FORMAT_CONVERSION_MAXCOMPUTE/format_convert/convert_doc.py

import inspect
import os
import re
import sys

import chardet
from bs4 import BeautifulSoup
sys.path.append(os.path.dirname(__file__) + "/../")
from format_convert.convert_tree import _Document, _Sentence, _Page
import logging
import traceback
from format_convert import get_memory_info
from format_convert.convert_docx import docx2text, DocxConvert
from format_convert.convert_need_interface import from_office_interface
from format_convert.utils import judge_error_code, get_logger, log


@get_memory_info.memory_decorator
def doc2text(path, unique_type_dir):
    log("into doc2text")
    try:
        # 调用office格式转换
        file_path = from_office_interface(path, unique_type_dir, 'docx')
        if judge_error_code(file_path):
            return file_path

        text = docx2text(file_path, unique_type_dir)
        return text
    except Exception as e:
        log("doc2text error!")
        print("doc2text", traceback.print_exc())
        return [-1]


class DocConvert:
    def __init__(self, path, unique_type_dir):
        self._doc = _Document(path)
        self.path = path
        self.unique_type_dir = unique_type_dir

    def convert(self):
        # 先判断特殊doc文件，可能是html文本
        is_html_doc = False
        try:
            try:
                with open(self.path, 'r') as f:
                    html_str = f.read()
            except UnicodeDecodeError:
                with open(self.path, 'r', errors='ignore') as f:
                    html_str = f.read()
            # if re.search('<div|<html|<body|<head|<tr|<br|<table|<td|<p>|<span', html_str):
            if len(re.findall('<div|<html|<body|<head|<tr|<br|<table|<td|<p>|<span', html_str)) >= 10:
                log('doc as html!')
                soup = BeautifulSoup(html_str, 'lxml')
                text = soup.text
                is_html_doc = True
        except:
            pass

        if is_html_doc:
            self._page = _Page(None, 0)
            _sen = _Sentence(text, (0, 0, 0, 0))
            self._page.add_child(_sen)
            self._doc.add_child(self._page)
        else:
            # 调用office格式转换
            file_path = from_office_interface(self.path, self.unique_type_dir, 'docx')
            if judge_error_code(file_path):
                self._doc.error_code = file_path
                return
            _docx = DocxConvert(file_path, self.unique_type_dir)
            _docx.convert()
            self._doc = _docx._doc

    def get_html(self):
        try:
            self.convert()
        except:
            traceback.print_exc()
            self._doc.error_code = [-1]
        if self._doc.error_code is not None:
            return self._doc.error_code
        # print(self._doc.children)
        return self._doc.get_html()


if __name__ == '__main__':
    c = DocConvert("C:/Users/Administrator/Downloads/-4274446916340743056.doc", "C:/Users/Administrator/Downloads/1")
    print(c.get_html())