def parse_related_documents(self, response):
xpath = '/html/body/div/table/tr[3]/td/table/tr/td/table/tr'
hxs = HtmlXPathSelector(response).select(xpath)
act = Loader(self, response, LegalAct(), hxs, required=('_id', ))
act.add_xpath('_id', 'td[2]/b/text()')
if not act.get_output_value('_id'):
p_id = unicode(self._get_query_attr(response.url, 'p_id'))
act.replace_value('_id', u'NONUMBER-%s' % p_id)
relations = defaultdict(list)
xpath = '/html/body/div/table/tr[3]/td/table/tr/td/align/table/tr'
for row in HtmlXPathSelector(response).select(xpath):
docid = get_all(row, 'td[4]/span//text()')
rel_type = row.select('td[6]/span/text()')
if rel_type:
rel_type = rel_type.extract()[0].strip().lower()
if rel_type in (u'pakeistas dokumentas',
u'ankstesnė dokumento redakcija'):
relations['amends'].append(docid)
elif rel_type == u'priimtas dokumentas':
relations['adopts'].append(docid)
elif rel_type == u'ryšys su taikymą nusakančiu dokumentu':
relations['defines_applicability'].append(docid)
elif rel_type == u'ryšys su galiojimą nusakančiu dokumentu':
relations['defines_validity'].append(docid)
elif rel_type == u'negalioja de jure':
relations['defines_as_no_longer_valid'].append(docid)
elif rel_type == u'kitas projekto variantas':
relations['new_draft_version'].append(docid)
elif rel_type == u'kitas projekto variantas':
relations['new_draft_version'].append(docid)
elif rel_type == u'ryšys su ratifikavimo dokumentu':
relations['ratification'].append(docid)
if relations:
act.add_value('relations', dict(relations))
yield act.load_item()
def _parse_law_act(self, response, hxs, base=False):
"""
Extracts basic document information and returns law act loader.
Parameters:
base
Return only base information about document. This is used, when
filling some information bits to a law act from several law act
documents.
"""
lang = hxs.select('tr[1]/td[4]/b/text()').extract()[0].strip().lower()
if lang not in (u'lietuvių', u'rusų', u'anglų', u'ispanų'):
self.error(response, 'Unknown language: %s' % lang)
if lang != u'lietuvių':
return None
act = Loader(self, response, LegalAct(), hxs, required=REQUIRED_FIELDS)
act.add_xpath('_id', 'tr[1]/td[2]/b/text()')
source = self._get_source(response.url, 'p_id')
if not act.get_output_value('_id'):
act.replace_value('_id', u'NONUMBER-%s' % source['id'])
if base:
return act
act.add_xpath('name', 'caption/text()')
act.add_xpath('kind', 'tr[1]/td[1]/b/text()')
act.add_xpath('number', 'tr[1]/td[2]/b/text()')
act.add_xpath('date', 'tr[1]/td[3]/b/text()')
act.add_value('source', source)
self._fix_name_case(act)
return act