Complete outstanding ingestion and document UX work

2026-08-22 18:54:07 +01:00
parent 932e889ee8
commit 5b5a9ecbc3
24 changed files with 696 additions and 41 deletions
@@ -0,0 +1,102 @@
+"""Focused tests for document metadata lookup and aggregation."""
+
+import pytest
+
+from kb import database
+from kb.routes import documents
+
+
+@pytest.fixture
+def document_db(tmp_path, monkeypatch):
+    db_path = tmp_path / "kb.db"
+    conn = database.get_connection(db_path)
+    database.init_schema(conn, 3)
+    conn.execute(
+        """
+        INSERT INTO documents(title, source_path, content_hash, doc_type, original_filename)
+        VALUES ('Vehicle Guide', '/data/staging/random.pdf', 'hash', 'pdf', 'M38T_manual.pdf')
+        """
+    )
+    doc_id = conn.execute("SELECT id FROM documents").fetchone()["id"]
+    conn.execute(
+        "INSERT INTO chunks(document_id, chunk_index, text, enriched_text) VALUES (?, 0, 'body', 'body')",
+        (doc_id,),
+    )
+    conn.commit()
+    conn.close()
+    monkeypatch.setattr(documents.cfg, "data_dir", tmp_path)
+    return doc_id
+
+
+@pytest.mark.asyncio
+async def test_list_filters_title_and_original_filename(document_db):
+    by_title = await documents.list_documents(
+        type=None, tags=None, title="vehicle", filename=None
+    )
+    by_filename = await documents.list_documents(
+        type=None, tags=None, title=None, filename="m38t"
+    )
+
+    assert [item["id"] for item in by_title] == [document_db]
+    assert by_filename[0]["original_filename"] == "M38T_manual.pdf"
+
+
+@pytest.mark.asyncio
+async def test_info_can_omit_chunks_without_losing_count(document_db):
+    result = await documents.get_document(document_db, include_chunks=False)
+
+    assert result["chunk_count"] == 1
+    assert result["chunks"] == []
+
+
+@pytest.mark.asyncio
+async def test_find_aggregates_chunk_hits_by_document(monkeypatch):
+    class Connection:
+        def close(self):
+            pass
+
+    monkeypatch.setattr(documents, "get_connection", lambda _path: Connection())
+    monkeypatch.setattr(
+        documents,
+        "hybrid_search",
+        lambda *_args, **_kwargs: {
+            "results": [
+                {
+                    "document_id": 7,
+                    "title": "Guide",
+                    "doc_type": "pdf",
+                    "source_path": "/staging/file",
+                    "original_filename": "guide.pdf",
+                    "tags": [],
+                    "score": 0.8,
+                    "text": "best hit",
+                },
+                {
+                    "document_id": 7,
+                    "title": "Guide",
+                    "doc_type": "pdf",
+                    "source_path": "/staging/file",
+                    "original_filename": "guide.pdf",
+                    "tags": [],
+                    "score": 0.7,
+                    "text": "second hit",
+                },
+            ]
+        },
+    )
+
+    result = await documents.find_documents(
+        documents.DocumentFindRequest(query="guide")
+    )
+
+    assert result == [{
+        "document_id": 7,
+        "title": "Guide",
+        "doc_type": "pdf",
+        "source_path": "/staging/file",
+        "original_filename": "guide.pdf",
+        "tags": [],
+        "score": 0.8,
+        "hit_count": 2,
+        "top_chunk": "best hit",
+    }]