Create backend/codebase_index.py

backend/codebase_index.py

@@ -1,4 +1,4 @@
-import os, faiss, pickle
+import os, faiss
 from pathlib import Path
 from typing import List, Dict
 from sentence_transformers import SentenceTransformer
@@ -13,17 +13,14 @@ class CodebaseIndex:
         self.chunks = []
 
     def build_index(self):
-        # Walk all code files
         exts = {'.py', '.js', '.ts', '.tsx', '.jsx', '.go', '.rs', '.java', '.cpp', '.c', '.h'}
         for filepath in self.project_path.rglob('*'):
             if filepath.suffix in exts and filepath.is_file():
                 self.file_list.append(str(filepath.relative_to(self.project_path)))
                 content = filepath.read_text()
-                # Split into chunks (by functions/paragraphs)
                 chunks = self._split_into_chunks(content)
                 for chunk in chunks:
                     self.chunks.append((str(filepath.relative_to(self.project_path)), chunk))
-        # Encode and build FAISS
         texts = [chunk[1] for chunk in self.chunks]
         embeddings = self.model.encode(texts, show_progress_bar=False)
         dim = embeddings.shape[1]
@@ -43,7 +40,6 @@ class CodebaseIndex:
         return results
 
     def _split_into_chunks(self, text, max_chars=1000):
-        # Simple split by double newlines, with overlap
         paragraphs = text.split('\n\n')
         chunks = []
         current = ""