reporter

indexer/config.py

@@ -32,3 +32,10 @@ BACKUP_PASSWORD = os.getenv("BACKUP_PASSWORD")
 
 DRY_RUN = os.getenv("DRY_RUN", "true").lower() in ("1", "true", "yes")
 BATCH_SIZE = int(os.getenv("BATCH_SIZE", 1000))
+
+# =========================
+# Logging
+# =========================
+
+LOG_LEVEL = os.getenv("LOG_LEVEL", "INFO")
+LOG_DIR = os.getenv("LOG_DIR", r"C:\Reporting\DropboxBackup\logs")

indexer/db.py

@@ -1,8 +1,13 @@
+import unicodedata
 import pymysql
 from datetime import datetime
 from indexer.config import DB_CONFIG, BATCH_SIZE
 
 
+def _nfc(s: str) -> str:
+    return unicodedata.normalize("NFC", s) if s else s
+
+
 def get_connection():
     return pymysql.connect(**DB_CONFIG)
 
@@ -50,7 +55,7 @@ def load_all_files(cur) -> dict:
     result = {}
     for row in cur.fetchall():
         file_id, rel_path, size, mtime, content_hash = row
-        result[rel_path] = {
+        result[_nfc(rel_path)] = {
             "id": file_id,
             "size": size,
             "mtime": mtime,
@@ -70,34 +75,24 @@ def batch_insert_files(cur, files_list: list, run_id: int) -> dict:
     Returns: {relative_path: file_id}
     """
     path_to_id = {}
-    for i in range(0, len(files_list), BATCH_SIZE):
-        chunk = files_list[i:i + BATCH_SIZE]
-        for f in chunk:
-            cur.execute(
-                """INSERT INTO files
-                   (relative_path, file_name, directory, file_size, mtime,
-                    content_hash, first_seen_run, last_seen_run, exists_now)
-                   VALUES (%s, %s, %s, %s, %s, %s, %s, %s, 1)
-                   ON DUPLICATE KEY UPDATE
-                    file_name = VALUES(file_name),
-                    directory = VALUES(directory),
-                    file_size = VALUES(file_size),
-                    mtime = VALUES(mtime),
-                    content_hash = VALUES(content_hash),
-                    last_seen_run = VALUES(last_seen_run),
-                    exists_now = 1""",
-                (f["relative_path"], f["file_name"], f["directory"],
-                 f["size"], f["mtime"], f["content_hash"], run_id, run_id)
-            )
-        # Fetch real IDs
-        paths = [f["relative_path"] for f in chunk]
-        placeholders = ",".join(["%s"] * len(paths))
+    for f in files_list:
         cur.execute(
-            f"SELECT id, relative_path FROM files WHERE relative_path IN ({placeholders})",
-            paths,
+            """INSERT INTO files
+               (relative_path, file_name, directory, file_size, mtime,
+                content_hash, first_seen_run, last_seen_run, exists_now)
+               VALUES (%s, %s, %s, %s, %s, %s, %s, %s, 1)
+               ON DUPLICATE KEY UPDATE
+                file_name = VALUES(file_name),
+                directory = VALUES(directory),
+                file_size = VALUES(file_size),
+                mtime = VALUES(mtime),
+                content_hash = VALUES(content_hash),
+                last_seen_run = VALUES(last_seen_run),
+                exists_now = 1""",
+            (f["relative_path"], f["file_name"], f["directory"],
+             f["size"], f["mtime"], f["content_hash"], run_id, run_id)
         )
-        for row in cur.fetchall():
-            path_to_id[row[1]] = row[0]
+        path_to_id[f["relative_path"]] = cur.lastrowid
     return path_to_id
 
 

indexer/hasher.py

@@ -1,4 +1,5 @@
 import ctypes
+import time
 
 from blake3 import blake3
 
@@ -17,6 +18,25 @@ def is_cloud_placeholder(path: str) -> bool:
     return bool(attrs & _CLOUD_MASK)
 
 
+def hydrate_file(path: str, timeout: int = 120, poll: int = 3) -> bool:
+    """
+    Přinutí Dropbox stáhnout cloud placeholder otevřením souboru.
+    Čeká max timeout sekund. Vrátí True pokud se soubor stáhl.
+    """
+    try:
+        with open(path, "rb") as f:
+            f.read(1)
+    except OSError:
+        pass
+
+    deadline = time.time() + timeout
+    while time.time() < deadline:
+        if not is_cloud_placeholder(path):
+            return True
+        time.sleep(poll)
+    return False
+
+
 def blake3_file(path, chunk_size=1024 * 1024):
     """Spočítá BLAKE3 hash souboru po blocích (bez načtení do paměti)."""
     h = blake3()

indexer/logger.py

@@ -0,0 +1,34 @@
+import logging
+import os
+import sys
+from logging.handlers import TimedRotatingFileHandler
+
+from indexer.config import LOG_LEVEL, LOG_DIR
+
+
+def setup_logging() -> logging.Logger:
+    os.makedirs(LOG_DIR, exist_ok=True)
+
+    level = getattr(logging, LOG_LEVEL.upper(), logging.INFO)
+    fmt = logging.Formatter(
+        "%(asctime)s [%(levelname)-8s] %(message)s",
+        datefmt="%Y-%m-%d %H:%M:%S",
+    )
+
+    file_handler = TimedRotatingFileHandler(
+        os.path.join(LOG_DIR, "backup.log"),
+        when="midnight",
+        backupCount=90,
+        encoding="utf-8",
+    )
+    file_handler.setFormatter(fmt)
+
+    console_handler = logging.StreamHandler(sys.stdout)
+    console_handler.setFormatter(fmt)
+
+    logging.root.setLevel(level)
+    logging.root.handlers.clear()
+    logging.root.addHandler(file_handler)
+    logging.root.addHandler(console_handler)
+
+    return logging.getLogger("backup")

indexer/scanner.py

@@ -1,4 +1,5 @@
 import os
+import unicodedata
 from datetime import datetime
 
 
@@ -18,8 +19,8 @@ def scan_files(root_path: str) -> dict:
                 stat = os.stat(full_path)
             except (FileNotFoundError, PermissionError):
                 continue
-            rel_path = os.path.relpath(full_path, root_path).replace("\\", "/")
-            rel_dir = os.path.relpath(root, root_path).replace("\\", "/")
+            rel_path = unicodedata.normalize("NFC", os.path.relpath(full_path, root_path).replace("\\", "/"))
+            rel_dir = unicodedata.normalize("NFC", os.path.relpath(root, root_path).replace("\\", "/"))
             # Truncate microseconds — MySQL DATETIME rounds to whole seconds,
             # which causes false "modified" detections on every run.
             mtime = datetime.fromtimestamp(stat.st_mtime).replace(microsecond=0)