hecvec 6.1.2__tar.gz → 6.2.0__tar.gz

{hecvec-6.1.2 → hecvec-6.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: hecvec
-Version: 6.1.2
+Version: 6.2.0
 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
@@ -107,7 +107,7 @@ import hecvec
 
 # Default: token chunking, Chroma at localhost:8000
 result = hecvec.Slicer.slice(path="/path/to/folder_or_file")
-# → {"files": N, "chunks": M, "collection": "folder_or_file_name_token"}
+# → {"files": N, "chunks": M, "collection": "folder_or_file_name_token_cs200_ov0_enccl100k_base"}
 
 # Custom host/port and semantic chunking
 result = hecvec.Slicer.slice(
@@ -151,7 +151,7 @@ All of these can be passed to `Slicer(...)` or to `Slicer.slice(..., key=value)`
 |-----------|---------|-------------|
 | `path` | *(required)* | File or directory to process (`.txt`/`.md` only). |
 | `root` | `path.parent` (file) or `path` (dir) | Safe root for resolving paths (used when listing under a directory). |
-| `collection_name` | `"hecvec"` | Base name for the Chroma collection. If `"hecvec"`, it is replaced by the file stem or directory name; the final name is always `{collection_name}_{chunking_method}` (e.g. `mydoc_token`). |
+| `collection_name` | `"hecvec"` | Base name for the Chroma collection. If `"hecvec"`, it is replaced by the file stem or directory name; the final name includes method + chunk config (so different `chunk_size`/`chunk_overlap` don’t collide). |
 | `db` | `"chroma"` | Database to use. Only `"chroma"` is supported. When `db="chroma"`, connection uses `host` and `port`. |
 | `host` | `"localhost"` | Server host (used when `db="chroma"`). Server must be listening. |
 | `port` | `8000` | Server port (used when `db="chroma"`). |
@@ -399,9 +399,12 @@ docker run -p 8000:8000 chromadb/chroma
   - **Directory:** `path.name` (e.g. `docs`)
 - The **final** collection name is always:
 
-  **`{base_name}_{chunking_method}`**
+  **`{base_name}_{chunking_method}_{chunk_config}`**
 
-  Examples: `mydoc_token`, `docs_semantic`, `CNSF-S0043-0032-2025_CONDUSEF-005190-08_token`.
+  Examples:
+  - token: `mydoc_token_cs200_ov0_enccl100k_base`
+  - text: `mydoc_text_cs400_ov0`
+  - llm/semantic: `mydoc_llm_cs200`
 
 - If a collection with that name **already exists**, the pipeline does **not** add documents again. It logs that the collection already exists and returns something like:
 

{hecvec-6.1.2 → hecvec-6.2.0}/README.md

@@ -70,7 +70,7 @@ import hecvec
 
 # Default: token chunking, Chroma at localhost:8000
 result = hecvec.Slicer.slice(path="/path/to/folder_or_file")
-# → {"files": N, "chunks": M, "collection": "folder_or_file_name_token"}
+# → {"files": N, "chunks": M, "collection": "folder_or_file_name_token_cs200_ov0_enccl100k_base"}
 
 # Custom host/port and semantic chunking
 result = hecvec.Slicer.slice(
@@ -114,7 +114,7 @@ All of these can be passed to `Slicer(...)` or to `Slicer.slice(..., key=value)`
 |-----------|---------|-------------|
 | `path` | *(required)* | File or directory to process (`.txt`/`.md` only). |
 | `root` | `path.parent` (file) or `path` (dir) | Safe root for resolving paths (used when listing under a directory). |
-| `collection_name` | `"hecvec"` | Base name for the Chroma collection. If `"hecvec"`, it is replaced by the file stem or directory name; the final name is always `{collection_name}_{chunking_method}` (e.g. `mydoc_token`). |
+| `collection_name` | `"hecvec"` | Base name for the Chroma collection. If `"hecvec"`, it is replaced by the file stem or directory name; the final name includes method + chunk config (so different `chunk_size`/`chunk_overlap` don’t collide). |
 | `db` | `"chroma"` | Database to use. Only `"chroma"` is supported. When `db="chroma"`, connection uses `host` and `port`. |
 | `host` | `"localhost"` | Server host (used when `db="chroma"`). Server must be listening. |
 | `port` | `8000` | Server port (used when `db="chroma"`). |
@@ -362,9 +362,12 @@ docker run -p 8000:8000 chromadb/chroma
   - **Directory:** `path.name` (e.g. `docs`)
 - The **final** collection name is always:
 
