feat(cli): support excluding files in queries.

Author: Zhe Yu

src/vectorcode/database/chroma0.py

@@ -40,7 +40,7 @@
 
 
 def __convert_chroma_query_results(
-    chroma_result: QueryResult, queries: list[str]
+    chroma_result: QueryResult, queries: Sequence[str]
 ) -> list[types.QueryResult]:
     """Convert chromadb query result to in-house query results"""
     assert chroma_result["documents"] is not None
@@ -272,6 +272,11 @@ async def query(self, collection_path, keywords_embeddings, opts):
         query_count = opts.count or (
             await self.count(collection_path, ResultType.chunk)
         )
+        query_filter = None
+        if len(opts.excluded_files):
+            query_filter = cast(
+                chromadb.Where, {"path": {"$nin": list(opts.excluded_files)}}
+            )
         query_result = await collection.query(
             query_embeddings=keywords_embeddings,
             include=[
@@ -280,6 +285,7 @@ async def query(self, collection_path, keywords_embeddings, opts):
                 IncludeEnum.distances,
             ],
             n_results=query_count,
+            where=query_filter,
         )
         return __convert_chroma_query_results(query_result, opts.keywords)
 

src/vectorcode/database/types.py

@@ -1,7 +1,7 @@
 import json
 from dataclasses import dataclass, field, fields
 from enum import StrEnum
-from typing import Any
+from typing import Any, Sequence
 
 import tabulate
 
@@ -17,9 +17,10 @@ class ResultType(StrEnum):
 
 @dataclass
 class QueryOpts:
-    keywords: list[str]
+    keywords: Sequence[str]
     count: int | None = None
     return_type: ResultType = ResultType.chunk
+    excluded_files: Sequence[str] = []
 
 
 @dataclass