Python readability.Document类代码示例

def run(index):
	print "Index %d" % index
	dirname = "data/%04d" % index

	# url of english article
	url = open(dirname + "/url_en.txt").read()

	# download html
	html = urllib.urlopen(url).read().decode('latin-1')

	# apply readability
	document = Document(html)
	article = document.summary()
	article = nltk.clean_html(article)

	# replace latin characters
	article = re.sub(u'
', u'\n', article)
	article = re.sub(u'\x92', u'`', article)
	article = re.sub(u'\x96', u'-', article)

	# article_en.txt
	output = codecs.open(dirname + "/article_en.txt", 'w', encoding='ascii', errors='ignore')
	output.write(article)
	output.close()

	# title.txt
	output = codecs.open(dirname + "/title.txt", 'w', encoding='ascii', errors='ignore')
	output.write(document.title())
	output.close()

def recommend_by_url(url):
    parsed = urlparse(url)
    doc = Document(requests.get(url).content)
    content = html.fromstring(doc.content()).xpath('string()')
    bigrams = make_bigrams(content)
    vec_bow = dictionary.doc2bow(bigrams)
    vec_lsi = lsi[vec_bow]
    sims = index[vec_lsi]
    #print sims
    docs = sorted(list(enumerate(sims)), key=lambda item: -item[1])
    results, seen = [], []
    for doc, score in docs:
        res = ARTICLES[doc]
        if not 'url' in res or res['url'] in seen:
            continue
        seen.append(res['url'])
        p = urlparse(res['url'])
        if p.hostname.endswith(parsed.hostname):
            continue
        res['score'] = float(score)
        if 'content' in res:
            del res['content']
        if 'html' in res:
            del res['html']
        if res['summary']:
            res['summary'] = res['summary'].strip()
        results.append(res)
        if len(results) > 14:
            break
    return results

def markdownify(url_list, **options):
    articles = []
    images = []
    paragraph_links = options['paragraph_links']
    wrap_text = options['wrap_text']
    preamble = options['preamble']
    for url in url_list:
        req = urllib2.Request(url,None,{'Referer': url_list[0]})
        html = urllib2.urlopen(req).read()
        document = Document(html, url=url)
        readable_title = document.short_title()
        summary = document.summary()
        summary_doc = build_doc(summary)
        images.extend([a.get('src') for a in summary_doc.findall('.//img')])
        articles.append(document.summary())

    markdown_articles = []
    for (article, url) in zip(articles, url_list):
        h = html2text.HTML2Text(baseurl=url)
        h.inline_links = False
        h.links_each_paragraph = (paragraph_links and 1) or 0
        h.body_width = (wrap_text and 78) or 0
        markdown_articles.append(h.handle(article))
    combined_article = u"\n\n----\n\n".join(markdown_articles)
    if preamble:
        combined_article = (u"Title:        %s  \nOriginal URL: %s\n\n" % (readable_title, url_list[0])) + combined_article
    return combined_article.encode("utf-8")

def getText():
    dataList = []
    for f in os.listdir('unsupervised\\documents'):
        filePath = 'unsupervised\\documents\\' + f
        #print filePath
        fileName, fileExtension = os.path.splitext(filePath)
        #print fileExtension
        if fileExtension.lower() == '.docx':
            print '' #'its a {0} {1}{2}'.format('word document', fileName, fileExtension)
            doc = docxDocument(filePath)
            for p in doc.paragraphs:
                dataList.append(p.text)     #print p.text
            #print "-------------------------------"
        elif fileExtension.lower() == '.pdf':
            print '' #'its a {0} {1}{2}'.format('pdf document', fileName, fileExtension)
            #TODO
        elif ((fileExtension.lower() == '.html') or (fileExtension.lower() == '.htm')):
            print '' #'its a {0} {1}{2}'.format('html file', fileName, fileExtension)
            with codecs.open (filePath, errors='ignore') as myfile:
                source = myfile.read()
                article = Document(source).summary()
                title = Document(source).title()
                soup = BeautifulSoup(article, 'lxml')
                final = replaceTwoOrMore((title.replace('\n', ' ').replace('\r', '') + '.' + soup.text.replace('\n', ' ').replace('\r', '')))
                dataList.append(final)
                #print '*** TITLE *** \n\"' + title + '\"\n'
                #print '*** CONTENT *** \n\"' + soup.text + '[...]\"'
        else:
            print '' # 'undectected document type'
            print '' #"-------------------------------"
    return dataList

def extract_article(url):
  r = requests.get(url)
  
