import re
import subprocess


def html_to_word_pandoc(html_file, output_file):
    subprocess.run(['pandoc', html_file, '-o', output_file])


def docx_to_markdown_pandoc(docx_file, output_file):
    subprocess.run(['pandoc', docx_file, '-f', 'docx', '-t', 'markdown', '-o', output_file])


docx_file = 'D:\dsWork\dsProject\dsRag\static\Txt\化学方程式_CHEMISTRY_1.docx'
output_file = 'c:/output.md'
docx_to_markdown_pandoc(docx_file, output_file)
finalFile = "c:/new.txt"

# 读取然后修改内容，输出到新的文件
idx=0
with open(finalFile, 'w', encoding='utf-8') as f1:
    with open(output_file, 'r', encoding='utf-8') as f:
        for line in f:
            if line.strip():
                # 改进后的正则表达式，匹配更多格式的MathType公式
                if re.search(r'!\[]\(media/image\d+\.\w+\)', line) or \
                        re.search(r'\.!\[]\(media/image\d+\.\w+\)\.', line):
                    idx=idx+1
                    f1.write("【MathType"+str(idx)+"】\n")
                else:
                    f1.write(line.strip()+"\n")

# 输出finalFile
with open(finalFile, 'r', encoding='utf-8') as f:
    print(f.read())