# Hardcoded string scanner for ObsiGate frontend JS files
# Finds user-visible strings that should be i18n'd but aren't.
# Also checks EN locale for French contamination.
#
# Usage:
#   cd /c/dev/git/python/obsigate
#   python3 << 'PYEOF' ... PYEOF

import re, os, json

js_dir = 'frontend/js'

# ── Part 1: Find hardcoded user-visible strings in JS ──
noise_patterns = [
    r'\.css', r'data-', r'console\.', r'getElementById', r'querySelector',
    r'\.json', r'/api/', r'CSS\.escape', r'encodeURI', r'font-size',
    r'background:', r'padding:', r'border', r'SSE parse', r'\[Offline',
    r'Failed to', r'Error ', r'Soft vault', r'\[Forge\]', r'CodeMirror',
    r'\[PWA\]', r'\[SW\]', r'DragDrop', r'Dashboard:',
]

french_chars = re.compile(r'[àâäéèêëîïôöùûüçÀÂÄÉÈÊËÎÏÔÖÙÛÜÇ]')
ui_words = ['loading', 'error', 'search', 'save', 'delete', 'create',
            'cancel', 'download', 'upload', 'share', 'copy', 'rename',
            'open', 'close', 'file', 'folder', 'vault']

for fname in sorted(os.listdir(js_dir)):
    if not fname.endswith('.js') or fname == 'i18n.js':
        continue
    fpath = os.path.join(js_dir, fname)
    with open(fpath, 'r', encoding='utf-8') as f:
        lines = f.readlines()

    hardcoded = []
    for i, line in enumerate(lines, 1):
        strings = re.findall(r"'([^']+)'|\"([^\"]+)\"|`([^`]+)`", line)
        for s_tuple in strings:
            s = s_tuple[0] or s_tuple[1] or s_tuple[2]
            if len(s) < 4: continue
            if not re.search(r'[A-Za-zÀ-ÿ]', s): continue
            if any(re.search(p, s) for p in noise_patterns): continue
            if s.startswith('#') or s.startswith('.') or s.startswith('http'): continue

            context = ''.join(lines[max(0,i-2):i+1])
            if 't(' in context or 'data-i18n' in context or 'tr(' in context:
                continue

            has_fr = bool(french_chars.search(s))
            is_ui = has_fr or any(w in s.lower() for w in ui_words)
            if is_ui:
                hardcoded.append((i, s.strip()[:120]))

    if hardcoded:
        print(f"\n=== {fname} ({len(hardcoded)}) ===")
        for ln, s in hardcoded:
            print(f"  L{ln}: {s}")

# ── Part 2: Check EN locale for French contamination ──
with open('frontend/locales/fr.json', encoding='utf-8') as f:
    fr = json.load(f)
with open('frontend/locales/en.json', encoding='utf-8') as f:
    en = json.load(f)

print(f"\n=== Locale alignment: FR={len(fr)} EN={len(en)} ===")

only_fr = set(fr.keys()) - set(en.keys())
only_en = set(en.keys()) - set(fr.keys())
if only_fr: print(f"Keys in FR only: {sorted(only_fr)}")
if only_en: print(f"Keys in EN only: {sorted(only_en)}")

# EN values with French chars
fr_in_en = []
for k in sorted(fr.keys()):
    if k in en and isinstance(en[k], str):
        if french_chars.search(en[k]):
            fr_in_en.append(k)

if fr_in_en:
    print(f"\n=== EN keys with French text ({len(fr_in_en)}) ===")
    for k in fr_in_en:
        if not k.startswith('auto.') and not k.startswith('help.desc_'):
            print(f"  {k}: {en[k][:80]}")