contextinator 1.2.4__tar.gz → 1.2.5__tar.gz

{contextinator-1.2.4/src/contextinator.egg-info → contextinator-1.2.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: contextinator
-Version: 1.2.4
+Version: 1.2.5
 Summary: Intelligent Codebase Understanding for AI Agents - Transform any codebase into semantically-aware, searchable knowledge
 Author-email: STARTHACK Team <founders@starthack.io>
 License: Apache-2.0
@@ -85,6 +85,7 @@ Dynamic: license-file
 Turn any codebase into semantically-aware, searchable knowledge for AI-powered workflows.
 </p>
 
+
 ### Key Features
 
 - **AST-Powered Chunking** - Extract functions, classes, and methods from 23+ programming languages

{contextinator-1.2.4 → contextinator-1.2.5}/README.md

@@ -4,6 +4,7 @@
 Turn any codebase into semantically-aware, searchable knowledge for AI-powered workflows.
 </p>
 
+
 ### Key Features
 
 - **AST-Powered Chunking** - Extract functions, classes, and methods from 23+ programming languages

{contextinator-1.2.4 → contextinator-1.2.5}/src/contextinator/_version.py

@@ -28,7 +28,7 @@ version_tuple: VERSION_TUPLE
 commit_id: COMMIT_ID
 __commit_id__: COMMIT_ID
 
-__version__ = version = '1.2.4'
-__version_tuple__ = version_tuple = (1, 2, 4)
+__version__ = version = '1.2.5'
+__version_tuple__ = version_tuple = (1, 2, 5)
 
-__commit_id__ = commit_id = 'g784ddc871'
+__commit_id__ = commit_id = 'g8ab564fe4'

contextinator-1.2.5/src/contextinator/tools/cat_file.py

@@ -0,0 +1,110 @@
+"""Cat file with TRUE async."""
+
+import asyncio
+import hashlib
+from typing import Dict, List, Optional
+from ..utils.logger import logger
+from ..config import USE_CHROMA_SERVER
+
+_async_chroma_client = None
+
+
+async def _get_async_chroma():
+    global _async_chroma_client
+    if _async_chroma_client is None:
+        if USE_CHROMA_SERVER:
+            from ..vectorstore.async_chroma import get_async_client
+            from ..config import CHROMA_SERVER_URL
+            from urllib.parse import urlparse
+
+            parsed = urlparse(CHROMA_SERVER_URL)
+            _async_chroma_client = await get_async_client(
+                host=parsed.hostname or "localhost", port=parsed.port or 8000
+            )
+        else:
+            return None
+    return _async_chroma_client
+
+
+async def cat_file(
+    collection_name: str, file_path: str, chromadb_dir: Optional[str] = None
+) -> str:
+    """TRUE async cat file."""
+    if not collection_name or not file_path:
+        raise ValueError("Collection name and file path required")
+
+    if file_path.startswith("/"):
+        file_path = file_path[1:]
+
+    client = await _get_async_chroma()
+    from ..config import sanitize_collection_name
+
+    collection = await client.get_collection(sanitize_collection_name(collection_name))
+
+    results = await collection.get(
+        where={"file_path": file_path}, include=["documents", "metadatas"]
+    )
+
+    if not results["ids"]:
+        raise ValueError(f"File not found: {file_path}")
+
+    logger.debug(f"Found {len(results['ids'])} chunks for file: {file_path}")
+
+    # Parse and enrich chunk metadata
+    chunks = []
+    for i, (doc, meta) in enumerate(zip(results["documents"], results["metadatas"])):
+        chunk = {
+            "content": doc,
+            "start_line": meta.get("start_line", 0),
+            "end_line": meta.get("end_line", 0), 
+            "split_index": meta.get("split_index", 0),
+            "parent_id": meta.get("parent_id"),
+            "is_split": meta.get("is_split", False),
+            "original_id": meta.get("original_id"),
+            "node_type": meta.get("node_type", "unknown"),
+            "chunk_id": results["ids"][i],
+        }
+        chunks.append(chunk)
+        
+    # Sort chunks by start_line first, then by split_index, then by end_line
+    chunks.sort(key=lambda x: (x["start_line"], x["split_index"], x["end_line"]))
+    
+    logger.debug(f"Chunk details: {[(c['start_line'], c['end_line'], c['is_split'], c['split_index'], len(c['content'])) for c in chunks]}")
+
+    # Reconstruct file
+    return _reconstruct_file(chunks)
+
+
+def _reconstruct_file(chunks: List[Dict]) -> str:
+    """Deduplicate and concatenate chunks, removing nested duplicates."""
+    if not chunks:
+        return ""
+    
+    # Sort by start_line, then by length (longer first)
+    chunks.sort(key=lambda x: (int(x.get('start_line', 0)), -int(x.get('end_line', 0))))
+    
+    # Remove chunks that are completely contained in other chunks
+    unique = []
+    for c in chunks:
+        c_start = int(c.get('start_line', 0))
+        c_end = int(c.get('end_line', 0))
+        
+        # Check if this chunk is contained in any already-kept chunk
+        is_contained = False
+        for kept in unique:
+            k_start = int(kept.get('start_line', 0))
+            k_end = int(kept.get('end_line', 0))
+            
+            # If c is completely inside kept, skip it
+            if k_start <= c_start and c_end <= k_end and (k_start != c_start or k_end != c_end):
+                is_contained = True
+                break
+        
+        if not is_contained:
+            unique.append(c)
+    
+    # Concatenate with double newline between chunks
+    return '\n\n'.join(c.get('content', '').strip() for c in unique if c.get('content', '').strip())
+
+
+__all__ = ["cat_file"]

