def text_only_search(text):
"""
performs a search and ranks results based on the text given (splits into individual words)
"""
search_words = text.split()
#remove duplicates
search_words = list(set(search_words))
results = []
ed = EditDistance(Keywords, 'keyword')
for word in search_words:
if word not in STOP_WORDS:
ed_list, ed_diff = ed.correct(word)
for res in ed_list:
cur_res = Keywords.objects.extra(select={'diff':"%s"}, select_params=[ed_diff]).filter(keyword__exact=res).order_by('diff','-frequency')
results = list(chain(results, cur_res))
# now we need to rank the results for images based on most exact matches
ranked_res = rank_results(results)
return ranked_res
def text_only_search(text):
"""
performs a search and ranks results based on the text given (splits into individual words)
"""
search_words = text.split()
#remove duplicates
search_words = list(set(search_words))
results = []
ed = EditDistance(Keywords, 'keyword')
for word in search_words:
if word not in STOP_WORDS:
#exact and substring keyword matches incase-sensitive
#####
#cur_res = Keywords.objects.extra(select={'diff':"length(keyword)-length(%s)"}, select_params=[word]).filter(keyword__contains=word).order_by('diff','-frequency')
#results = list(chain(results, cur_res))
#print cur_res
####
ed_list, ed_diff = ed.correct(word)
for res in ed_list:
#### Advanced search features for later
#cur_res = Keywords.objects.extra(select={'diff':"length(keyword)-length(%s)+%d"}, select_params=[res, diff]).filter(keyword__contains=word)order_by('diff','-frequency')
####
cur_res = Keywords.objects.extra(select={'diff':"%s"}, select_params=[ed_diff]).filter(keyword__exact=res).order_by('diff','-frequency')
results = list(chain(results, cur_res))
print cur_res
#for res in cur_res:
# for kw in res:
# print "Keyword %s has frequency %d and diff %d" % (res.keyword, res.frequency, res.diff)
print results
#now we need to rank the results for images based on most exact matches
ranked_res = rank_results(results)
return ranked_res
