added docs

backend/tests/test_document_rag.py

@@ -0,0 +1,160 @@
+from __future__ import annotations
+
+import sys
+import types
+from io import BytesIO
+
+import pytest
+from fastapi import FastAPI
+from fastapi.testclient import TestClient
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+from sqlalchemy.pool import StaticPool
+
+from app.auth.deps import get_current_user
+from app.api.routes.documents import router as documents_router
+from app.db.base import Base, get_db
+from app.db.models import Document, DocumentChunk
+from app.db.models import User
+from app.rag.chunker import chunk_blocks
+from app.rag.extractors import ExtractedBlock, extract_document
+from app.rag.ingest import _chunk_payload
+
+
+def test_extract_text_document_blocks() -> None:
+    extracted = extract_document("1. Общие положения\n\nТекст пункта".encode(), "gost.txt")
+
+    assert extracted.mime_type == "text/plain"
+    assert extracted.text.startswith("1. Общие положения")
+    assert [block.paragraph_index for block in extracted.blocks] == [0, 2]
+
+
+def test_extract_docx_document_sections() -> None:
+    docx = pytest.importorskip("docx")
+    document = docx.Document()
+    document.add_heading("1. Термины", level=1)
+    document.add_paragraph("Определение из ГОСТ.")
+    buffer = BytesIO()
+    document.save(buffer)
+
+    extracted = extract_document(buffer.getvalue(), "terms.docx")
+
+    assert extracted.mime_type.endswith("wordprocessingml.document")
+    assert any(block.section_path == "1. Термины" for block in extracted.blocks)
+    assert "Определение из ГОСТ." in extracted.text
+
+
+def test_extract_pdf_document_pages(monkeypatch: pytest.MonkeyPatch) -> None:
+    class FakePage:
+        def __init__(self, text: str) -> None:
+            self._text = text
+
+        def extract_text(self) -> str:
+            return self._text
+
+    class FakePdfReader:
+        def __init__(self, _: BytesIO) -> None:
+            self.pages = [FakePage("Страница 1"), FakePage("Страница 2")]
+
+    fake_pypdf = types.SimpleNamespace(PdfReader=FakePdfReader)
+    monkeypatch.setitem(sys.modules, "pypdf", fake_pypdf)
+
+    extracted = extract_document(b"%PDF fake", "doc.pdf")
+
+    assert extracted.mime_type == "application/pdf"
+    assert [block.page_start for block in extracted.blocks] == [1, 2]
+    assert "Страница 2" in extracted.text
+
+
+def test_chunk_blocks_keeps_source_metadata() -> None:
+    chunks = chunk_blocks(
+        [
+            ExtractedBlock(text="1. Область применения", page_start=3, page_end=3),
+            ExtractedBlock(text="Требование к оформлению.", page_start=3, page_end=3),
+        ],
+        chunk_size=200,
+    )
+
+    assert len(chunks) == 1
+    assert chunks[0].page_start == 3
+    assert chunks[0].source_ref == "стр. 3"
+    assert "Требование" in chunks[0].content
+
+
+def test_qdrant_payload_contains_citation_fields() -> None:
+    doc = Document(id=7, user_id=2, title="ГОСТ", filename="gost.pdf", mime_type="application/pdf")
+    row = DocumentChunk(
+        id=11,
+        document_id=7,
+        chunk_index=4,
+        content="Пункт 5.3",
+        source_filename="gost.pdf",
+        mime_type="application/pdf",
+        page_start=12,
+        page_end=13,
+        section_title="5. Требования",
+        section_path="5. Требования > 5.3 Оформление",
+        paragraph_index=9,
+        source_ref="стр. 12-13, 5. Требования > 5.3 Оформление",
+    )
+
+    payload = _chunk_payload(doc, row)
+
+    assert payload["document_id"] == 7
+    assert payload["chunk_id"] == 11
+    assert payload["page_start"] == 12
+    assert payload["source_ref"].startswith("стр. 12-13")
+
+
+def test_upload_document_endpoint_stores_chunks(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setenv("RAG_ENABLED", "false")
+    from app.config import get_settings
+
+    get_settings.cache_clear()
+    engine = create_engine(
+        "sqlite://",
+        connect_args={"check_same_thread": False},
+        poolclass=StaticPool,
+    )
+    TestingSessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
+    Base.metadata.create_all(bind=engine)
+
+    db = TestingSessionLocal()
+    user = User(username="owner", display_name="Owner", api_token_hash="hash", is_active=True)
+    db.add(user)
+    db.commit()
+    db.refresh(user)
+    user_id = user.id
+    db.close()
+
+    app = FastAPI()
+    app.include_router(documents_router, prefix="/api/v1")
+
+    def override_get_db():
+        session = TestingSessionLocal()
+        try:
+            yield session
+        finally:
+            session.close()
+
+    def override_current_user():
+        session = TestingSessionLocal()
+        try:
+            return session.get(User, user_id)
+        finally:
+            session.close()
+
+    app.dependency_overrides[get_db] = override_get_db
+    app.dependency_overrides[get_current_user] = override_current_user
+
+    with TestClient(app) as client:
+        response = client.post(
+            "/api/v1/documents/upload",
+            files={"file": ("tz.txt", b"1. Scope\n\nRequirement text", "text/plain")},
+        )
+
+    assert response.status_code == 200
+    payload = response.json()["document"]
+    assert payload["filename"] == "tz.txt"
+    assert payload["chunk_count"] == 1
+    assert payload["index_status"] == "stored"