Merge pull request 'fix(indexer): restore module-level embedding/batch constants (cronjobs cassés)' (#59) from fix/indexer-missing-constants into main

Reviewed-on: #59

.gitea/workflows/main.yaml

@@ -22,7 +22,9 @@ jobs:
         with:
           python-version: "3.14"
       - name: Install dependencies
-        run: pip install -r requirements.txt pytest
+        run: pip install -r requirements.txt pytest ruff==0.17.1
+      - name: Lint (undefined names)
+        run: ruff check src/ tests/ --select F821
       - name: Run unit tests
         run: pytest tests/ -v
         env:

src/indexer.py

@@ -29,6 +29,13 @@ MAILDIR_FOLDERS = os.environ.get("MAILDIR_FOLDERS", "")
 QDRANT_URL = os.environ.get("QDRANT_URL", "")
 COLLECTION_NAME = os.environ.get("COLLECTION_NAME", "")
 
+# Embedding model and batching. EMBEDDING_MODEL_NAME MUST stay in sync with the
+# one used by server.py: indexer and server have to embed with the same model,
+# otherwise the vectors stored in Qdrant are not comparable with query vectors.
+EMBEDDING_MODEL_NAME = os.environ.get("EMBEDDING_MODEL_NAME", "BAAI/bge-small-en-v1.5")
+BATCH_SIZE = int(os.environ.get("BATCH_SIZE", "100"))
+EMBEDDING_BATCH_SIZE = int(os.environ.get("EMBEDDING_BATCH_SIZE", "64"))
+
 METADATA_COLLECTION = "mcp_indexer_metadata"
 INCREMENTAL_DAYS = int(os.environ.get("INCREMENTAL_DAYS", "7"))
 FORCE_REINDEX = os.environ.get("FORCE_REINDEX", "").lower() in ("1", "true", "yes")

tests/test_indexer.py

@@ -21,6 +21,9 @@ from indexer import (
     parse_email_message,
     get_recent_keys,
     METADATA_COLLECTION,
+    EMBEDDING_MODEL_NAME,
+    BATCH_SIZE,
+    EMBEDDING_BATCH_SIZE,
 )
 
 
@@ -227,3 +230,25 @@ class TestGetRecentKeys:
         assert key.split(":")[0] in keys
         for k in keys:
             assert ":" not in k
+
+
+# ---------------------------------------------------------------------------
+# Regression: module-level constants consumed by main()
+# ---------------------------------------------------------------------------
+
+
+def test_indexer_module_constants_are_defined():
+    """Regression (e32806a): the "make indexer importable without env" refactor
+    dropped these three module-level definitions while main() kept using them,
+    so every indexer run died with NameError before touching Qdrant."""
+    assert isinstance(EMBEDDING_MODEL_NAME, str) and EMBEDDING_MODEL_NAME
+    assert isinstance(BATCH_SIZE, int) and BATCH_SIZE > 0
+    assert isinstance(EMBEDDING_BATCH_SIZE, int) and EMBEDDING_BATCH_SIZE > 0
+
+
+def test_embedding_model_is_the_same_as_the_server_one():
+    """Indexer and server must embed with the same model, otherwise the vectors
+    stored by the indexer are not comparable with the query vectors."""
+    from server import EMBEDDING_MODEL_NAME as server_embedding_model
+
+    assert EMBEDDING_MODEL_NAME == server_embedding_model