lucky_shop/scripts/i18n_clean_dead.py

"""
i18n_clean_dead.py - 从 lang/en-us.js + lang/zh-cn.js 删除未被任何代码调用的死 key

设计：
- 解析 lang/<locale>.js 的全部 keys（带嵌套追踪）
- 扫描所有 .vue / .js（非 lang/、非 node_modules）的 $lang('xxx') / $lang("xxx") / this.$lang('xxx')
- 计算 dead keys = defined - used
- 对每个 dead key：
  - 如果是嵌套对象内的叶子值（字符串），从父对象里删
  - 如果是嵌套对象的最后一个子项，整个对象也删（如果是空对象了）
  - 同步删除 en-us 和 zh-cn 两侧

退出码：
  0 = 成功
  非 0 = 有 dead key 列出但未删（dry-run 默认）
  --apply 才会真删
"""
import re
import json
import sys
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'
EXCLUDE_DIRS = {'node_modules', 'unpackage', 'stub-app-projects', '.git', 'lang'}


def parse_keys_with_lines(path):
    """Parse lang file with nesting + line tracking.
    Returns: dict of full_key -> (value, line_no, indent, parent_path)
    """
    text = path.read_text(encoding='utf-8')
    out = {}
    indent_stack = [(-1, '', -1)]  # (indent, parent_path, parent_line)
    lines = text.split('\n')
    for i, ln in enumerate(lines, 1):
        stripped = ln.lstrip()
        if stripped.startswith('//') or not stripped.strip():
            continue
        indent = len(ln) - len(stripped)
        m = re.match(r"""^['\"]?([A-Za-z_][A-Za-z0-9_]*)['\"]?\s*:\s*(.*)$""", stripped)
        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_path = indent_stack[-1][1]
        full_key = f'{parent_path}.{key}' if parent_path else key
        if rest.endswith('{'):
            indent_stack.append((indent, full_key, i))
            out[full_key] = ('<object>', i, indent, parent_path)
        else:
            v = rest
            if (v.startswith("'") and v.endswith("'")) or (v.startswith('"') and v.endswith('"')):
                if len(v) >= 2:
                    v = v[1:-1]
            out[full_key] = (v, i, indent, parent_path)
    return out


def collect_used_keys():
    """Scan all .vue / .js files for $lang('xxx') / this.$lang('xxx') calls"""
    used = set()
    call_re = re.compile(r"\$lang\(\s*['\"]([^'\"]+)['\"]")
    this_re = re.compile(r"this\.\$lang\(\s*['\"]([^'\"]+)['\"]")
    for p in list(ROOT.rglob('*.vue')) + list(ROOT.rglob('*.js')):
        if any(d in p.parts for d in EXCLUDE_DIRS):
            continue
        if p.name.startswith('i18n_') or p.name.startswith('_verify') or p.name.startswith('_find'):
            continue
        try:
            text = p.read_text(encoding='utf-8')
        except Exception:
            continue
        if '$lang' not in text:
            continue
        # strip comments to avoid false positives
        text_no_comments = re.sub(r'//.*?$|/\*.*?\*/', '', text, flags=re.DOTALL | re.MULTILINE)
        for m in call_re.finditer(text_no_comments):
            used.add(m.group(1))
        for m in this_re.finditer(text_no_comments):
            used.add(m.group(1))
    return used


def remove_dead_keys_from_file(path, dead_keys):
    """从 lang 文件删除 dead_keys。返回 (new_text, removed_count)"""
    text = path.read_text(encoding='utf-8')
    lines = text.split('\n')

    # Build line removal set
    lines_to_remove = set()
    removed = 0

    # Parse again with full info
    keys_info = parse_keys_with_lines(path)
    for dead_key in dead_keys:
        if dead_key in keys_info:
            val, line_no, indent, parent = keys_info[dead_key]
            lines_to_remove.add(line_no - 1)  # 0-indexed
            removed += 1

    new_lines = [ln for i, ln in enumerate(lines) if i not in lines_to_remove]
    return '\n'.join(new_lines), removed


def main():
    apply = '--apply' in sys.argv

    en_keys = parse_keys_with_lines(EN_FILE)
    zh_keys = parse_keys_with_lines(ZH_FILE)
    used = collect_used_keys()

    en_defined = {k for k, v in en_keys.items() if v[0] != '<object>'}
    zh_defined = {k for k, v in zh_keys.items() if v[0] != '<object>'}
    all_defined = en_defined | zh_defined

    dead = sorted(all_defined - used)

    print(f'en-us defined: {len(en_defined)}')
    print(f'zh-cn defined: {len(zh_defined)}')
    print(f'unique used: {len(used)}')
    print(f'dead keys: {len(dead)}')
    print()

    if not dead:
        print('[OK] No dead keys to remove')
        return 0

    print('Dead keys to remove:')
    for k in dead[:50]:
        print(f'  - {k}')
    if len(dead) > 50:
        print(f'  ... and {len(dead) - 50} more')

    if not apply:
        print('\n[DRY-RUN] 未实际删除文件。加 --apply 参数真正删除。')
        return 1

    # 真删
    new_en, removed_en = remove_dead_keys_from_file(EN_FILE, dead)
    new_zh, removed_zh = remove_dead_keys_from_file(ZH_FILE, dead)

    EN_FILE.write_text(new_en, encoding='utf-8')
    ZH_FILE.write_text(new_zh, encoding='utf-8')

    print(f'\n[OK] en-us: removed {removed_en} keys')
    print(f'[OK] zh-cn: removed {removed_zh} keys')

    # 写报告
    out = ROOT / 'docs' / 'i18n-clean-dead.log'
    out.write_text(
        f'en-us removed: {removed_en}\n'
        f'zh-cn removed: {removed_zh}\n\n'
        'Removed keys:\n' + '\n'.join(dead),
        encoding='utf-8'
    )
    print(f'[OK] Log -> {out.relative_to(ROOT)}')
    return 0


if __name__ == '__main__':
    sys.exit(main())