#!/usr/bin/env python3
import os, json, uuid, re
from datetime import datetime

PDF_DIR = "/opt/hub/pdfs/files"
META = os.path.join(PDF_DIR, ".metadata.json")

TAG_RULES = [
    (r'fm[-_ ]?\d|atp[-_ ]?\d|tm[-_ ]?\d|tc[-_ ]?\d|army|military|combat|tactical|sniper|infantry|guerrilla|special.forces', 'defense'),
    (r'medic|first.aid|surgery|trauma|nursing|hesperian|midwif|dentist|disease|emergency.med|tccc|where.there.is.no.doctor', 'medical'),
    (r'survival|prepper|shtf|bug.out|wilderness|evasion', 'survival'),
    (r'radio|ham|antenna|arrl|fcc|comms|aprs|signal', 'comms'),
    (r'food|cook|cann|preserv|recipe|kitchen|baking|ferment', 'food'),
    (r'farm|garden|crop|seed|livestock|chicken|poultry|cattle|sheep|goat|beekeep|honey|agriculture|permaculture', 'agriculture'),
    (r'vet|animal|dog|cat|horse|pig|pet', 'veterinary'),
    (r'navig|map|compass|gps|bowditch|topo', 'navigation'),
    (r'solar|wind|battery|grid|power|nrel|energy|generator', 'energy'),
    (r'electric|plumb|carpent|hvac|welding|machinist|trade|repair|audel', 'trades'),
    (r'chem|biolog|physic|crc.handbook|merck.index', 'chemistry'),
    (r'engineer|mechanic|machine|handbook', 'engineering'),
    (r'law|legal|constitution|statute|code|court|right', 'law'),
    (r'finance|tax|irs|money|invest|insurance', 'finance'),
    (r'history|war|civil|world.war|geopolit', 'history'),
]

def infer_tag(filename):
    name = filename.lower()
    for pattern, tag in TAG_RULES:
        if re.search(pattern, name):
            return tag
    return 'general'

def clean_title(filename):
    name = filename.rsplit('.', 1)[0]
    name = re.sub(r'[_-]+', ' ', name).strip()
    return name[:200]

def main():
    if os.path.exists(META):
        with open(META) as f:
            meta = json.load(f)
    else:
        meta = {}

    registered = {info['filename'] for info in meta.values()}
    added = 0
    skipped = 0
    for fname in sorted(os.listdir(PDF_DIR)):
        if not fname.lower().endswith('.pdf'): continue
        if fname.startswith('.'): continue
        if fname in registered:
            skipped += 1
            continue
        path = os.path.join(PDF_DIR, fname)
        if os.path.getsize(path) < 50000: continue
        fid = uuid.uuid4().hex[:12]
        meta[fid] = {
            'filename': fname,
            'title': clean_title(fname),
            'tag': infer_tag(fname),
            'uploaded': datetime.utcnow().isoformat(),
        }
        added += 1

    with open(META, 'w') as f:
        json.dump(meta, f, indent=2)

    print(f"Added: {added}   Already registered: {skipped}   Total in library: {len(meta)}")

if __name__ == '__main__':
    main()