"""docx → HTML 高层入口（集成清洗 + 转码 + 美化）"""
import os
import re
import zipfile
from typing import Callable

from .cleaner import clean_docx
from .post import post_docx
from .render import clean_html, stats as render_stats, copy_mathjax_to


def _zip_name(info) -> str:
    """zip 条目名：带 UTF-8 标志位(0x800)直接用；否则按 cp437→GBK 还原"""
    if info.flag_bits // 2048 % 2:
        return info.filename
    try:
        return info.filename.encode('cp437').decode('gbk')
    except Exception:
        return info.filename


def convert_one(docx_path: str, auto_clean: bool = True) -> dict:
    """单份 docx → HTML 字符串 + stats

    Returns: {ok, html, stats, error?}
    """
    try:
        if auto_clean:
            docx_path = clean_docx(docx_path)
        raw = post_docx(docx_path)
        html = clean_html(raw, mathjax=True)
        return {'ok': True, 'html': html, 'stats': render_stats(html)}
    except Exception as e:
        return {'ok': False, 'error': str(e)[:300]}


def batch_convert(path: str, out_dir: str, auto_clean: bool = True,
                  progress_cb: Callable | None = None) -> dict:
    """目录或 zip → 每份 docx 转 HTML 落 out_dir

    Returns: {count, errors[{file, error}], files[{name, group, stats}]}
    """
    os.makedirs(out_dir, exist_ok=True)
    copy_mathjax_to(out_dir)

    items: list[tuple[str, str]] = []
    if os.path.isdir(path):
        for f in sorted(os.listdir(path)):
            if f.lower().endswith('.docx'):
                items.append((f, os.path.join(path, f)))
    else:
        with zipfile.ZipFile(path) as z:
            for i in z.infolist():
                if i.is_dir() or i.filename.startswith('__MACOSX/') or not i.filename.lower().endswith('.docx'):
                    continue
                parts = [p for p in _zip_name(i).split('/') if p]
                rel = '/'.join(parts[1:]) if len(parts) > 1 else parts[-1]
                # 不解压到磁盘，直接 post bytes
                items.append((rel, zipfile.ZipFile(path).read(i)))

    files, errors = [], []
    for rel, src in items:
        try:
            if isinstance(src, bytes):
                # 写临时文件
                tmp = f'/tmp/_wpc_{abs(hash(rel)) & 0xFFFFFFFF:x}.docx'
                with open(tmp, 'wb') as f:
                    f.write(src)
                src = tmp
            r = convert_one(src, auto_clean=auto_clean)
            if r['ok']:
                html = r['html']
                rel_html = re.sub(r'\.docx$', '', rel, flags=re.I) + '.html'
                out = os.path.join(out_dir, rel_html)
                os.makedirs(os.path.dirname(out) or out_dir, exist_ok=True)
                with open(out, 'w', encoding='utf-8') as f:
                    f.write(html)
                files.append({
                    'name': rel_html,
                    'group': os.path.dirname(rel_html),
                    **r['stats'],
                })
            else:
                errors.append({'file': rel, 'error': r.get('error', '?')})
        except Exception as e:
            errors.append({'file': rel, 'error': str(e)[:200]})

        if progress_cb:
            progress_cb(len(files) + len(errors), len(items))

    return {'count': len(files), 'errors': errors, 'files': files}