  # the the url exists, continue
  if r.status_code == 200:
    
    # extract and parse response url
    url = parse_url(r.url)

    # extract html
    html = r.content.decode('utf-8', errors='ignore')

    # run boilerpipe
    # boilerpipe_extractor = Extractor(html=html)

    # run readability
    readability_extractor = Document(html)

    html = readability_extractor.summary()
    # return article data
    return {
      'title': readability_extractor.short_title(),
      'html': html,
      'content': strip_tags(html).encode('utf-8', errors='ignore'),
      'url': url
    }

  # otherwise return an empty dict
  else:
    return {}

def get_webpage_by_html(url, html=None):
    html = get_html_str(url, html)
    summary_obj = predefined_site(url, html)
    article = video_site(url)
    if summary_obj is None:
        doc = Document(html, url=url, debug=True, multipage=False)
        summary_obj = doc.summary_with_metadata(enclose_with_html_tag=False)
    title = summary_obj.short_title
    if article is None:
        article = summary_obj.html
    from urllib.parse import urlparse
    webpage = Webpage()
    webpage.url = url
    webpage.domain = urlparse(url).hostname
    webpage.title = title
    webpage.favicon = ""
    webpage.top_image = None
    webpage.excerpt = summary_obj.description
    webpage.author = None
    webpage.content = article
    webpage.tags = get_suggest_tags(title, article, summary_obj.keywords)
    webpage.movies = []
    webpage.raw_html = html
    webpage.publish_date = None
    webpage.segmentation = get_segmentation(title, article)
    return webpage.__dict__

def extract_article(url):
  r = requests.get(url)
  
  # the the url exists, continue
  if r.status_code == 200:
    
    # extract and parse response url
    url = parse_url(r.url)

    # extract html
    html = r.content.decode('utf-8', errors='ignore')

    # run boilerpipe
    BP = Extractor(html=html)

    # run readability
    Rdb = Document(html)

    html = Rdb.summary()
    # return article data
    return {
      'extracted_title': Rdb.short_title().strip(),
      'extracted_content': strip_tags(BP.getText()),
    }

  # otherwise return an empty dict
  else:
    return {}

class Gist:

    keyword_pattern = re.compile(r'^[^\d]+$')
    stop_words = set(get_stop_words('en'))

    def __init__(self, html):
        self.html = html
        self.document = Document(html)

    @property
    def title(self):
        return self.document.short_title()

    @cached_property
    def text(self):
        text = self.document.summary()
        text = re.sub('<br[^>]+>', '\n', text)
        text = re.sub('</?p[^>]+>', '\n\n', text)
        text = re.sub('<[^>]+>', '', text)
        text = re.sub('^[ \t]+$', '', text)
        text = re.sub('\n{3,}', '\n\n', text, flags=re.MULTILINE)
        return text

    @staticmethod
    def _common_prefix(one, two):
        parallelity = [x == y for x, y in zip(one, two)] + [False]
        return parallelity.index(False)

    @classmethod
    def _find_representative(cls, stem, text):
        tokens = text.split()
        prefixes = {token: cls._common_prefix(token, stem) for token in tokens}
        best = lambda token: (-token[1], len(token[0]))
        return sorted(prefixes.items(), key=best)[0][0]

    @classmethod
    def _is_good_keyword(cls, word):
        return (word not in cls.stop_words) and \
                cls.keyword_pattern.match(word)

