def on_search(self, request):
if request.method == "GET":
return self.render_template('search.html')
elif request.method == "POST":
search_string = request.form.get("keywords", None)
if search_string:
search_string_norm = normalize_keyword(search_string)
words = search_string_norm.split()
id_ = self.searcher.start_search(words)
return redirect("/search/%s" % "+".join(words))
return redirect("/")
def on_search(self, request):
if request.method == "GET":
return self.render_template('search.html')
elif request.method == "POST":
search_string = request.form.get("keywords", None)
search_string = urllib.unquote_plus(search_string)
if search_string:
search_string_norm = normalize_keyword(search_string)
words = search_string_norm.split()
self.searcher.start_search(words)
return redirect("/search/%s" % "+".join(words))
return redirect("/")
def on_search_results(self, request, key):
search_string_norm = normalize_keyword(key)
words = search_string_norm.split()
start = int(request.args.get("start", 0))
quantity = int(request.args.get("quantity", config.SEARCH_RESULTS))
id_ = self.searcher.start_search(words)
results = self.searcher.get_results(id_, start, quantity)
CLEAN = re.compile("[(),]")
# group by link, giving priority to the title of the original articles
grouped_results = {}
for link, title, ptje, original, texto in results:
# remove 3 dirs from link and add the proper base url
link = "%s/%s" % (ARTICLES_BASE_URL, to3dirs.from_path(link))
# convert tokens to lower case
tit_tokens = set(CLEAN.sub("", x.lower()) for x in title.split())
if link in grouped_results:
(tit, prv_ptje, tokens, txt) = grouped_results[link]
tokens.update(tit_tokens)
if original:
# save the info of the original article
tit = title
txt = texto
grouped_results[link] = (tit, prv_ptje + ptje, tokens, txt)
else:
grouped_results[link] = (title, ptje, tit_tokens, texto)
# clean the tokens
for link, (tit, ptje, tokens, texto) in grouped_results.iteritems():
tit_tokens = set(CLEAN.sub("", x.lower()) for x in tit.split())
tokens.difference_update(tit_tokens)
# sort the results
candidates = ((k, ) + tuple(v) for k, v in grouped_results.iteritems())
sorted_results = sorted(candidates, key=operator.itemgetter(2),
reverse=True)
return self.render_template('search.html',
search_words=words,
results=sorted_results,
start=start,
quantity=quantity
)
def on_search_results(self, request, key):
search_string_norm = urllib.unquote_plus(normalize_keyword(key))
words = search_string_norm.split()
start = int(request.args.get("start", 0))
quantity = int(request.args.get("quantity", config.SEARCH_RESULTS))
id_ = self.searcher.start_search(words)
results = self.searcher.get_results(id_, start, quantity)
CLEAN = re.compile("[(),]")
# group by link, giving priority to the title of the original articles
grouped_results = {}
for link, title, ptje, original, texto in results:
# remove 3 dirs from link and add the proper base url
link = "%s/%s" % (ARTICLES_BASE_URL, to3dirs.from_path(link))
# convert tokens to lower case
tit_tokens = set(CLEAN.sub("", x.lower()) for x in title.split())
if link in grouped_results:
(tit, prv_ptje, tokens, txt) = grouped_results[link]
tokens.update(tit_tokens)
if original:
# save the info of the original article
tit = title
txt = texto
grouped_results[link] = (tit, prv_ptje + ptje, tokens, txt)
else:
grouped_results[link] = (title, ptje, tit_tokens, texto)
# clean the tokens
for link, (tit, ptje, tokens, texto) in grouped_results.iteritems():
tit_tokens = set(CLEAN.sub("", x.lower()) for x in tit.split())
tokens.difference_update(tit_tokens)
# sort the results
candidates = ((k, ) + tuple(v) for k, v in grouped_results.iteritems())
sorted_results = sorted(candidates,
key=operator.itemgetter(2),
reverse=True)
return self.render_template('search.html',
search_words=words,
results=sorted_results,
start=start,
quantity=quantity)