{contextinator-1.2.4 → contextinator-1.2.5/src/contextinator.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: contextinator
-Version: 1.2.4
+Version: 1.2.5
 Summary: Intelligent Codebase Understanding for AI Agents - Transform any codebase into semantically-aware, searchable knowledge
 Author-email: STARTHACK Team <founders@starthack.io>
 License: Apache-2.0
@@ -85,6 +85,7 @@ Dynamic: license-file
 Turn any codebase into semantically-aware, searchable knowledge for AI-powered workflows.
 </p>
 
+
 ### Key Features
 
 - **AST-Powered Chunking** - Extract functions, classes, and methods from 23+ programming languages

contextinator-1.2.4/src/contextinator/tools/cat_file.py

@@ -1,85 +0,0 @@
-"""Cat file with TRUE async."""
-
-import asyncio
-from typing import Optional
-from ..utils.logger import logger
-from ..config import USE_CHROMA_SERVER
-
-_async_chroma_client = None
-
-
-async def _get_async_chroma():
-    global _async_chroma_client
-    if _async_chroma_client is None:
-        if USE_CHROMA_SERVER:
-            from ..vectorstore.async_chroma import get_async_client
-            from ..config import CHROMA_SERVER_URL
-            from urllib.parse import urlparse
-
-            parsed = urlparse(CHROMA_SERVER_URL)
-            _async_chroma_client = await get_async_client(
-                host=parsed.hostname or "localhost", port=parsed.port or 8000
-            )
-        else:
-            return None
-    return _async_chroma_client
-
-
-async def cat_file(
-    collection_name: str, file_path: str, chromadb_dir: Optional[str] = None
-) -> str:
-    """TRUE async cat file."""
-    if not collection_name or not file_path:
-        raise ValueError("Collection name and file path required")
-
-    if file_path.startswith("/"):
-        file_path = file_path[1:]
-
-    client = await _get_async_chroma()
-    from ..config import sanitize_collection_name
-
-    collection = await client.get_collection(sanitize_collection_name(collection_name))
-
-    results = await collection.get(
-        where={"file_path": file_path}, include=["documents", "metadatas"]
-    )
-
-    if not results["ids"]:
-        raise ValueError(f"File not found: {file_path}")
-
-    chunks = sorted(
-        [
-            {
-                "content": doc,
-                "start_line": meta.get("start_line", 0),
-                "end_line": meta.get("end_line", 0),
-                "split_index": meta.get("split_index", 0),
-                "parent_id": meta.get("parent_id"),
-            }
-            for doc, meta in zip(results["documents"], results["metadatas"])
-        ],
-        key=lambda x: (x["start_line"], x["split_index"]),
-    )
-
-    # Deduplicate overlapping chunks and filter out nested children
-    seen_ranges = set()
-    unique_chunks = []
-    for chunk in chunks:
-        range_key = (chunk["start_line"], chunk["end_line"])
-        # Skip if same range already seen OR if this chunk is nested inside another
-        if range_key in seen_ranges:
-            continue
-        # Check if this chunk is contained within any already added chunk
-        is_nested = any(
-            uc["start_line"] <= chunk["start_line"] and uc["end_line"] >= chunk["end_line"]
-            and (uc["start_line"], uc["end_line"]) != range_key
-            for uc in unique_chunks
-        )
-        if not is_nested:
-            seen_ranges.add(range_key)
-            unique_chunks.append(chunk)
-
-    return "\n".join(c["content"] for c in unique_chunks)
-
-
-__all__ = ["cat_file"]