Tighten asset linking and batch conversion tests

src/doclift/convert.py

@@ -9,6 +9,7 @@ from .legacy_doc import (
     extract_references,
     extract_tables,
     extract_title,
+    link_related_assets,
     normalize_text_preserve_layout,
     render_markdown,
     run_catdoc,
@@ -32,7 +33,7 @@ def convert_doc(source_path: Path, out_root: Path, figure_assets: list | None =
     layout = build_layout_manifest(layout_body)
     table_refs = extract_references(body, r"\bTable\s+\d+\b")
     figure_refs = extract_references(body, r"\b(?:Fig\.?\s*[\d.]+|Figure\s+[\d.]+)\b")
-    related_assets = list(figure_assets or [])
+    related_assets = link_related_assets(figure_refs, list(figure_assets or []))
 
     doc_out = _document_output_dir(out_root, source_path, title)
     doc_out.mkdir(parents=True, exist_ok=True)
@@ -93,6 +94,15 @@ def convert_directory(source_root: Path, out_root: Path, asset_root: Path | None
             "summary": {
                 "documents_with_tables": sum(1 for bundle in bundles if bundle.table_count > 0),
                 "documents_with_figure_references": sum(1 for bundle in bundles if bundle.figure_reference_count > 0),
+                "documents": [
+                    {
+                        "document_id": bundle.document_id,
+                        "title": bundle.title,
+                        "table_count": bundle.table_count,
+                        "figure_reference_count": bundle.figure_reference_count,
+                    }
+                    for bundle in bundles
+                ],
             }
         },
     )

src/doclift/legacy_doc.py

@@ -221,6 +221,28 @@ def collect_figure_assets(root: Path) -> list[FigureAsset]:
     return assets
 
 
+def link_related_assets(figure_refs: list[str], figure_assets: list[FigureAsset]) -> list[FigureAsset]:
+    if not figure_refs:
+        return []
+
+    matched: list[FigureAsset] = []
+    seen: set[str] = set()
+    ref_keys: set[str] = set()
+    for ref in figure_refs:
+        key = slugify(ref.replace("Figure", "Fig").replace("figure", "fig"))
+        ref_keys.add(key)
+
+    for asset in figure_assets:
+        asset_key = slugify(asset.name.rsplit(".", 1)[0])
+        for ref_key in ref_keys:
+            if ref_key and ref_key in asset_key:
+                if asset.asset_id not in seen:
+                    seen.add(asset.asset_id)
+                    matched.append(asset)
+                break
+    return matched
+
+
 def build_layout_manifest(layout_body: str) -> list[dict]:
     manifest: list[dict] = []
     for line_no, line in enumerate(layout_body.splitlines(), start=1):

tests/test_convert.py

@@ -0,0 +1,48 @@
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+from doclift import convert as convert_module
+
+
+def test_convert_directory_writes_manifest_and_conversion_report(tmp_path: Path, monkeypatch) -> None:
+    source_root = tmp_path / "src"
+    asset_root = tmp_path / "assets"
+    out_root = tmp_path / "out"
+    source_root.mkdir()
+    asset_root.mkdir()
+    (source_root / "sample.doc").write_text("stub", encoding="utf-8")
+    (asset_root / "Fig. 5.1.bmp").write_text("img", encoding="utf-8")
+
+    sample_text = "\n".join(
+        [
+            "Lecture 1. Example legacy document",
+            "",
+            "See Fig. 5.1 and Table 1.",
+            "",
+            "Table 1. Example caption",
+            "",
+            "Metric\tRest\tSwim",
+            "O2\t1.0\t2.0",
+        ]
+    )
+
+    monkeypatch.setattr(convert_module, "run_catdoc", lambda path: sample_text)
+
+    report = convert_module.convert_directory(source_root, out_root, asset_root=asset_root)
+
+    assert report.document_count == 1
+    manifest = json.loads((out_root / "manifest.json").read_text(encoding="utf-8"))
+    conversion_report = json.loads((out_root / "conversion_report.json").read_text(encoding="utf-8"))
+    figures_payload = json.loads(
+        (out_root / "documents" / "sample-lecture-1-example-legacy-document" / "document.figures.json").read_text(
+            encoding="utf-8"
+        )
+    )
+
+    assert manifest["document_count"] == 1
+    assert conversion_report["summary"]["documents_with_tables"] == 1
+    assert conversion_report["summary"]["documents_with_figure_references"] == 1
+    assert figures_payload["figure_references"] == ["Fig. 5.1"]
+    assert len(figures_payload["related_assets"]) == 1

tests/test_legacy_doc.py

@@ -1,4 +1,4 @@
-from doclift.legacy_doc import extract_references, extract_tables
+from doclift.legacy_doc import FigureAsset, extract_references, extract_tables, link_related_assets
 
 
 def test_extract_references_dedupes() -> None:
@@ -22,3 +22,26 @@ def test_extract_tables_parses_tabbed_rows() -> None:
     assert tables[0].caption == "Table 1. Example caption"
     assert tables[0].column_count_guess == 3
     assert tables[0].parsed_rows[1] == ["O2", "1.0", "2.0"]
+
+
+def test_link_related_assets_matches_explicit_figure_refs() -> None:
+    assets = [
+        FigureAsset(
+            asset_id="a1",
+            path="/tmp/Fig. 5.1.bmp",
+            relative_path="vol/Fig. 5.1.bmp",
+            name="Fig. 5.1.bmp",
+            container="vol",
+            looks_like_figure=True,
+        ),
+        FigureAsset(
+            asset_id="a2",
+            path="/tmp/Slide 1.jpg",
+            relative_path="vol/Slide 1.jpg",
+            name="Slide 1.jpg",
+            container="vol",
+            looks_like_figure=False,
+        ),
+    ]
+    matched = link_related_assets(["Fig. 5.1"], assets)
+    assert [asset.asset_id for asset in matched] == ["a1"]