hecvec 0.4.2__tar.gz → 0.4.3__tar.gz

{hecvec-0.4.2 → hecvec-0.4.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: hecvec
-Version: 0.4.2
+Version: 0.4.3
 Summary: List directories (safe root), filter .txt/.md files, read as text, chunk, embed, and push to Chroma.
 License-Expression: MIT
 Keywords: chunking,document-pipeline,listdir,text-files

{hecvec-0.4.2 → hecvec-0.4.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "hecvec"
-version = "0.4.2"
+version = "0.4.3"
 description = "List directories (safe root), filter .txt/.md files, read as text, chunk, embed, and push to Chroma."
 readme = "README.md"
 requires-python = ">=3.9,<3.14"

{hecvec-0.4.2 → hecvec-0.4.3}/src/hecvec/__init__.py

@@ -34,4 +34,4 @@ __all__ = [
     "__version__",
 ]
 
-__version__ = "0.4.2"
+__version__ = "0.4.3"

{hecvec-0.4.2 → hecvec-0.4.3}/src/hecvec/chroma_client.py

@@ -13,13 +13,28 @@ DEFAULT_HOST = "localhost"
 DEFAULT_PORT = 8000
 
 
-def get_client(host: str = DEFAULT_HOST, port: int = DEFAULT_PORT):
-    """Return a Chroma HTTP client, or an in-memory client if the server is not reachable."""
+def get_client(
+    host: str = DEFAULT_HOST,
+    port: int = DEFAULT_PORT,
+    persist_path: str | None = None,
+):
+    """
+    Return a Chroma client and mode.
+
+    - If persist_path is set: use PersistentClient(path). Data is stored on disk.
+      Returns (client, "persistent"). host/port are ignored.
+    - Else: try HttpClient(host, port). If that fails, use EphemeralClient().
+      Returns (client, "server") or (client, "ephemeral").
+    """
     import chromadb
+    if persist_path is not None:
+        client = chromadb.PersistentClient(path=persist_path)
+        return client, "persistent"
     try:
-        return chromadb.HttpClient(host=host, port=port)
+        client = chromadb.HttpClient(host=host, port=port)
+        return client, "server"
     except Exception:
-        return chromadb.EphemeralClient()
+        return chromadb.EphemeralClient(), "ephemeral"
 
 
 def get_or_create_collection(
@@ -39,11 +54,14 @@ def add_documents(
     ids: list[str],
     embeddings: list[list[float]],
     documents: list[str],
-) -> None:
+) -> dict:
     """
     Add documents to a collection. If dimension mismatch, deletes and recreates the collection.
+    Returns {"collection_existed": bool} so the caller can log whether the collection was pre-existing.
     """
     import chromadb
+    existing_names = [c.name for c in client.list_collections()]
+    collection_existed = collection_name in existing_names
     coll = get_or_create_collection(client, collection_name)
     try:
         coll.add(ids=ids, embeddings=embeddings, documents=documents)
@@ -56,3 +74,5 @@ def add_documents(
             metadata={"hnsw:space": "cosine"},
         )
         coll.add(ids=ids, embeddings=embeddings, documents=documents)
+        collection_existed = False  # we recreated it
+    return {"collection_existed": collection_existed}

{hecvec-0.4.2 → hecvec-0.4.3}/src/hecvec/pipeline.py

@@ -209,9 +209,17 @@ class Slicer:
         # 5/5 Push to Chroma
         stage_start = perf_counter()
         logger.info("[5/5] Writing to Chroma | host=%s | port=%s | collection=%s", chroma_host, chroma_port, collection_name)
-        client = get_client(host=chroma_host, port=chroma_port)
+        client, chroma_mode = get_client(host=chroma_host, port=chroma_port)
+        if chroma_mode == "server":
+            logger.info("[5/5] Using Chroma server at %s:%s", chroma_host, chroma_port)
+        else:
+            logger.info("[5/5] Chroma server unreachable; using in-memory (ephemeral) client")
         logger.info("[5/5] Adding %d document chunk(s) to collection...", len(documents))
-        add_documents(client, collection_name, ids, embeddings, documents)
+        add_result = add_documents(client, collection_name, ids, embeddings, documents)
+        if add_result["collection_existed"]:
+            logger.info("[5/5] Collection %r already existed; documents appended", collection_name)
+        else:
+            logger.info("[5/5] Collection %r created (new)", collection_name)
         logger.info("[5/5] Chroma write completed in %.2fs", perf_counter() - stage_start)
 
         logger.info("Slice finished in %.2fs | files=%d | chunks=%d | collection=%s", perf_counter() - total_start, len(path_and_text), len(documents), collection_name)

{hecvec-0.4.2 → hecvec-0.4.3}/uv.lock

@@ -599,7 +599,7 @@ wheels = [
 
 [[package]]
 name = "hecvec"
-version = "0.4.1"
+version = "0.4.2"
 source = { editable = "." }
 dependencies = [
     { name = "chromadb" },