pdf内容差异对比

场景：防止抄袭！word对比就很容易，所以这里介绍pdf内容对比技术。

一、环境

• win10
• python3.9.6

二、思路

pdf转图片jpg,图片转文字，文字进行对比。

1. 将 PDF 文件的每一页转换为一个图片
2. 图片转为文字，对文字进行比较，再得到一个差异图。
3. 将所有生成的差异图像拼接成一个 PDF 文件

三、实现效果

使用方法：python diff-pdf.py test1.pdf test2.pdf
结果如下：

四、部分程序

主要使用模块：pdfminer，io中的StringIO。

4.1 读取pdf

    def read_pdf(self, file_name):
        rsrcmgr = PDFResourceManager()
        laparams = LAParams()
        fp = open(file_name, 'rb')
        parser = PDFParser(fp)
        document = PDFDocument(parser)
        device = PDFPageAggregator(rsrcmgr, laparams=laparams)
        interpreter = PDFPageInterpreter(rsrcmgr, device)

        pages = {}
        for page in PDFPage.create_pages(document):
            dictionary = {}
            dictionary['textbox'] = []
            dictionary['textline'] = []

            interpreter.process_page(page)
            layout = device.get_result()
            for item in layout:
                if isinstance(item, LTTextBox):
                    dictionary['textbox'].append(item)
                    for child in item:
                        if isinstance(child, LTTextLine):
                            dictionary['textline'].append(child)
            pages[layout.pageid] = dictionary
        return pages

4.2 pdf转文本

    def convert_pdf_to_txt(self, path, page_no=-1):
        rsrcmgr = PDFResourceManager()

        retstr = StringIO()
        laparams = LAParams()
        device = TextConverter(rsrcmgr, retstr, laparams=laparams)
        fp = open(path, 'rb')
        interpreter = PDFPageInterpreter(rsrcmgr, device)
        password = ""
        maxpages = 0
        caching = True
        pagenos = set()
        i = 0
        for page in PDFPage.get_pages(fp, pagenos, maxpages=maxpages, password=password, caching=caching,
                                      check_extractable=True):
            i += 1
            if i != page_no:
                continue
            interpreter.process_page(page)

        fp.close()
        device.close()
        str = retstr.getvalue()
        retstr.close()

        return str

4.3 pdf比较

    def compare_pdf(self, file1, file2, header_text, x_margin=10, compare_margin=0.2):
        rsrcmgr = PDFResourceManager()
        retstr = StringIO()
        laparams = LAParams()
        fp = open(file1, 'rb')
        parser = PDFParser(fp)
        document = PDFDocument(parser)
        device = PDFPageAggregator(rsrcmgr, laparams=laparams)

        out = StringIO()

        layoutmode = 'normal'
        scale = 1.3
        fontscale = 1

        html_coverter = HTMLPrivateConverter(rsrcmgr, out, scale=scale,
                                             layoutmode=layoutmode, laparams=laparams, fontscale=fontscale,
                                             imagewriter=None, header_text=header_text, x_margin=x_margin)
        interpreter = PDFPageInterpreter(rsrcmgr, device)
        testpages = PDFPage.create_pages(document)

        file_dict = self.read_pdf(file2)

        for page in testpages:
            interpreter.process_page(page)
            layout = device.get_result()

            html_coverter.page_begin(layout)
            if file_dict.get(layout.pageid) == None:
                break
            compare_page = file_dict[layout.pageid]

            for item in layout:
                if isinstance(item, LTTextBox):
                    html_coverter.begin_div('textbox', 1, item.x0 + html_coverter.x_margin, item.y1, item.width,
                                            item.height,
                                            item.get_writing_mode())

                    for child in item:
                        if isinstance(child, LTTextLine):
                            self.compare_textline(child, compare_page, html_coverter, compare_margin)
                            html_coverter.put_newline()
                    html_coverter.end_div()
            html_coverter.page_end()

        fp.close()
        device.close()
        retstr.close()

        return out.getvalue()

4.4 完整程序

点击：完整源码