Python：提取word中关键内容并导出到其它word和excel表格中

一、前言

新部门接到一个新需求，要求根据以前的会议纪要，提取相关信息（如下图所示）

包括了会议名称、时间、地点、主持人、出席人员、列席人员、缺席人员、会议内容、汇报人、列席人等等，然后要生成两样东西：

1、会议通知

右下角是会议通知时间，根据会议时间往前倒推两天自动生成。

2、会议总表

二、主要难点

1、原来的文件都是doc格式的，python的docx库不能读取，所以必须要靠win32转换成docx；
2、对docx的库使用不多，所以提取和写入表格的代码都是百度了好久获得的；
3、写入excel不难，只是参会人要把“出席人员”、“列席人员”和“缺席人员”组合起来，会议时间也只要日期不要时间；
4、最后还需要把word转成pdf，又使用了win32的库。

三、具体代码

# coding=gbk
import docx
from win32com.client import Dispatch
from docx.shared import Pt
from docx.enum.text import WD_ALIGN_PARAGRAPH
import os
import datetime
import traceback

def excel_pre():
    '''启动excel和路径设置'''
    global xl
    xl = Dispatch("Excel.Application")
    xl.Visible = False #True是显示， False是隐藏
    xl.DisplayAlerts = 0

def doc2Docx(fileName):
    '''将doc转换为docx'''
    word = Dispatch("Word.Application")
    doc = word.Documents.Open(fileName)
    doc.SaveAs(fileName + "x", 12, False, "", True, "", False, False, False, False)
    os.remove(fileName)
    doc.Close()
    word.Quit()

def dict_judge(text):
    '''判断text中是否存在dict中的key'''
    num_dict = {'一、':1, '二、':2, '三、':3, '四、':4, '五、':5, '六、':6, '七、':7, '八、':8, '九、':9}
    for key,value in num_dict.items():
        if key in text:
            return value
    return 0

def notice_time(meeting_time, timedelta=-2):
    '''根据会议时间倒推会议通知时间'''
    a = datetime.datetime.strptime(meeting_time, "%Y-%m-%d")
    b = a + datetime.timedelta(days=timedelta)
    c = b.strftime('%Y-%m-%d')
    return c

def report_time(meeting_time, num):
    '''模拟生成汇报时间report_time'''
    if '午' in meeting_time:
        meeting_time = meeting_time.replace('午', '0')
    a = datetime.datetime.strptime(meeting_time, '%H:%M')
    if num ==1:
        return meeting_time
    else:
        b = a + datetime.timedelta(minutes=10*(num-1))
        c = b.strftime('%H:%M')
        return c

def doc2pdf(input_file):
    '''把word转成pdf'''
    word = Dispatch('Word.Application')
    doc = word.Documents.Open(input_file)
    doc.SaveAs(input_file.replace(".docx", ".pdf"), FileFormat=17)
    doc.Close()
    word.Quit()

def set_font(paragraph):
    '''设定word中的字体大小'''
    paragraph.paragraph_format.left_indent = Pt(0) #取消左缩进
    paragraph.paragraph_format.right_indent = Pt(0)  # 取消右缩进
    run = paragraph.runs
    font = run[0].font
    font.size= Pt(14) #14对应四号字体
    # font.bold = True #加粗

def get_meeting_info(docname):
    '''提取会议纪要里的内容'''
    doc = docx.Document(meeting_file_path + docname)
    meeting_name = docname.replace('纪要.docx', '')
    meeting_name = '上海NOC ' + meeting_name
    meeting_dict = {'会议名称':meeting_name, '会议时间': '', '主持人': '', '参会人': ''}
    notice_dict = {'会议名称':meeting_name, '会议时间':'', '会议地点':'', '会议主持':'', '出席人员': ''}
    joiner = ''
    for i, paragraph in enumerate(doc.paragraphs):
        this_text = paragraph.text
        num = dict_judge(this_text)
        if '会议时间' in this_text:
            this_text = this_text.split('：')[1]
            notice_dict['会议时间'] = this_text
            meeting_date = this_text[:10] #会议日期
            meeting_time = this_text[-5:] #会议时间
            notice_date = notice_time(meeting_date)
            meeting_dict['会议时间'] = meeting_date
            notice_dict['通知时间'] = notice_date
        elif '会议地点' in this_text:
            this_text = this_text.split('：')[1]
            notice_dict['会议地点'] = this_text
        elif '会议主持' in this_text:
            this_text = this_text.split('：')[1]
            meeting_dict['主持人'] = notice_dict['会议主持'] = this_text
        elif '出席人员' in this_text:
            this_text = this_text.split('：')[1]
            # notice_dict['出席人员'] = this_text
            joiner = joiner + this_text + '、'
        elif '列席人员' in this_text:
            this_text = this_text.split('：')[1]
            joiner = joiner + this_text + '、'
        elif '缺席人员' in this_text:
            this_text = this_text.split('：')[1]
            if this_text == '无':
                joiner = joiner[:-1]
            else:
                joiner = joiner + this_text
        elif num:
            this_title = this_text.split('、', 1)[1].replace('。', '')
            try:
                this_reporter = doc.paragraphs[i+1].text.split('：', 1)[1]
                if '回避' in this_reporter:
                    this_reporter = this_reporter.split('（', 1)[0]
            except:
                this_reporter = ''
            if '列席人' in this_reporter:
                this_reporter, liexiren = this_reporter.split('（', 1)
                liexiren = liexiren.split('：')[1].replace('）', '')
            else:
                liexiren = ''
            this_report_time = report_time(meeting_time, num)
            notice_dict[num] = [this_title, this_reporter, liexiren, this_report_time]
    meeting_dict['参会人'] = joiner
    if '（开会）' in joiner:
        joiner = joiner.replace('（开会）', '')
    notice_dict['出席人员'] = joiner
    return meeting_dict, notice_dict

