proceedings_xml.py 文件源码

def get_language(self, s_intervention, p, i_lang, new_paragraphs):
        language = p.xpath('.//span[@class="italic"][text()[re:test(.,"^[\xad\s\.—–\-?,\(]*({})[\xad\s\.—–\-?,\)]*")]]'.format('|'.join(self.langs)), namespaces=self.ns)
        if len(language) > 0 and not self.explanations_of_vote.match(language[0].text):
            lang = re.match(
                r'.*({}).*'.format('|'.join(self.langs)),
                language[0].text)
            output = lang.group(1)
            for l in language:
                l.drop_tree()
        else:
            p = html.tostring(p, with_tail=True, encoding='utf-8').decode('utf-8')
            lang_in_text = re.search(
                r'\(({})\)'.format('|'.join(self.langs)),
                p)
            if lang_in_text is not None:
                output = lang_in_text.group(1)
                p = re.sub(r'\(({})\) *'.format('|'.join(self.langs)), r'', p)
            else:
                if len(new_paragraphs) == 0:
                    if 'role' in s_intervention.keys():
                        president_pattern = '|'.join(self.loc['president'])
                        if re.match(r'{}\Z'.format(president_pattern), s_intervention['role']):
                                output = 'unknown'
                        else:
                            if i_lang is None:
                                output = self.language.upper()
                            else:
                                output = i_lang
                    else:
                        if i_lang is None:
                            output = self.language.upper()
                        else:
                            output = i_lang
                else:
                    output = new_paragraphs[-1]['language']
            p = html.fromstring(p)
        return output, p