def mutate_facet_tokens(book):
facet_fields = [
["authors"],
["publishers"],
["subjects"],
["subject", "subjects"],
["source"],
["language"],
["languages"],
["language_code"],
["has_fulltext"],
["facet_year"],
]
all_tokens = []
for ff in facet_fields:
fieldname = ff[0]
label = fieldname if len(ff) == 1 else ff[1]
# if 'sub' in label: pdb.set_trace()
vs = book.get(fieldname)
if type(vs) != list:
vs = [vs]
all_tokens.extend(facet_token(label, v) for v in vs if v)
all_tokens = sorted(set(all_tokens))
if all_tokens:
book["facet_tokens"] = " ".join(all_tokens)
def get_facet_tokens(buf, dd):
# for each facet field:
# generate facet tokens
# insert tokens into buf
global token_counts
# should get facet fields from schema, not put them here. @@
facet_fields = ('authors',
'publisher',
'subject',
('subjects', 'subject'),
'source',
'language',
'language_code',
'has_fulltext',
'facet_year',
)
tokens = []
for xfield in facet_fields:
if type(xfield) == tuple:
field, flabel = xfield
else:
field = flabel = xfield
for idx,v in dd.get(field, []):
assert buf[idx][0] == field
token_counts[field] += 1
tokens.append(facet_token(flabel, v))
return tokens
def exact_facet_count(self, query, facet_name, facet_value):
ftoken = facet_token(facet_name, facet_value)
result_json = self.raw_search(
'%s facet_tokens:%s'% (query, ftoken),
rows=0,
wt='json')
result = simplejson.loads(result_json)
n = result['response']['numFound']
return n
def Xfacet_token_inverse(self,
token,
facet_list = default_facet_list):
# for now, just pull this straight from the SE
# need to add an LRU cache for performance. @@
if not re.match('^[a-z]+$', token):
raise SolrError, 'invalid facet token'
m = simplejson.loads(self.raw_search('facet_tokens:%s'% token,
rows=1, wt='json'))
facet_set = set(facet_list)
for d in m['response']['docs']:
for k,vx in d.iteritems():
kfs = k in facet_set
# if not kfs: continue
vvx = {str:(vx,), list:vx}.get(type(vx),())
for v in map(unicode, vvx):
ft = facet_token(k,v)
if facet_token(k,v) == token:
return (k,v)
return None
def Xfacet_token_inverse(self, token, facet_list=default_facet_list):
# for now, just pull this straight from the SE
# need to add an LRU cache for performance. @@
if not re.match('^[a-z]+$', token):
raise SolrError, 'invalid facet token'
m = simplejson.loads(
self.raw_search('facet_tokens:%s' % token, rows=1, wt='json'))
facet_set = set(facet_list)
for d in m['response']['docs']:
for k, vx in d.iteritems():
kfs = k in facet_set
# if not kfs: continue
vvx = {str: (vx, ), list: vx}.get(type(vx), ())
for v in map(unicode, vvx):
if facet_token(k, v) == token:
return (k, v)
return None
def exact_facet_count(self, query, selected_facets, facet_name,
facet_value):
ftoken = facet_token(facet_name, facet_value)
# this function is temporarily broken because of how facets are handled
# under dismax. @@
# well, ok, the use of dismax is temporarily backed out, but leave
# this signal here to verify that we're not actually using exact
# counts right now.
raise NotImplementedError
sf = list(s for s in selected_facets if re.match('^[a-z]{12}$', s))
fs = ' '.join(sf + [ftoken])
result_json = self.raw_search(self.basic_query(query),
fq='facet_tokens:(%s)' % fs,
rows=0,
wt='json')
result = simplejson.loads(result_json)
n = result['response']['numFound']
return n
def exact_facet_count(self, query, selected_facets,
facet_name, facet_value):
ftoken = facet_token(facet_name, facet_value)
# this function is temporarily broken because of how facets are handled
# under dismax. @@
# well, ok, the use of dismax is temporarily backed out, but leave
# this signal here to verify that we're not actually using exact
# counts right now.
raise NotImplementedError
sf = list(s for s in selected_facets if re.match('^[a-z]{12}$', s))
fs = ' '.join(sf+[ftoken])
result_json = self.raw_search(
self.basic_query(query),
fq='facet_tokens:(%s)'% fs,
rows=0,
wt='json')
result = simplejson.loads(result_json)
n = result['response']['numFound']
return n
