feat: implement prometheus exporter

Author: NicolasIRAGNE

.gitignore

@@ -10,6 +10,7 @@ Thumbs.db
 
 # Python virtual-envs & tooling
 .venv*/
+venv/
 .python-version
 __pycache__/
 *.egg-info/

pyproject.toml

@@ -16,6 +16,7 @@ dependencies = [
     "tiktoken>=0.7.0",  # Support for o200k_base encoding
     "typing_extensions>= 4.0.0; python_version < '3.10'",
     "uvicorn>=0.11.7",  # Minimum safe release (https://osv.dev/vulnerability/PYSEC-2020-150)
+    "prometheus-client",
 ]
 
 license = {file = "LICENSE"}

requirements.txt

@@ -8,3 +8,4 @@ slowapi
 starlette>=0.40.0  # Vulnerable to https://osv.dev/vulnerability/GHSA-f96h-pmfr-66vw
 tiktoken>=0.7.0  # Support for o200k_base encoding
 uvicorn>=0.11.7  # Vulnerable to https://osv.dev/vulnerability/PYSEC-2020-150
+prometheus-client

src/server/main.py

@@ -12,7 +12,7 @@
 from slowapi.errors import RateLimitExceeded
 from starlette.middleware.trustedhost import TrustedHostMiddleware
 
-from server.routers import dynamic, index, ingest
+from server.routers import dynamic, index, ingest, metrics
 from server.server_config import templates
 from server.server_utils import lifespan, limiter, rate_limit_exception_handler
 
@@ -159,6 +159,8 @@ def openapi_json() -> JSONResponse:
 
 
 # Include routers for modular endpoints
+if os.getenv("GITINGEST_PROMETHEUS_TOKEN") is not None:
+    app.include_router(metrics)
 app.include_router(index)
 app.include_router(ingest)
 app.include_router(dynamic)

src/server/routers/__init__.py

@@ -3,5 +3,6 @@
 from server.routers.dynamic import router as dynamic
 from server.routers.index import router as index
 from server.routers.ingest import router as ingest
+from server.routers.metrics import router as metrics
 
-__all__ = ["dynamic", "index", "ingest"]
+__all__ = ["dynamic", "index", "ingest", "metrics"]

src/server/routers/ingest.py

@@ -2,13 +2,16 @@
 
 from fastapi import APIRouter, HTTPException, Request, status
 from fastapi.responses import FileResponse, JSONResponse
+from prometheus_client import Counter
 
 from gitingest.config import TMP_BASE_PATH
 from server.models import IngestRequest
 from server.routers_utils import COMMON_INGEST_RESPONSES, _perform_ingestion
 from server.server_config import MAX_DISPLAY_SIZE
 from server.server_utils import limiter
 
+ingest_counter = Counter("gitingest_ingest_total", "Number of ingests", ["status", "url"])
+
 router = APIRouter()
 
 
@@ -33,13 +36,16 @@ async def api_ingest(
     - **JSONResponse**: Success response with ingestion results or error response with appropriate HTTP status code
 
     """
-    return await _perform_ingestion(
+    response = await _perform_ingestion(
         input_text=ingest_request.input_text,
         max_file_size=ingest_request.max_file_size,
         pattern_type=ingest_request.pattern_type,
         pattern=ingest_request.pattern,
         token=ingest_request.token,
     )
+    # limit URL to 255 characters
+    ingest_counter.labels(status=response.status_code, url=ingest_request.input_text[:255]).inc()
+    return response
 
 
 @router.get("/api/{user}/{repository}", responses=COMMON_INGEST_RESPONSES)
@@ -72,13 +78,16 @@ async def api_ingest_get(
     **Returns**
     - **JSONResponse**: Success response with ingestion results or error response with appropriate HTTP status code
     """
-    return await _perform_ingestion(
+    response = await _perform_ingestion(
         input_text=f"{user}/{repository}",
         max_file_size=max_file_size,
         pattern_type=pattern_type,
         pattern=pattern,
         token=token or None,
     )
+    # limit URL to 255 characters
+    ingest_counter.labels(status=response.status_code, url=f"{user}/{repository}"[:255]).inc()
+    return response
 
 
 @router.get("/api/download/file/{ingest_id}", response_class=FileResponse)

src/server/routers/metrics.py

@@ -0,0 +1,51 @@
+"""Prometheus metrics endpoint for the API."""
+
+import os
+from fastapi import APIRouter, HTTPException, Request, status
+from fastapi.responses import HTMLResponse
+from prometheus_client import REGISTRY, generate_latest
+
+router = APIRouter()
+
+@router.get("/metrics")
+async def metrics(request: Request) -> HTMLResponse:
+    """Serve Prometheus metrics with token authentication.
+    
+    This endpoint requires the GITINGEST_PROMETHEUS_TOKEN to be provided
+    in the Authorization header as 'Bearer <token>'.
+    
+    Parameters
+    ----------
+    request : Request
+        The incoming HTTP request containing headers
+        
+    Returns
+    -------
+    HTMLResponse
+        Prometheus metrics in text format
+        
+    Raises
+    ------
+    HTTPException
+        401 if no token provided or invalid token
+    """
+    # Get the expected token from environment
+    expected_token = os.getenv("GITINGEST_PROMETHEUS_TOKEN")
+    if not expected_token:
+        raise HTTPException(
+            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+        )
+    # Check Authorization header
+    auth_header = request.headers.get("Authorization")
+    if not auth_header:
+        raise HTTPException(
+            status_code=status.HTTP_401_UNAUTHORIZED,
+        )
+
+    # Extract and verify token
+    provided_token = auth_header[7:]  # Remove "Bearer " prefix
+    if provided_token != expected_token:
+        raise HTTPException(
+            status_code=status.HTTP_401_UNAUTHORIZED,
+        )
+    return HTMLResponse(content=generate_latest(REGISTRY), status_code=200, media_type="text/plain")