-  **`{base_name}_{chunking_method}`**
+  **`{base_name}_{chunking_method}_{chunk_config}`**
 
-  Examples: `mydoc_token`, `docs_semantic`, `CNSF-S0043-0032-2025_CONDUSEF-005190-08_token`.
+  Examples:
+  - token: `mydoc_token_cs200_ov0_enccl100k_base`
+  - text: `mydoc_text_cs400_ov0`
+  - llm/semantic: `mydoc_llm_cs200`
 
 - If a collection with that name **already exists**, the pipeline does **not** add documents again. It logs that the collection already exists and returns something like:
 

{hecvec-6.1.2 → hecvec-6.2.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "hecvec"
-version = "6.1.2"
+version = "6.2.0"
 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-6.1.2 → hecvec-6.2.0}/scripts/test_slice.py

@@ -11,9 +11,9 @@ def main():
 
     from hecvec import Slicer
     path = Path(sys.argv[1]).expanduser().resolve() if len(sys.argv) >= 2 else (ROOT / "tests/CNSF-S0043-0032-2025_CONDUSEF-005190-08.txt")
-    slicer = Slicer(db="chroma", host="localhosttt", port=8000)
+    slicer = Slicer(db="chroma", host="localhost", port=8000)
     print("hecvec slicer config:", {"db": slicer.db, "host": slicer.host, "port": slicer.port, "auth": slicer.auth})
-    test = slicer.slice(path=path, collection_name="test", chunking_method="text")
+    test = slicer.slice(path=path, collection_name="test", chunking_method="token",chunk_size=400)
     print(test)
 
     collections = slicer.collections()

{hecvec-6.1.2 → hecvec-6.2.0}/src/hecvec/__init__.py

@@ -35,4 +35,4 @@ __all__ = [
     "__version__",
 ]
 
-__version__ = "6.1.2"
+__version__ = "6.2.0"

{hecvec-6.1.2 → hecvec-6.2.0}/src/hecvec/pipeline.py

@@ -221,7 +221,16 @@ class Slicer:
 
         if collection_name == "hecvec":
             collection_name = path.stem if path.is_file() else path.name
-        collection_name = f"{collection_name}_{chunking_method}"
+        # Include chunking configuration in the collection name so reruns with a different
+        # chunk_size/chunk_overlap do not collide with an existing collection.
+        if chunking_method == "token":
+            cfg_suffix = f"cs{chunk_size}_ov{chunk_overlap}_enc{encoding_name}"
+        elif chunking_method == "text":
+            cfg_suffix = f"cs{chunk_size}_ov{chunk_overlap}"
+        else:
+            # For semantic/llm we keep the suffix minimal since advanced knobs aren't exposed on Slicer.
+            cfg_suffix = f"cs{chunk_size}"
+        collection_name = f"{collection_name}_{chunking_method}_{cfg_suffix}"
 
         logger.info("Resolved collection name: %s", collection_name)
 

{hecvec-6.1.2 → hecvec-6.2.0}/uv.lock

@@ -599,7 +599,7 @@ wheels = [
 
 [[package]]
 name = "hecvec"
-version = "6.1.1"
+version = "6.1.3"
 source = { editable = "." }
 dependencies = [
     { name = "chromadb" },