test: cover the real /ingest pipeline end-to-end (slow)

Author: NorthernLightx

tests/unit/test_ingest_route.py

@@ -1,13 +1,18 @@
 """POST /ingest (ADR 0029): flag gate, validation, and the happy-path append."""
 
+from pathlib import Path
 from unittest import mock
 
+import fitz
 import pytest
 from fastapi.testclient import TestClient
 
 from src.api.deps import get_settings
 from src.api.main import create_app
 from src.config.settings import Settings
+from src.rag.bm25 import Bm25Index
+from src.rag.vectorstore import QdrantVectorStore
+from tests.fakes import FakeEmbedder
 
 
 def _app(enable_upload: bool) -> TestClient:
@@ -79,3 +84,42 @@ async def fake_ingest(**kwargs: object) -> mock.Mock:
     assert resp.status_code == 422
     assert "internal detail" not in resp.text  # no exception/path leakage to the client
     assert "Could not parse" in resp.text
+
+
+@pytest.mark.slow
+async def test_ingest_route_real_pipeline(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
+    """The real /ingest pipeline, unmocked: Docling parse + embed + Qdrant upsert
+    + the live chunks_by_id append, then retrievable. Uses in-memory Qdrant + a
+    fake embedder (the repo's e2e pattern), so it needs no services — only the
+    real Docling parse, hence `slow`."""
+    embedder = FakeEmbedder(dim=8)
+    vectorstore = QdrantVectorStore(url=":memory:", collection_name="ingest_slow", dim=8)
+    await vectorstore.ensure_collection()
+    bm25 = Bm25Index()
+    chunks: dict[str, object] = {}
+    monkeypatch.setattr(
+        "src.api.routes.ingest.get_corpus_handles", lambda: (embedder, vectorstore, bm25)
+    )
+    monkeypatch.setattr("src.api.routes.ingest.get_chunks", lambda: chunks)
+
+    doc = fitz.open()
+    doc.new_page().insert_text(
+        (72, 72),
+        "The zorblax7731 calibration constant equals 88.5 in this synthetic test document.",
+        fontsize=11,
+    )
+    pdf = tmp_path / "zorblax.pdf"
+    doc.save(pdf)
+    doc.close()
+
+    resp = _app(enable_upload=True).post(
+        "/ingest", files={"file": ("zorblax.pdf", pdf.read_bytes(), "application/pdf")}
+    )
+
+    assert resp.status_code == 200, resp.text
+    body = resp.json()
+    assert body["chunks_added"] >= 1
+    assert body["paper_id"] == "zorblax"
+    assert len(chunks) == body["chunks_added"]  # route appended to the live index
+    hits = bm25.search("zorblax7731", top_k=5)  # the uploaded doc is now retrievable
+    assert hits and any(h.chunk_id in chunks for h in hits)