"] is True
+    assert (tmp_path / "pack" / "groundrecall_query_bundle.json").exists()

tests/test_pack_emitter.py

@@ -27,3 +27,29 @@ def test_emit_pack(tmp_path: Path) -> None:
     assert (tmp_path / "review_report.md").exists()
     assert (tmp_path / "source_corpus.json").exists()
     assert (tmp_path / "knowledge_graph.json").exists()
+
+
+def test_emit_pack_can_write_groundrecall_query_bundle(tmp_path: Path) -> None:
+    course = parse_markdown_course(SAMPLE, "Sample Course")
+    concepts = extract_concept_candidates(course)
+    ctx = RuleContext(course=course, concepts=concepts)
+    run_rules(ctx, build_default_rules())
+    groundrecall_bundle = {
+        "bundle_kind": "groundrecall_query_bundle",
+        "concept": {"concept_id": "concept::topic-a", "title": "Topic A"},
+        "review_candidates": [{"candidate_id": "concept::topic-a", "rationale": "Topic A | lane=knowledge_capture | priority=20"}],
+    }
+    draft = build_draft_pack(
+        course,
+        ctx.concepts,
+        "Tester",
+        "REVIEW",
+        ctx.review_flags,
+        groundrecall_query_bundle=groundrecall_bundle,
+    )
+    write_draft_pack(draft, tmp_path)
+
+    pack_yaml = (tmp_path / "pack.yaml").read_text(encoding="utf-8")
+    bundle_payload = (tmp_path / "groundrecall_query_bundle.json").read_text(encoding="utf-8")
+    assert "groundrecall_query_bundle.json" in pack_yaml
+    assert '"bundle_kind": "groundrecall_query_bundle"' in bundle_payload