def row_to_index(self, row):
key = row.get(self.entity_key)
q = {self.alias_key: key}
aliases = self.alias.traverse(**q)
aliases = map(lambda a: a.get(self.alias_text), aliases)
row['alias'] = aliases
row['title.n'] = normalize(row.get('title'))
row['alias.n'] = map(normalize, aliases)
row['__type__'] = self.name
row['__key__'] = key
row['__id__'] = self.name + ':' + key
sys.stdout.write('.')
sys.stdout.flush()
return row
def row_to_index(self, row):
key = row.get(self.entity_key)
q = {self.alias_key: key}
aliases = self.alias.traverse(**q)
aliases = map(lambda a: a.get(self.alias_text), aliases)
row['alias'] = aliases
row['title.n'] = normalize(row.get('title'))
row['alias.n'] = map(normalize, aliases)
row['__type__'] = self.name
row['__key__'] = key
row['__id__'] = self.name + ':' + key
sys.stdout.write('.')
sys.stdout.flush()
return row
def query(solr, q, filters=(), **kw):
fq = ['+' + field(k, v) for k, v in filters]
#fq.append('_collection:%s' % entities.name)
if len(q) and q != '*:*':
nq = normalize(q)
_q = [
field('title', q, boost=10),
field('title.n', nq, boost=7),
field('alias', q, boost=8),
field('alias.n', nq, boost=5),
field('text', q, boost=2),
field('text', nq)
]
q = ' OR '.join(_q)
result = solr.raw_query(q=q, fq=fq, wt='json',
sort='score desc, title desc', fl='*,score', **kw)
result = json.loads(result)
return result
def find_fuzzy(cls, text, filters=(), facet_type=False, **kw):
fq = ['+' + query_filter(k, v) for k, v in filters]
if len(text) and text != '*:*':
ntext = normalize(text)
_q = [
query_filter('title', text, boost=10),
query_filter('title.n', ntext, boost=7),
query_filter('alias', text, boost=8),
query_filter('alias.n', ntext, boost=5),
query_filter('text', text, boost=2),
query_filter('text', ntext)
]
for token in tokenize(ntext):
_q.append(query_filter('title', token, fuzzy=True, boost=4))
_q.append(query_filter('alias', token, fuzzy=True, boost=3))
text = ' OR '.join(_q)
if facet_type:
kw['facet'] = 'true'
kw['facet.field'] = kw.get('facet.field', []) + ['__type__']
kw['facet.limit'] = 50
return cls.find(text, fq, **kw)
def find_fuzzy(cls, text, filters=(), facet_type=False, **kw):
fq = ['+' + query_filter(k, v) for k, v in filters]
if len(text) and text != '*:*':
ntext = normalize(text)
_q = [
query_filter('title', text, boost=10),
query_filter('title.n', ntext, boost=7),
query_filter('alias', text, boost=8),
query_filter('alias.n', ntext, boost=5),
query_filter('text', text, boost=2),
query_filter('text', ntext)
]
for token in tokenize(ntext):
_q.append(query_filter('title', token, fuzzy=True, boost=4))
_q.append(query_filter('alias', token, fuzzy=True, boost=3))
text = ' OR '.join(_q)
if facet_type:
kw['facet'] = 'true'
kw['facet.field'] = kw.get('facet.field', []) + ['__type__']
kw['facet.limit'] = 50
return cls.find(text, fq, **kw)