#!/usr/bin/env python3 import os import sys import pickle import configparser from concurrent.futures import ThreadPoolExecutor # ============================================================ # CONFIG config = configparser.ConfigParser() config.read(r"C:\ProgramData\Program-Search\config.ini") # PATHS SEARCH_ROOT = config["Paths"]["SEARCH_ROOT"] INDEX_FILE = config["Paths"]["INDEX_FILE"] # FEATURES WORKER_COUNT = config.getint("Features", "WORKER_COUNT", fallback=8) # ============================================================ def read_file(path, mtime, size): try: with open(path, "r", errors="ignore") as f: return ( path, { "filename": os.path.basename(path).lower(), "content": f.read().lower(), "mtime": mtime, "size": size, } ) except Exception: return None def scan_files(root): """Return {path: (mtime, size)} for every file under root.""" found = {} stack = [root] while stack: current = stack.pop() try: with os.scandir(current) as it: for entry in it: try: if entry.is_dir(follow_symlinks=False): stack.append(entry.path) elif entry.is_file(): st = entry.stat() found[entry.path] = (st.st_mtime, st.st_size) except OSError: pass except OSError: pass return found def load_index(): try: with open(INDEX_FILE, "rb") as f: return pickle.load(f) except Exception: return {} def save_index(index): os.makedirs(os.path.dirname(INDEX_FILE), exist_ok=True) # Write to a temp file first so a crash never leaves a corrupt index. tmp_file = INDEX_FILE + ".tmp" with open(tmp_file, "wb") as f: pickle.dump(index, f, protocol=pickle.HIGHEST_PROTOCOL) os.replace(tmp_file, INDEX_FILE) def update_index(index, verbose=False): """ Bring index up to date with SEARCH_ROOT in place. Only new or changed files are re-read; deleted files are removed. Returns True if anything changed. """ if verbose: print("Scanning files...") on_disk = scan_files(SEARCH_ROOT) removed = [p for p in index if p not in on_disk] for p in removed: del index[p] to_read = [] for path, (mtime, size) in on_disk.items(): data = index.get(path) if ( data is None or data.get("mtime") != mtime or data.get("size") != size ): to_read.append((path, mtime, size)) if verbose: print( f"Found {len(on_disk)} files: " f"{len(to_read)} new/changed, {len(removed)} removed" ) completed = 0 with ThreadPoolExecutor(max_workers=WORKER_COUNT) as executor: for result in executor.map(lambda a: read_file(*a), to_read): completed += 1 if verbose and completed % 100 == 0: print(f"{completed}/{len(to_read)}") if result: path, data = result index[path] = data return bool(removed or to_read) def build_index(full=False): index = {} if full else load_index() changed = update_index(index, verbose=True) if not changed and os.path.exists(INDEX_FILE): print("\nIndex already up to date.") return print("\nSaving index...") save_index(index) size_mb = os.path.getsize(INDEX_FILE) / 1024 / 1024 print( f"Index saved:\n" f"{INDEX_FILE}\n" f"{len(index)} files\n" f"{size_mb:.2f} MB" ) if __name__ == "__main__": # Run with --full to rebuild the index from scratch. build_index(full="--full" in sys.argv)