feat: enhance /ingest endpoint to support file paths and file uploads

Author: Bessouat40

README.md

@@ -208,7 +208,8 @@ raglight serve --port 8080
 |---|---|---|---|
 | `GET` | `/health` | — | `{"status": "ok"}` |
 | `POST` | `/generate` | `{"question": "..."}` | `{"answer": "..."}` |
-| `POST` | `/ingest` | `{"data_path": "...", "github_url": "...", "github_branch": "main"}` | `{"message": "..."}` |
+| `POST` | `/ingest` | `{"data_path": "...", "file_paths": [...], "github_url": "...", "github_branch": "main"}` | `{"message": "..."}` |
+| `POST` | `/ingest/upload` | `multipart/form-data` — champ `files` (un ou plusieurs fichiers) | `{"message": "..."}` |
 | `GET` | `/collections` | — | `{"collections": [...]}` |
 
 The interactive API documentation (Swagger UI) is automatically available at `http://localhost:8000/docs`.
@@ -234,6 +235,11 @@ curl -X POST http://localhost:8000/ingest \
   -H "Content-Type: application/json" \
   -d '{"github_url": "https://github.com/Bessouat40/RAGLight", "github_branch": "main"}'
 
+# Upload files directly (multipart)
+curl -X POST http://localhost:8000/ingest/upload \
+  -F "files=@./rapport.pdf" \
+  -F "files=@./notes.txt"
+
 # List collections
 curl http://localhost:8000/collections
 ```

pyproject.toml

@@ -38,6 +38,7 @@ dependencies = [
     "nltk>=3.9.2",
     "fastapi>=0.115.0",
     "uvicorn[standard]>=0.34.0",
+    "python-multipart>=0.0.9",
 ]
 [project.scripts]
 raglight = "raglight.cli.main:app"

src/raglight/config/settings.py

@@ -34,7 +34,7 @@ def setup_logging() -> None:
     MISTRAL_API_KEY = os.environ.get("MISTRAL_API_KEY", "")
     LMSTUDIO = "LmStudio"
     HUGGINGFACE = "HuggingFace"
-    DEFAULT_LLM = "llama3"
+    DEFAULT_LLM = "llama3.2:1b"
     DEFAULT_OPENAI_CLIENT = os.environ.get(
         "OPENAI_CLIENT_URL", "https://api.openai.com/v1"
     )

src/raglight/rag/builder.py

@@ -60,6 +60,7 @@ def with_embeddings(self, type: str, **kwargs) -> Builder:
         """
         logging.info("⏳ Creating an Embeddings Model...")
         if type == Settings.HUGGINGFACE:
+            kwargs.pop("api_base", None)
             self.embeddings = HuggingfaceEmbeddingsModel(**kwargs)
         elif type == Settings.OLLAMA:
             self.embeddings = OllamaEmbeddingsModel(**kwargs)

tests/tests_api/test_router.py

@@ -69,7 +69,8 @@ def test_ingest_local(client, pipeline):
 def test_ingest_missing_params(client):
     response = client.post("/ingest", json={})
     assert response.status_code == 400
-    assert "data_path" in response.json()["detail"] or "github_url" in response.json()["detail"]
+    detail = response.json()["detail"]
+    assert any(k in detail for k in ("data_path", "file_paths", "github_url"))
 
 
 def test_ingest_github(client, pipeline):
@@ -94,6 +95,56 @@ def test_ingest_github(client, pipeline):
 
 # ── /collections ──────────────────────────────────────────────────────────────
 
+def test_ingest_file_paths(client, pipeline):
+    vector_store = pipeline.get_vector_store.return_value
+    vector_store.add_documents.return_value = None
+    vector_store.add_class_documents.return_value = None
+    vector_store._flatten_metadata.side_effect = lambda docs: docs
+    vector_store._process_file = lambda fp, factory, flatten: (["chunk"], [])
+
+    with patch("os.path.isfile", return_value=True):
+        response = client.post(
+            "/ingest",
+            json={"file_paths": ["/some/file.pdf", "/some/other.txt"]},
+        )
+
+    assert response.status_code == 200
+    assert "success" in response.json()["message"].lower()
+
+
+def test_ingest_upload(client, pipeline):
+    vector_store = pipeline.get_vector_store.return_value
+    vector_store.add_documents.return_value = None
+    vector_store.add_class_documents.return_value = None
+    vector_store._flatten_metadata.side_effect = lambda docs: docs
+    vector_store._process_file = lambda fp, factory, flatten: (["chunk"], [])
+
+    response = client.post(
+        "/ingest/upload",
+        files=[
+            ("files", ("doc1.txt", b"hello world", "text/plain")),
+            ("files", ("doc2.txt", b"another doc", "text/plain")),
+        ],
+    )
+    assert response.status_code == 200
+    assert "2 file(s)" in response.json()["message"]
+
+
+def test_ingest_upload_no_files(client):
+    response = client.post("/ingest/upload", files=[])
+    assert response.status_code in (400, 422)
+
+
+def test_ingest_file_paths_missing_file(client, pipeline):
+    with patch("os.path.isfile", return_value=False):
+        response = client.post(
+            "/ingest",
+            json={"file_paths": ["/nonexistent/file.pdf"]},
+        )
+    assert response.status_code == 500
+    assert "not found" in response.json()["detail"].lower()
+
+
 def test_collections(client, pipeline):
     vector_store = pipeline.get_vector_store.return_value
     vector_store.get_available_collections.return_value = ["default", "project_x"]

tests/tests_api/test_server_config.py

@@ -10,13 +10,10 @@
 
 
 def test_defaults():
-    with patch.dict(os.environ, {}, clear=False):
-        # Remove RAGLIGHT_* vars if present
-        env_overrides = {k: "" for k in os.environ if k.startswith("RAGLIGHT_")}
-        for k in env_overrides:
-            os.environ.pop(k, None)
+    clean_env = {k: v for k, v in os.environ.items() if not k.startswith("RAGLIGHT_")}
+    with patch.dict(os.environ, clean_env, clear=True):
         cfg = ServerConfig()
-    assert cfg.llm_model == "llama3"
+    assert cfg.llm_model == Settings.DEFAULT_LLM
     assert cfg.llm_provider == Settings.OLLAMA
     assert cfg.embeddings_model == Settings.DEFAULT_EMBEDDINGS_MODEL
     assert cfg.embeddings_provider == Settings.HUGGINGFACE