#!/usr/bin/env python3
"""Purge GOLD backups: keep max 5 per basename (doctrine 59 no-delete = archive tar)"""
import os, re, glob, tarfile, datetime
from collections import defaultdict
from pathlib import Path

VAULT = "/opt/wevads/vault"
ARCHIVE_DIR = f"{VAULT}/archived_gold_purge"
KEEP = 5
LOG = "/var/log/weval/gold-purge.log"
Path(LOG).parent.mkdir(parents=True, exist_ok=True)
Path(ARCHIVE_DIR).mkdir(parents=True, exist_ok=True)

def log(m):
    with open(LOG, "a") as f: f.write(f"[{datetime.datetime.now().isoformat()}] {m}\n")
    print(m)

# Group by basename stripping date suffix
groups = defaultdict(list)
for f in glob.glob(f"{VAULT}/*"):
    if os.path.isdir(f): continue
    name = os.path.basename(f)
    if name.startswith("archived_") or name.startswith("auto-"): continue
    # Extract basename by stripping date patterns
    base = re.sub(r"[._-]?(gold|GOLD|20260\d{3}[-_][\d]+|pre[-_].*)[-_.].*$", "", name, count=1)
    base = re.sub(r"[._-]?\d{14,}$", "", base)  # trailing timestamp
    base = re.sub(r"\.gold-.*$", "", base, flags=re.IGNORECASE)
    base = re.sub(r"\.GOLD-.*$", "", base)
    groups[base].append((f, os.path.getmtime(f)))

total_archived = 0
saved_bytes = 0
for base, files in groups.items():
    if len(files) <= KEEP: continue
    # Sort by mtime desc (newest first)
    files.sort(key=lambda x: x[1], reverse=True)
    to_archive = files[KEEP:]
    total_archived += len(to_archive)
    
    # Tar them together per basename 
    if to_archive:
        tar_name = f"{ARCHIVE_DIR}/{base.replace('/','_')}-olderthan-top{KEEP}-{datetime.date.today().isoformat()}.tar.gz"
        try:
            with tarfile.open(tar_name, "w:gz") as tar:
                for f, _ in to_archive:
                    try:
                        saved_bytes += os.path.getsize(f)
                        tar.add(f, arcname=os.path.basename(f))
                    except: pass
            # Now remove originals
            for f, _ in to_archive:
                try: os.remove(f)
                except: pass
        except Exception as e:
            log(f"ARCH_ERR {base}: {e}")

log(f"TOTAL_ARCHIVED {total_archived} files, saved_bytes={saved_bytes/1024/1024:.1f}MB")
print(f"Archived {total_archived} old GOLDs, freed {saved_bytes/1024/1024:.1f}MB")