python -m pip install pandas openpyxl XMind==1.2.0

from pathlib import Path
import shutil
import tempfile
import xml.etree.ElementTree as ET
import zipfile

import pandas as pd
import xmind

LEVELS = ['L1', 'L2', '驱动杠杆', '业务域', '管理重点', '分析主题', '指标标准名称']
MANIFEST = 'META-INF/manifest.xml'
NS = 'urn:xmind:xmap:xmlns:manifest:1.0'


def normalize(value):
    if value is None or pd.isna(value):
        return ''
    return str(value).strip()


def rebuild_archive(source, target):
    """重建 ZIP，移除旧 manifest，生成一个且仅一个清单条目。"""
    ET.register_namespace('', NS)
    manifest = ET.Element(f'{{{NS}}}manifest')
    with zipfile.ZipFile(source, 'r') as src:
        names = src.namelist()
        if len(names) != len(set(names)):
            raise ValueError('输入 XMind 压缩包包含重复条目')
        if 'content.xml' not in names or src.testzip() is not None:
            raise ValueError('输入 XMind 压缩包缺少正文或校验失败')
        ET.fromstring(src.read('content.xml'))
        with zipfile.ZipFile(target, 'w', compression=zipfile.ZIP_DEFLATED) as dst:
            for info in src.infolist():
                if info.filename == MANIFEST:
                    continue
                dst.writestr(info, src.read(info.filename))
                if not info.is_dir() and not info.filename.startswith('META-INF/'):
                    ET.SubElement(manifest, f'{{{NS}}}file-entry', {'full-path': info.filename})
            dst.writestr(MANIFEST, ET.tostring(manifest, encoding='utf-8', xml_declaration=True))
    with zipfile.ZipFile(target) as check:
        if check.namelist().count(MANIFEST) != 1 or check.testzip() is not None:
            raise ValueError('输出 XMind 压缩包校验失败')
        ET.fromstring(check.read(MANIFEST))


def dataframe_to_xmind(df, output, levels=LEVELS, root_title='指标体系'):
    output = Path(output)
    if output.suffix.lower() != '.xmind':
        raise ValueError('输出文件必须使用 .xmind 扩展名')
    if not levels or len(levels) != len(set(levels)):
        raise ValueError('层级列必须非空且不能重复')
    if df.columns.has_duplicates:
        raise ValueError('输入表包含重复列名')
    missing = [name for name in levels if name not in df.columns]
    if missing:
        raise ValueError(f'缺少层级列：{missing}')
    if output.exists():
        raise FileExistsError(f'输出文件已存在：{output}')
    if not output.parent.is_dir():
        raise FileNotFoundError(f'输出目录不存在：{output.parent}')

    paths = []
    for row_number, values in enumerate(df.loc[:, levels].itertuples(index=False, name=None), start=2):
        titles = [normalize(value) for value in values]
        if not any(titles):
            continue
        end = next((i for i, title in enumerate(titles) if not title), len(titles))
        if any(titles[end:]):
            raise ValueError(f'Excel 第 {row_number} 行有中间缺失层级')
        paths.append(titles[:end])
    if not paths:
        raise ValueError('没有可转换的非空路径')

    # 在临时目录创建新工作簿，绝不加载目标文件后继续追加。
    with tempfile.TemporaryDirectory() as temp_dir:
        source = Path(temp_dir) / 'new.xmind'
        fixed = Path(temp_dir) / 'fixed.xmind'
        workbook = xmind.load(str(source))
        sheet = workbook.getPrimarySheet()
        sheet.setTitle(root_title)
        root = sheet.getRootTopic()
        root.setTitle(root_title)
        nodes = {(): root}
        for titles in paths:
            for depth, title in enumerate(titles, start=1):
                key = tuple(titles[:depth])
                if key not in nodes:
                    child = nodes[key[:-1]].addSubTopic()
                    child.setTitle(title)
                    nodes[key] = child
        xmind.save(workbook, str(source))
        rebuild_archive(source, fixed)
        # 独占创建处理“检查存在后，其他进程刚好创建文件”的竞争。
        with fixed.open('rb') as src, output.open('xb') as dst:
            try:
                shutil.copyfileobj(src, dst)
            except BaseException:
                dst.close()
                output.unlink(missing_ok=True)
                raise
    return output


if __name__ == '__main__':
    frame = pd.read_excel('input.xlsx', sheet_name=0, engine='openpyxl',
                          dtype=object, keep_default_na=False)
    result = dataframe_to_xmind(frame, 'output.xmind')
    print(f'已生成：{result}')