109 lines
4.0 KiB
Python
109 lines
4.0 KiB
Python
import os, re, json, h5py
|
|
from pathlib import Path
|
|
from datetime import datetime, timedelta
|
|
from tqdm import tqdm
|
|
|
|
DATA_ROOTS = [Path("/mnt/kingston"), Path("/mnt/data_sdb1")]
|
|
POSITIONS_CSV = Path("/mnt/kingston/Copie de SETE_AUV_DARFV4-Copier(1).csv")
|
|
OUTPUT_INDEX = Path("/tmp/index.json")
|
|
SAMPLE_RATE = 200
|
|
|
|
def load_pos():
|
|
positions = {}
|
|
if not POSITIONS_CSV.exists(): return {}
|
|
with open(POSITIONS_CSV, 'r', encoding='utf-8', errors='replace') as f:
|
|
lines = f.readlines()
|
|
if len(lines) < 5: return {}
|
|
headers = lines[3].strip().split(',')
|
|
try:
|
|
ni = headers.index('NodeCode')
|
|
ei = headers.index('Aslaid Easting') if 'Aslaid Easting' in headers else headers.index('Preplot Easting')
|
|
oi = headers.index('Aslaid Northing') if 'Aslaid Northing' in headers else headers.index('Preplot Northing')
|
|
except: return {}
|
|
for line in lines[4:]:
|
|
parts = line.strip().split(',')
|
|
try:
|
|
nid = parts[ni].strip()
|
|
positions[nid] = {
|
|
'easting': float(parts[ei]),
|
|
'northing': float(parts[oi]),
|
|
'depth': float(parts[headers.index('Aslaid Depth')]) if 'Aslaid Depth' in headers else 0.0
|
|
}
|
|
except: continue
|
|
return positions
|
|
|
|
def scan():
|
|
pos = load_pos()
|
|
nodes = {}
|
|
all_dates = set()
|
|
file_count = 0
|
|
|
|
print("🔍 Scanning all H5 files (Trusting filenames/folders for dates)...")
|
|
all_h5_files = []
|
|
for root in DATA_ROOTS:
|
|
all_h5_files.extend(list(root.rglob("*.h5")))
|
|
|
|
for h5_path in tqdm(all_h5_files):
|
|
try:
|
|
# Pattern: auto_{julian}_{time}_b{node}_..._{ts}.h5
|
|
match = re.search(r'auto_(\d+)_(\d{6})_b(\d+)_.*?_(\d{10})\.h5$', h5_path.name)
|
|
if not match: continue
|
|
|
|
julian_day = int(match.group(1))
|
|
time_str = match.group(2)
|
|
node_id = match.group(3)
|
|
internal_ts = int(match.group(4))
|
|
|
|
# Calculer la date réelle à partir du Julian Day (2020)
|
|
# Julian 1 = Jan 1. Julian 255 = Sept 11.
|
|
date_ref = datetime(2020, 1, 1) + timedelta(days=julian_day - 1)
|
|
date_str = date_ref.strftime('%Y-%m-%d')
|
|
|
|
# Heure du fichier
|
|
hours = int(time_str[:2])
|
|
minutes = int(time_str[2:4])
|
|
seconds = int(time_str[4:6])
|
|
|
|
# Timestamp calculé (plus fiable pour le matching que l'interne buggé)
|
|
actual_start_ts = int(datetime(2020, 1, 1).timestamp() + (julian_day - 1) * 86400 + hours * 3600 + minutes * 60 + seconds)
|
|
|
|
with h5py.File(h5_path, 'r') as f:
|
|
if 'adc_values' not in f: continue
|
|
duration = f['adc_values'].shape[0] / SAMPLE_RATE
|
|
actual_end_ts = actual_start_ts + duration
|
|
|
|
all_dates.add(date_str)
|
|
|
|
if node_id not in nodes:
|
|
nodes[node_id] = {
|
|
'id': node_id,
|
|
'position': pos.get(node_id),
|
|
'files': []
|
|
}
|
|
|
|
nodes[node_id]['files'].append({
|
|
'path': str(h5_path),
|
|
'start': actual_start_ts,
|
|
'end': actual_end_ts,
|
|
'julian': julian_day,
|
|
'channels': ['ch0', 'ch1', 'ch2', 'ch3']
|
|
})
|
|
file_count += 1
|
|
except: continue
|
|
|
|
# Sauvegarder l'index complet
|
|
result = {
|
|
'generated_at': datetime.now().isoformat(),
|
|
'sample_rate_hz': SAMPLE_RATE,
|
|
'nodes': nodes,
|
|
'dates': sorted(list(all_dates))
|
|
}
|
|
|
|
with open(OUTPUT_INDEX, 'w') as f:
|
|
json.dump(result, f, indent=2)
|
|
|
|
print(f"✅ Indexing complete: {file_count} files, {len(nodes)} nodes, {len(all_dates)} dates.")
|
|
print(f"📅 Dates covered: {sorted(list(all_dates))}")
|
|
|
|
if __name__ == '__main__': scan()
|