from collections import defaultdict # python 2.5
from operator import itemgetter
snd = itemgetter(1) # snd(t) is 2nd element of tuple t
def facet_counts(result_list, facet_fields):
"""Return list of facet counts for a search result set.
The list of field names to fact on is `facet_fields'.
The result list from solr is `result_list'. The structures
look like:
result_list = [ { fieldname1 : [values...] }, ... ]
facet_fields = ('author', 'media_type', ...)
>>> fnames = ('author', 'topic', 'format')
"""
facets = defaultdict(lambda: defaultdict(int))
for r in result_list:
for k in set(r.keys()) & set(facet_fields):
facets_k = facets[k] # move lookup out of loop for speed
for x in r[k]:
facets_k[x] += 1
return filter(snd, ((f, sorted(facets[f].items(),
key=snd,
reverse=True))
for f in facet_fields))
fnames = ('author', 'topic', 'format')
results = [{'title': ['Julius Caesar'],
'author': ['William Shakespeare'],
'format': ['folio'] },
{'title': ['Richard III'],
'author': ['William Shakespeare'],
'format': ['folio'] },
{'title': ['Tom Sawyer'],
'author': ['Mark Twain'],
'format': ['paperback'] },
{'title': ['The Space Merchants'],
'author': ['Frederik Pohl', 'C. M. Kornbluth'],
'format': ['paperback'] },
]
import urllib
from time import time
def query(q=None, max_rows=2000, facets_to_show=5):
global h,h0,h1,docs,fc
timings = []
# record a timestamped message in timings[]
def a(m=''): timings.append((time(), m))
q = q or raw_input('enter query: ')
url='http://127.0.0.1:8993/solr/select?' + \
urllib.urlencode({'q':q, 'rows':0, 'wt':'python'})
a('get #hits')
h = urllib.urlopen(url).read()
a('eval 1st response, %d bytes'% len(h))
h0 = eval(h)
n = h0['response']['numFound']
a('get n=%d rows'% n)
if n > max_rows:
a('limit retrieval to %d rows'% max_rows)
n = max_rows
url='http://127.0.0.1:8993/solr/select?' + \
urllib.urlencode({'q':q, 'rows':n, 'wt':'python'})
h = urllib.urlopen(url).read()
a('eval full response %d bytes'% len(h))
h1 = eval(h)
a('extract doc list')
docs = h1['response']['docs']
a('got %d docs'% len(docs))
fc = facet_counts(docs, ('authors','subject','language'))
a('got facet counts')
for fname,facets in fc:
m = min(len(facets), facets_to_show)
print 'top %d (of %d) "%s" facets'% (m, len(facets), fname)
print '','\n '.join(repr(t) for t in facets[:facets_to_show]),'\n'
a('done')
def deltas(timings):
for (a1,b1),(a2,b2) in zip(timings[:-1], timings[1:]):
print ' (%s)->(%s): %.3e sec'% (b1,b2,a2-a1)
print ' total: %.3e sec'% (timings[-1][0] - timings[0][0])
print 'speed:'
deltas(timings)
query()
