docxconvert/parsedocx.py

#-*-coding:utf-8 -*-
import re,os
import json,uuid
from bs4 import BeautifulSoup
from win32com.client import Dispatch
import pythoncom


class DocxConverter(object):
    """
    """
    def __init__(self,docpath="test4.docx"):
        """
        """
        self.docpath = docpath

    def docx2html(self):
        """
        """
        pythoncom.CoInitialize()
        self.word = Dispatch("Word.Application")
        self.word.Visible = 0
        self.doc = self.word.Documents.Open(self.docpath)
        html = os.path.join(os.path.dirname(self.docpath),str(uuid.uuid4())+".html")
        self.doc.SaveAs(html,10)
        self.doc.Close()
        self.word.Quit()
        pythoncom.CoUninitialize()
        return html

class QuestionsParser(object):
    """试题解析
    """
    def __init__(self,name="test4.html"):
        self.html = open(name,"r").read()
        self.soup = BeautifulSoup(self.html,"html.parser")

    def get_paragraphs(self):
        """
        """
        wordsection = self.soup.find("div",class_="WordSection1")
        #print wordsection
        pars = wordsection.find_all("p")
        return pars

    def parse_questions(self):
        """提取试题
        """
        que_type_dct = {}
        paragraphs = self.get_paragraphs()
        for i,p in enumerate(paragraphs):
            print p.text
            if u"【题型】" in p.text:
                que_type_dct["type"] = p.text.split("、")[-1] 

    def parse_questions(self):
        """提取试题
        """
        data = []
        tmp_val = {}
        tx_name = ""
        key = ""
        paragraphs = self.get_paragraphs()
        for i,p in enumerate(paragraphs):
            if u"【题型】" in p.text:
                tx_name = p.text 
            if u"【题干】" in p.text:
                key = "tg"
                tmp_val["tx"] = tx_name
                if tmp_val.get("tg"):
                    data.append(tmp_val)
                tmp_val = {"tg":"","tx":"","zsd":"","nd":"","da":"","jx":""}
            if u"【知识点】" in p.text:
                key = "zsd"
            if u"【难度】" in p.text:
                key = "nd"
            if u"【答案】" in p.text:
                key = "da"
            if u"【解析】" in p.text:
                key = "jx"

            if key != "":
                tmp_val[key] += p.__str__()

        data.append(tmp_val)

        return data

    def get_questions(self):
        """
        """
        questions = self.parse_questions()
        for que in questions:
            que["tx"] = que["tx"].split(u"、")[-1]
            #que["tg"] = que["tg"].replace(u"【题干】","")
            #que["zsd"] = que["zsd"].replace(u"【知识点】","")
            #que["da"] = que["da"].replace(u"【答案】","")
            #que["jx"] = que["jx"].replace(u"【解析】","")
            que["qno"] = self.get_qno(que["tg"])
        return questions

    def get_qno(self,tg):
        """提取题号
        """
        tgsoup = BeautifulSoup(tg,"html.parser")
        tgtext = tgsoup.text
        qno = re.search(r"\d+",tgtext.split(u"、")[0]).group()
        return qno
        
#docxconverter = DocxConverter() 
#questionparser =  QuestionsParser()

if __name__ == "__main__":
    #ques = questionparser.get_questions()
    #with open("t.json","w+") as f:
    #    f.write(json.dumps(ques))
    #docxconverter.docx2html()
    pass