def make_report(report_dict):
    '''制作会议通知'''
    doc = docx.Document('会议通知模板.docx')
    #插入主要通知内容
    doc_cell_dict = {'会议名称': (0, 1), '会议时间': (1, 1), '会议地点': (2, 1), '会议主持': (3, 1), '出席人员': (4, 1)}
    for key, value in doc_cell_dict.items():
        doc.tables[0].rows[value[0]].cells[value[1]].text = report_dict[key]
        paragraph = doc.tables[0].rows[value[0]].cells[value[1]].paragraphs[0]
        set_font(paragraph)
    #插入几个会议内容
    for i in range(1, len(report_dict)-5):
        doc.tables[0].rows[i + 5].cells[1].text = report_dict[i][0]  # 会议内容
        doc.tables[0].rows[i + 5].cells[2].text = report_dict[i][1]  # 汇报人
        doc.tables[0].rows[i + 5].cells[3].text = report_dict[i][2]  # 专题列席人
        doc.tables[0].rows[i + 5].cells[4].text = report_dict[i][3]  # 汇报时间
        for j in range(1, 5):
            paragraph = doc.tables[0].rows[i + 5].cells[j].paragraphs[0]
            set_font(paragraph)
    #创建新的格式
    try:
        style1 = doc.styles['style head1']
    except:
        style1 = doc.styles.add_style('style head1', 2)
    finally:
        style1.font.bold = True
        style1.font.name = u'宋体 (中文正文)'
        style1.font.size = Pt(14)
    #在最后插入通知时间
    e = doc.add_paragraph()
    e.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.RIGHT #右对齐
    e.add_run(report_dict['通知时间'], style=style1)
    #最后根据名字保存会议通知
    doc.save(report_file_path + '会议通知：'+ report_dict['会议名称'] + '.docx')
    #把会议通知转成PDF格式
    doc2pdf(report_file_path + '会议通知：'+ report_dict['会议名称'] + '.docx')

def make_newxls(meeting_dict, newrow):
    '''把meetring的信息导入到excel里'''
    ws.Cells(newrow, 1).Value = meeting_dict['会议名称']
    ws.Cells(newrow, 3).Value = '党委（党组）会'
    ws.Cells(newrow, 4).Value = '现场会议'
    ws.Cells(newrow, 5).Value = meeting_dict['会议时间']
    ws.Cells(newrow, 6).Value = meeting_dict['主持人']
    ws.Cells(newrow, 7).Value = meeting_dict['参会人']

if __name__ == "__main__":
    #如果原始文档是doc格式的话，就需要先批量转换为docx
    # for f in os.listdir('会议纪要库'):
    #     if f.endswith('.doc'):
    #         doc2Docx(meeting_file_path + f)
    meeting_file_path = os.path.abspath('.') + '\\' + '会议纪要库' + '\\'
    report_file_path = os.path.abspath('.') + '\\' + '会议通知库' + '\\'
    this_path = os.path.abspath('.') + '\\'
    excel_pre()
    wb = xl.Workbooks.Open(this_path + '决策会议采集模板.xls')
    ws = wb.Sheets('决策会议')
    n = 3
    try:
        for f in os.listdir('会议纪要库'):
            if f.endswith('.docx'):
                print(f)
                # doc2pdf(meeting_file_path + f)   # 把会议通知转成PDF格式（在“会议纪要库”里）
                meeting_dict, notice_dict = get_meeting_info(f)
                make_report(notice_dict) # 生成会议通知（在“会议通知库”里）
                make_newxls(meeting_dict, n) #把记录放在《决策会议采集模板.xls》里
                n +=1
    except:
        traceback.print_exc()
    finally:
        wb.Save()
        wb.Close()

成果：