lucky_shop/scripts/i18n_rebuild_from_vue.py

"""
i18n_rebuild_from_vue.py - 用真实 .vue 中文重建 lang/zh-cn.js

数据源：
1. docs/i18n-vue-extract.json - 1186 个 (pre-f258987 key, 真实中文) 对
2. lang/_migration-map.json - 326 个 (old_key, new_key) 对

组合：
- final[new_key] = vue_extract[old_key]  如果 old_key in migration_map
- final[new_key] = vue_extract[old_key]  如果 old_key == new_key（未改名）

然后用 en-us 端已有的合理英文翻译 + 真实中文重建
"""
import re
import json
import sys
import subprocess
from pathlib import Path
from collections import defaultdict

ROOT = Path(__file__).resolve().parent.parent
EN_FILE = ROOT / 'lang' / 'en-us.js'
ZH_FILE = ROOT / 'lang' / 'zh-cn.js'
EXTRACT_FILE = ROOT / 'docs' / 'i18n-vue-extract.json'
MAP_FILE = ROOT / 'lang' / '_migration-map.json'

USER_DIRS = ['components', 'components-diy', 'pages', 'pages_goods',
             'pages_order', 'pages_promotion', 'pages_tool', 'store']
SOURCE_REF = 'f258987'


def main():
    # 1. 读 1186 个 (old_key, value)
    extract = json.loads(EXTRACT_FILE.read_text(encoding='utf-8'))
    print(f'从 .vue 提取 (旧 key, 真实中文): {len(extract)}')

    # 2. 读 migration map
    mig = json.loads(MAP_FILE.read_text(encoding='utf-8')).get('old_to_new', {})
    print(f'迁移映射 (old_key, new_key): {len(mig)}')

    # 3. 组合: 旧 key -> 新 key
    final_zh = {}
    unmapped = []
    for old_key, zh in extract.items():
        new_key = mig.get(old_key, old_key)  # 未迁移的保持原 key
        if new_key in final_zh:
            continue  # 已存在
        final_zh[new_key] = zh
        if old_key != new_key and old_key not in mig:
            unmapped.append(old_key)
    print(f'组合后 unique 新 key: {len(final_zh)}')
    if unmapped:
        print(f'  未迁移的 (保持原 key): {len(unmapped)}')
        for k in unmapped[:5]:
            print(f'    {k}')

    # 4. 收集当前 .vue 调用的所有 key (检查覆盖)
    used_keys = set()
    for d in USER_DIRS:
        for p in (ROOT / d).rglob('*.vue'):
            try:
                text = p.read_text(encoding='utf-8', errors='ignore')
            except: continue
            for m in re.finditer(r"""\$\s*lang\s*\(\s*['"]([^'"]+)['"]\s*\)""", text):
                used_keys.add(m.group(1))
            for m in re.finditer(r"""this\s*\.\s*\$\s*lang\s*\(\s*['"]([^'"]+)['"]\s*\)""", text):
                used_keys.add(m.group(1))
    print(f'当前 .vue 调用的 key: {len(used_keys)}')

    # 5. 用 current lang/en-us.js 的 en 翻译作为基础（保留已有英文翻译）
    # 解析 en-us.js
    en_text = EN_FILE.read_bytes().decode('utf-8', errors='ignore')
    en_dict = {}
    indent_stack = [(-1, '')]
    for ln in en_text.split('\n'):
        if ln.strip().startswith('//') or not ln.strip(): continue
        indent = len(ln) - len(ln.lstrip())
        m = re.match(r"""^['\"]?([A-Za-z_][A-Za-z0-9_.]*)['\"]?\s*:\s*(.*)$""", ln.lstrip())
        if not m: continue
        key, rest = m.group(1), m.group(2).strip().rstrip(',')
        while len(indent_stack) > 1 and indent_stack[-1][0] >= indent: indent_stack.pop()
        parent = indent_stack[-1][1]
        full_key = f'{parent}.{key}' if parent else key
        if rest.endswith('{'): indent_stack.append((indent, full_key)); continue
        v = rest
        if (v.startswith('"') and v.endswith('"')) or (v.startswith("'") and v.endswith("'")):
            if len(v) >= 2: v = v[1:-1]
        en_dict[full_key] = v

    # 6. 输出统计
    print(f'\n=== 最终统计 ===')
    print(f'en-us 已有: {len(en_dict)} keys')
    print(f'zh-cn 从 .vue 提取: {len(final_zh)} keys')
    used_in_zh = sum(1 for k in used_keys if k in final_zh)
    print(f'vue 调用 key 在 zh-cn 有翻译: {used_in_zh} / {len(used_keys)}')

    # 写
    apply = '--apply' in sys.argv
    if not apply:
        print('\n[DRY-RUN] 加 --apply 真写文件')
        return 0

    # 写 zh-cn.js: 用 final_zh 优先，en-us 没有的留空
    # 写 en-us.js: 保留已有 en_dict，但 final_zh 里 .vue 调用的 key 缺英文的，标记 TODO
    write_lang(ZH_FILE, final_zh, 'zh-cn')
    write_lang(EN_FILE, en_dict, 'en-us')
    return 0


def write_lang(path, d, locale):
    """写带双引号的 lang/<locale>.js"""
    grouped = defaultdict(dict)
    top_level = {}
    for k, v in d.items():
        parts = k.split('.')
        if len(parts) == 1:
            top_level[k] = v
        else:
            ns = parts[0]
            sub = '.'.join(parts[1:])
            grouped[ns][sub] = v

    lines = [
        f'// Auto-generated by i18n_rebuild_from_vue.py',
        f'// Single authoritative lang source. Do not edit manually.',
        f'// locale: {locale}',
        f'// Generated: 2026-06-23',
        '',
        'export const lang = {',
    ]
    # namespace groups
    written = set()
    nested_ns = {'pageTitles', 'common', 'tabBar', 'customer'}
    for ns in sorted(set(grouped.keys()) | set(top_level.keys())):
        if ns in nested_ns and ns in grouped:
            lines.append(f'\t"{ns}": {{')
            for sub_k in sorted(grouped[ns].keys()):
                sub_v = grouped[ns][sub_k]
                escaped = sub_v.replace('\\', '\\\\').replace('"', '\\"')
                lines.append(f'\t\t"{sub_k}": "{escaped}",')
            lines.append('\t},')
            written.add(ns)
    for ns in sorted(grouped.keys()):
        if ns in written: continue
        lines.append(f'\t"{ns}": {{')
        for sub_k in sorted(grouped[ns].keys()):
            sub_v = grouped[ns][sub_k]
            escaped = sub_v.replace('\\', '\\\\').replace('"', '\\"')
            lines.append(f'\t\t"{sub_k}": "{escaped}",')
        lines.append('\t},')
        written.add(ns)
    for k in sorted(top_level.keys()):
        if k in written: continue
        v = top_level[k]
        escaped = v.replace('\\', '\\\\').replace('"', '\\"')
        lines.append(f'\t"{k}": "{escaped}",')

    lines.append('};')
    lines.append('')
    content = '\n'.join(lines)
    path.write_bytes(content.encode('utf-8'))
    print(f'[OK] {path.relative_to(ROOT)}: {len(d)} keys')


if __name__ == '__main__':
    main()