def extract_fields_from_html(self, data):
root = lxml.html.fromstring(data, parser=self.htmlparser)
text = root.find('body/div')
if text is None:
raise ValueError("Structure of html is not body > div")
metaarea = root.cssselect('#metaarea')
author = []
if metaarea:
author = ' '.join(
unique(e.text_content()
for e in metaarea[0].cssselect('.author')))
metaarea[0].drop_tree()
for section in root.iter('section'):
for sib in section.itersiblings():
if sib.tag == 'section':
break
sib.drop_tree()
for p in root.iter('p'):
p.tail = '\n\n' + (p.tail or '')
for e in root.cssselect('.add'):
e.drop_tree()
hgroup = text.cssselect('.hgroup')[0]
division = hgroup[0]
title = hgroup[-1]
if title == division:
division = None
others = hgroup[1:-1]
hgroup.drop_tree()
content = self.fix_text(text.text_content())
if title is not None:
title = self.fix_text(title.text_content())
else:
title = ''
if division is not None:
division = self.fix_text(division.text_content())
else:
division = ''
return {
'content': content,
'author': author,
'heading': {
'title': title,
'division': division,
'subhead': [e.text_content().strip() for e in others]
},
'boost': self.boost_factor(content)
}
def extract_fields_from_html(self, data):
root = lxml.html.fromstring(data, parser=self.htmlparser)
text = root.find('body/div')
if text is None:
raise ValueError("Structure of html is not body > div")
metaarea = root.cssselect('#metaarea')
author = []
if metaarea:
author = ' '.join(unique(e.text_content() for e in metaarea[0].cssselect('.author')))
metaarea[0].drop_tree()
for section in root.iter('section'):
for sib in section.itersiblings():
if sib.tag == 'section':
break
sib.drop_tree()
for p in root.iter('p'):
p.tail = '\n\n' + (p.tail or '')
for e in root.cssselect('.add'):
e.drop_tree()
hgroup = text.cssselect('.hgroup')[0]
division = hgroup[0]
title = hgroup[-1]
if title == division:
division = None
others = hgroup[1:-1]
hgroup.drop_tree()
content = self.fix_text(text.text_content())
if title is not None:
title = self.fix_text(title.text_content())
else:
title = ''
if division is not None:
division = self.fix_text(division.text_content())
else:
division = ''
return {
'content': content,
'author': author,
'heading': {
'title': title,
'division': division,
'subhead': [e.text_content().strip() for e in others]
},
'boost': self.boost_factor(content)
}
def suggest_pi_dict(word):
_cache = cache["pi"]
if word not in _cache:
body = {
"query": {
"term": {
"term.folded": asciify(word)
}
}
}
is_title = word.istitle()
r = es.search('en-dict', 'definition', body)
matching_terms = [hit['_source']['term'] for hit in r['hits']['hits']]
if is_title:
matching_terms = (term.title() for term in matching_terms)
matching_terms = (term for term in matching_terms if ' ' not in term)
matching_terms = unique(matching_terms)
result = [(term, 1) for term in matching_terms]
_cache[word] = result
return _cache[word]