    @classmethod
    def find_keywords(cls, text):
        whoosh_backend = SearchForm().searchqueryset.query.backend
        if not whoosh_backend.setup_complete:
            whoosh_backend.setup()
        with whoosh_backend.index.searcher() as searcher:
            keywords = searcher.key_terms_from_text(
                'text', text, numterms=10, normalize=False)
        keywords = list(zip(*keywords))[0] if keywords else []
        keywords = [cls._find_representative(keyword, text) for keyword in keywords]
        keywords = [keyword for keyword in keywords if cls._is_good_keyword(keyword)]
        #no double keywords in list
        keywords = list(set(keywords))
        #no punctuation in suggested keywords
        keywords = [''.join(c for c in s if c not in string.punctuation) for s in keywords]
        return keywords

    @property
    def keywords(self):
        return self.find_keywords(self.text)

    async def enrich(self, result):
        if not self.soup:
            return result

        result.set('title', self.soup.title.string, 0, 'textlength')

        if result.has('content'):
            return result

        parts = []
        for txt in self.soup.find_all("noscript"):
            if txt.string is not None:
                parts.append(txt.string)
        html = " ".join(parts).strip()
        if not html:
            html = self.soup.all_text()

        try:
            doc = Document(html, url=self.url)
            content = doc.summary(html_partial=True)
            result.set('content', sanitize_html(content))
        # pylint: disable=bare-except
        except:
            pass

        return result

class Article:

    def __init__(self, url):
        print('Saving page: {}'.format(url))
        res = requests.get(url)
        self.url = url
        self.article = Document(res.content)
        self._add_title()
        self._save_images()

    def _add_title(self):
        self.root = etree.fromstring(self.article.summary())
        body = self.root.find('body')

        title = self.article.title()
        ascii_title = unidecode(title) if type(title) == unicode else title

        title_header = etree.HTML('<h2>{}</h2>'.format(ascii_title))
        body.insert(0, title_header)

    def _save_images(self):
        tmppath = tempfile.mkdtemp()
        images = self.root.xpath('//img')
        for img in images:
            imgsrc = img.get('src')

            # handle scheme-agnostic URLs
            if 'http' not in imgsrc and '//' in imgsrc:
                imgsrc = 'http:{}'.format(imgsrc)

            # handle relative file paths
            elif 'http' not in imgsrc:
                parsed = urlparse(self.url)
                imgsrc = '{}://{}{}'.format(parsed.scheme, parsed.netloc, imgsrc)

            filename = os.path.basename(imgsrc)
            dest = os.path.join(tmppath, filename)

            try:
                res = requests.get(imgsrc)
            except Exception as e:
                print('Could not fetch image ({}) from "{}"'.format(str(e), imgsrc))
                return

            if res.status_code == 404:
                print('Could not fetch image (HTTP 404), attempted fetch: "{}", source URL: {}'.format(imgsrc, img.get('src')))
                continue

            with open(dest, 'wb') as f:
                f.write(res.content)

            img.set('src', dest)

    @property
    def title(self):
        return self.article.title()

    @property
    def html(self):
        return etree.tostring(self.root)

def get_announcement_body(url):

        now = datetime.datetime.now()
        resp = ["","","","","",""]
        images = []
        html = br.open(url).read()

        readable_announcement = Document(html).summary()
        readable_title = Document(html).title()
        soup = BeautifulSoup(readable_announcement, "lxml")
        final_announcement = soup.text
        links = soup.findAll('img', src=True)
        for lin in links:
                li = urlparse.urljoin(url,lin['src'])
                images.append( li)
                
        resp[0] = str(final_announcement.encode("ascii","ignore"))
        resp[1] = str(readable_title.encode("ascii","ignore"))
        resp[2] = str(now.month)+" "+str(now.day)+" "+str(now.year)+"-"+str(now.hour)+":"+str(now.minute)+":"+str(now.second)
        resp[3] = url
        resp[4] = url
        resp[5] = ""
        #insertDB(resp)
        #print "inserted resp"
                 
