140 lines
3.0 KiB
Python
140 lines
3.0 KiB
Python
#!/usr/bin/env python3
|
|
# -*- coding: utf-8 -*-
|
|
|
|
"""
|
|
TOWER PATH NORMALIZATION MIGRATION
|
|
----------------------------------
|
|
✔ Normalizes full_path → NFC canonical
|
|
✔ Recalculates path_hash
|
|
✔ Uses two DB connections (streaming safe)
|
|
✔ Idempotent (safe to rerun)
|
|
✔ Production safe
|
|
"""
|
|
|
|
import pymysql
|
|
import hashlib
|
|
import posixpath
|
|
import unicodedata
|
|
import time
|
|
|
|
# =========================
|
|
# CONFIG
|
|
# =========================
|
|
|
|
HOST_TO_FIX = "tower"
|
|
BATCH_FETCH = 5000
|
|
COMMIT_EVERY = 2000
|
|
|
|
DB_CONFIG = {
|
|
"host": "192.168.1.50",
|
|
"port": 3306,
|
|
"user": "root",
|
|
"password": "Vlado9674+",
|
|
"database": "torrents",
|
|
"charset": "utf8mb4",
|
|
"autocommit": False,
|
|
}
|
|
|
|
# =========================
|
|
# CANONICALIZATION
|
|
# =========================
|
|
|
|
def canonical_path(path_str: str) -> str:
|
|
path_str = path_str.replace("\\", "/")
|
|
path_str = posixpath.normpath(path_str)
|
|
path_str = unicodedata.normalize("NFC", path_str)
|
|
return path_str
|
|
|
|
def md5_bytes(path_str: str) -> bytes:
|
|
return hashlib.md5(path_str.encode("utf-8")).digest()
|
|
|
|
# =========================
|
|
# MAIN
|
|
# =========================
|
|
|
|
def main():
|
|
|
|
print("=" * 70)
|
|
print("🚀 TOWER PATH NORMALIZATION MIGRATION")
|
|
print(f"Host: {HOST_TO_FIX}")
|
|
print("=" * 70)
|
|
|
|
start = time.time()
|
|
|
|
# --- TWO CONNECTIONS ---
|
|
db_read = pymysql.connect(**DB_CONFIG)
|
|
db_write = pymysql.connect(**DB_CONFIG)
|
|
|
|
read_cur = db_read.cursor(pymysql.cursors.SSCursor)
|
|
write_cur = db_write.cursor()
|
|
|
|
read_cur.execute("""
|
|
SELECT id, full_path, path_hash
|
|
FROM file_md5_index
|
|
WHERE host_name = %s
|
|
""", (HOST_TO_FIX,))
|
|
|
|
total = 0
|
|
updated = 0
|
|
skipped = 0
|
|
pending_commit = 0
|
|
|
|
while True:
|
|
|
|
rows = read_cur.fetchmany(BATCH_FETCH)
|
|
|
|
if not rows:
|
|
break
|
|
|
|
for rec_id, full_path, stored_hash in rows:
|
|
|
|
total += 1
|
|
|
|
new_path = canonical_path(full_path)
|
|
new_hash = md5_bytes(new_path)
|
|
|
|
if new_path == full_path and new_hash == stored_hash:
|
|
skipped += 1
|
|
continue
|
|
|
|
write_cur.execute("""
|
|
UPDATE file_md5_index
|
|
SET full_path = %s,
|
|
path_hash = %s
|
|
WHERE id = %s
|
|
""", (new_path, new_hash, rec_id))
|
|
|
|
updated += 1
|
|
pending_commit += 1
|
|
|
|
if pending_commit >= COMMIT_EVERY:
|
|
db_write.commit()
|
|
pending_commit = 0
|
|
|
|
print(
|
|
f"Checked {total:,} | Updated {updated:,} | Skipped {skipped:,}"
|
|
)
|
|
|
|
if pending_commit:
|
|
db_write.commit()
|
|
|
|
elapsed = time.time() - start
|
|
|
|
print("\n" + "=" * 70)
|
|
print("✅ MIGRATION FINISHED")
|
|
print("=" * 70)
|
|
print(f"Total checked : {total:,}")
|
|
print(f"Rows updated : {updated:,}")
|
|
print(f"Rows skipped : {skipped:,}")
|
|
print(f"Runtime : {elapsed:.1f}s")
|
|
print("=" * 70)
|
|
|
|
read_cur.close()
|
|
write_cur.close()
|
|
db_read.close()
|
|
db_write.close()
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|