def get_page_summary(self): parser = HtmlParser(self.src_content, Tokenizer(self.LANGUAGE), self.base_url) stemmer = Stemmer(self.LANGUAGE) summarizer = Summarizer(stemmer) summarizer.stop_words = get_stop_words(self.LANGUAGE) summary = "" try: for sentence in summarizer(parser.document, 10): summary = summary + " " + str(sentence) return summary except Exception: return ""
def summarize(self, method='luhn'): """ Summarize text """ method = self._check_method(method) if self.url: parser = HtmlParser.from_url(self.url, Tokenizer(self.LANGUAGE)) elif self.html: parser = HtmlParser(self.html, Tokenizer(self.LANGUAGE)) stemmer = Stemmer(self.LANGUAGE) summarizer = method(stemmer) summarizer.stop_words = get_stop_words(self.LANGUAGE) sumy = summarizer(parser.document, self.SENTENCES_COUNT) summary = ''.join([str(i) for i in list(sumy)]) return summary
def summarize(article: Article): parser = HtmlParser(article.html, Tokenizer(LANGUAGE)) stemmer = Stemmer(LANGUAGE) summarizer = Summarizer(stemmer) summarizer.stop_words = get_stop_words(LANGUAGE) return summarizer(parser.document, SENTENCES_COUNT)