        title_article = []
        title_article.append(final_announcement)
        title_article.append(readable_title)
        title_article.append(images)                
        return title_article

    def getTextFromHTML(self, url_id):
        """ Runs Readability (Document) on the HTML text
        """
        html_row = get_html(self.pg_conn, url_id)

        if not html_row or 'html' not in html_row:
            return False

        if html_row['readabletext'] and html_row['readabletext'] != '':
            return html_row['readabletext']

        html = html_row['html']

        try:
            html_summary = Document(html).summary(html_partial=True)
            html_summary = html_summary.replace('\n','').replace('\t','')

            if len(html_summary) < 150 or "Something's wrong here..." in html_summary or "<h1>Not Found</h1><p>The requested URL" in html_summary or html_summary == "<html><head/></html>" or "403 Forbidden" in html_summary:
                return False

            raw_text = lxml.html.document_fromstring(html_summary).text_content()
        except:
            raw_text = False

        if raw_text:
            save_readabletext(self.pg_conn, url_id, raw_text, 'meta')
        else:
            save_readabletext(self.pg_conn, url_id, '', 'meta')

        return raw_text

def main():
    #print 'Hello there'
    # Command line args are in sys.argv[1], sys.argv[2] ...
    # sys.argv[0] is the script name itself and can be ignored

    dataList = []

    for f in os.listdir('documents'):
        filePath = 'documents\\' + f
        #print filePath
        fileName, fileExtension = os.path.splitext(filePath)
        #print fileExtension
        if fileExtension.lower() == '.docx':
            print '' #'its a {0} {1}{2}'.format('word document', fileName, fileExtension)
            doc = docxDocument(filePath)
            for p in doc.paragraphs:
                dataList.append(p.text)     #print p.text
            #print "-------------------------------"
        elif fileExtension.lower() == '.pdf':
            print '' #'its a {0} {1}{2}'.format('pdf document', fileName, fileExtension)
            # with open(filePath) as f:
            #     doc = slate.PDF(f)
            #     print doc[1]
            #     exit()


            #TODO
        elif ((fileExtension.lower() == '.html') or (fileExtension.lower() == '.htm')):
            print '' #'its a {0} {1}{2}'.format('html file', fileName, fileExtension)
            with codecs.open (filePath, errors='ignore') as myfile:
                source = myfile.read()
                article = Document(source).summary()
                title = Document(source).title()
                soup = BeautifulSoup(article, 'lxml')
                final = replaceTwoOrMore((title.replace('\n', ' ').replace('\r', '') + '.' + soup.text.replace('\n', ' ').replace('\r', '')))
                dataList.append(final)
                #print '*** TITLE *** \n\"' + title + '\"\n'
                #print '*** CONTENT *** \n\"' + soup.text + '[...]\"'
        else:
            print '' # 'undectected document type'
            print '' #"-------------------------------"

    #print dataList
    #for i in dataList:
    #    print i
    cachedStopWords = stopwords.words("english")
    combined = ' '.join(dataList)

    #print combined
    bloblist = [tb(combined)]

    for i, blob in enumerate(bloblist):
        print("Top words in document {}".format(i + 1))
        scores = {word: tfidf(word, blob, bloblist) for word in blob.words if word not in nltk.corpus.stopwords.words('english')}
        #print scores
        sorted_words = sorted(scores.items(), key=lambda x: x[1], reverse=True)
        #print sorted_words
        for word, score in sorted_words:
            print("\tWord: {}, TF-IDF: {}".format(word, round(score, 5)))

 def _getResponseText(self, response):
     '''
     (reponse) -> Text
     Returns text within the body of an HttpResponse object.
     '''
     readability = Document(response.body)
     content = readability.title() + readability.summary()
     return content

def main():
    html = urllib.urlopen("http://habrahabr.ru/post/150756/").read()
    doc = Document(html)
    short_title = doc.short_title()
    readable_article = doc.summary()
    f = open("C:\\users\\mykola\\documents\\%s.html" % short_title, "wb")
    f.write(readable_article.encode("utf-8"))
    f.close()