main.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
HTML 表格 → 结构化文本
支持命令行两种调用方式：
1. 位置参数: python script.py /path/to/table.html 3
2. 选项参数: python script.py -html /path/to/table.html -nrows 3
"""

import pandas as pd
from pathlib import Path
import argparse
import sys

def merge_multi_header(header_rows):
    num_levels = len(header_rows)
    num_cols = len(header_rows[0])
    merged = []

    for col in range(num_cols):
        parts = []
        last = None
        for row in range(num_levels):
            val = str(header_rows[row][col])
            if val == last:
                continue
            parts.append(val)
            last = val
        merged.append("-".join(parts))
    return merged

def html_to_structured_text(html_input: str, header_rows_num=2):
    if html_input.strip().startswith("<"):
        tables = pd.read_html(html_input)
    else:
        html_file = Path(html_input)
        if not html_file.exists():
            raise FileNotFoundError(f"HTML文件不存在: {html_input}")
        tables = pd.read_html(html_file.read_text(encoding="utf-8"))

    if not tables:
        raise ValueError("未找到表格")

    df = tables[0]

    df_header = df.iloc[:header_rows_num, :]
    new_header = merge_multi_header(df_header.values.tolist())

    df_data = df.iloc[header_rows_num:, :]

    lines = []
    for _, row in df_data.iterrows():
        pairs = [(new_header[i], v) for i, v in enumerate(row) if pd.notna(v)]
        formatted = []
        last_key = None
        last_val = None
        for key, val in pairs:
            text = f"{key}: {val}"
            if formatted and key == last_key:
                if val == last_val:
                    continue
                formatted[-1] = f"{formatted[-1]}-{val}"
            else:
                formatted.append(text)
            last_key = key
            last_val =val
        lines.append("; ".join(formatted))

    return new_header, lines

def save_text(lines, path: Path):
    path.parent.mkdir(parents=True, exist_ok=True)
    path.write_text("\n".join(lines), encoding="utf-8")

def main(html_file: str, nrows=2):
    html_path = Path(html_file)
    if not html_path.exists():
        print(f"HTML 文件不存在: {html_file}")
        return

    txt_path = html_path.with_suffix(".txt")

    print(f"开始处理: {html_file}")
    headers, text_lines = html_to_structured_text(html_file, header_rows_num=nrows)
    save_text(text_lines, txt_path)
    print(f"处理完成。\n表头数量: {len(headers)}\n数据行数: {len(text_lines)}")
    print(f"输出文件: {txt_path}")

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="HTML 表格 → 结构化文本")

    # 位置参数
    parser.add_argument("html_file_pos", nargs="?", help="HTML 文件路径（位置参数方式）")
    parser.add_argument("nrows_pos", nargs="?", type=int, default=2, help="表头行数（位置参数方式）")

    # 选项参数
    parser.add_argument("-html", "--html_file", help="HTML 文件路径（选项参数方式）")
    parser.add_argument("-n", "--nrows", type=int, default=None, help="表头行数（选项参数方式）")

    args = parser.parse_args()

    # 优先使用选项参数
    html_file = args.html_file or args.html_file_pos
    nrows = args.nrows if args.nrows is not None else args.nrows_pos

    if not html_file:
        print("请提供 HTML 文件路径，例如：")
        print("  python script.py /path/to/table.html 3")
        print("  python script.py -html /path/to/table.html -n 3")
        sys.exit(1)

    main(html_file, nrows)