#!/usr/bin/env python3
"""Rebuild H5 metadata database - V2 (capture ALL patterns)."""
import os
import re
import sqlite3
from datetime import datetime

H5_ROOTS = ['/mnt/data_sdb1', '/mnt/kingston']
DB_PATH = '/home/floppyrj45/docker/seismic-nodes-viewer/h5_data.db'

# Pattern plus permissif - capture TOUS les b###
FILE_PATTERN = re.compile(r'b(\d+)')
CHANNEL_PATTERN = re.compile(r'ch(\d+)')

SCHEMA = [
    'CREATE TABLE IF NOT EXISTS positions (node_code INTEGER PRIMARY KEY, has_data BOOLEAN, has_aux BOOLEAN, sample_count INTEGER, last_seen TEXT)',
    'CREATE TABLE IF NOT EXISTS files (id INTEGER PRIMARY KEY AUTOINCREMENT, path TEXT, node_code INTEGER, channel INTEGER, dataset TEXT, size INTEGER, mtime INTEGER, FOREIGN KEY(node_code) REFERENCES positions(node_code))',
    'CREATE INDEX IF NOT EXISTS idx_files_node ON files(node_code)'
]

def rebuild_db():
    conn = sqlite3.connect(DB_PATH)
    cur = conn.cursor()

    for stmt in SCHEMA:
        cur.execute(stmt)

    cur.execute('DELETE FROM files')
    cur.execute('DELETE FROM positions')

    files_counter = 0
    summary = {}

    for root in H5_ROOTS:
        for dirpath, _, filenames in os.walk(root):
            for filename in filenames:
                if not filename.endswith('.h5'):
                    continue
                
                filepath = os.path.join(dirpath, filename)
                
                # Extraire node_code
                node_match = FILE_PATTERN.search(filename)
                if not node_match:
                    continue
                
                node_code = int(node_match.group(1))
                
                # Extraire channel (peut ne pas exister)
                channel_match = CHANNEL_PATTERN.search(filename)
                channel = int(channel_match.group(1)) if channel_match else -1
                
                # Déterminer dataset (data vs aux)
                dataset = 'aux' if 'aux' in filename else 'data'
                
                stat = os.stat(filepath)
                mtime = int(stat.st_mtime)
                size = stat.st_size
                
                # Mise à jour summary
                summary.setdefault(node_code, {'data': False, 'aux': False, 'count': 0, 'last': 0})
                summary[node_code]['count'] += 1
                summary[node_code]['last'] = max(summary[node_code]['last'], mtime)
                if dataset == 'data':
                    summary[node_code]['data'] = True
                else:
                    summary[node_code]['aux'] = True
                
                # Insertion fichier
                cur.execute(
                    'INSERT INTO files (path, node_code, channel, dataset, size, mtime) VALUES (?, ?, ?, ?, ?, ?)',
                    (filepath, node_code, channel, dataset, size, mtime)
                )
                files_counter += 1

    print(f"✓ Indexed {files_counter} H5 files")

    # Insertion positions
    for node_code, stats in summary.items():
        has_data = 1 if stats['data'] else 0
        has_aux = 1 if stats['aux'] else 0
        last_seen = datetime.fromtimestamp(stats['last']).isoformat()
        cur.execute(
            'INSERT INTO positions (node_code, has_data, has_aux, sample_count, last_seen) VALUES (?, ?, ?, ?, ?)',
            (node_code, has_data, has_aux, stats['count'], last_seen)
        )

    conn.commit()
    
    # Stats finales
    total_positions = len(summary)
    with_data = sum(1 for s in summary.values() if s['data'])
    with_aux = sum(1 for s in summary.values() if s['aux'])
    
    print(f"✓ Rebuilt DB: {total_positions} positions total")
    print(f"  • With data files: {with_data}")
    print(f"  • With aux files: {with_aux}")
    print(f"  • Both: {sum(1 for s in summary.values() if s['data'] and s['aux'])}")
    print(f"  • Coverage: {(with_data/205*100):.1f}% (assuming 205 planned)")
    
    conn.close()

if __name__ == '__main__':
    rebuild_db()