从其位置获取注释文本PDFMiner

问题描述

我想从其位置提取注释文本（例如，超链接的突出显示的文本）。为此，我可以使用PDFminer抓取位置和网址，如下面的代码所示。是否可以将此位置传递给布局对象并获取文本？

这是我用于此目的的代码块。

第一部分包括一个名为parse_annotation的函数，用于解析每个页面中的注释。

def parse_annotations(page):
    positions = []
    urls = []
    for annot in pdftypes.resolve1(page.annots):
        if isinstance(annot,pdftypes.PDFObjRef):
            annotationDict = annot.resolve()
            # Skip over any annotations that are not links
            if str(annotationDict["Subtype"]) != "/'Link'":
                continue
            destID = 0
            position = annotationDict["Rect"]
            uriDict = "None"
            if any(k in annotationDict for k in {"Dest","D"}):                
                destID = (annotationDict["Dest"][0]).objid                
                url = "Cross reference"
            elif "A" in annotationDict:
                # Key A contains PDFObjRef,then resolve it again
                if isinstance(annotationDict["A"],pdftypes.PDFObjRef):
                    uriDict = pdftypes.resolve1(annotationDict["A"])
                    if any(k in uriDict for k in {"Dest","D"}): 
                        destID = (uriDict["D"][0]).objid
                else:
                    uriDict = annotationDict["A"]
                # Check if the key exists within resolved uriDict
                if str(uriDict["S"]) == "/'GoTo'":
                    url = "Cross reference"
                elif str(uriDict["S"]) == "/'URI'":
                    url = str(uriDict["URI"])
                    url = url.lstrip("b")
                    url = url.replace("'","")
                else:
                    # Skip if key S in uriDict does not contain value URI,GoTo
                    continue
            else:
                sys.stderr.write("Warning: unknown key in annotationDict : ",annotationDict)
            #print(annot,'\n',annotationDict,destID,position,uriDict,url,'\n')
            print(position,'\n')
            positions.append(position)
            urls.append(url)
        else:
            sys.stderr.write("Warning: unknown annotation: %s\n" % annot)
    return positions,urls

示例PDF文件可从下面的链接中找到。

https://www2.ed.gov/about/offices/list/ocr/docs/20200512-qa-psi-covid-19.pdf

现在，通过使用PDFMiner创建一个文档对象，并开始循环浏览PDF中找到的页面。

manager = PDFResourceManager()
output = StringIO()
codec = 'utf-8'
laparams = LAParams()
converter = TextConverter(manager,output,codec=codec,laparams=laparams)
device = PDFPageAggregator(manager,laparams=laparams)
interpreter = PDFPageInterpreter(manager,device)
page_interpreter = PDFPageInterpreter(manager,converter)
filename = '20200512-qa-psi-covid-19.pdf'
fp = open(filename,'rb')
parser = PDFParser(fp)
document = PDFDocument(parser)
if not document.is_extractable:
    raise PDFTextExtractionNotAllowed
page_no = 0
for pageNumber,page in enumerate(PDFPage.create_pages(document)):
    print("\n================ PageNumber ",pageNumber+1,"===================\n")
    if pageNumber == page_no:
        page_interpreter.process_page(page)
        raw_text = output.getvalue()
        output.truncate(0)
        output.seek(0)
    interpreter.process_page(page)
    layout = device.get_result()
    if page.annots:
        positions,urls = parse_annotations(page)
        for obj in layout:
            print('Object name and position %s \t %s \n' % (obj.__class__.__name__,obj.bbox))
    page_no += 1
fp.close()
converter.close()
output.close()
device.close()

预先感谢， 答：

解决方法

暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！

如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。

小编邮箱:dio#foxmail.com (将#修改为@）