def get_keyword_relevancy_map(self, name_list, n_list, terms, sortcontext,
enginetype='BING',
license=None):
"""
http://www.clips.ua.ac.be/pages/pattern-web#sort
"""
results_list = []
self.name_list = name_list
results = sort(
terms=[],
context=sortcontext, # Term used for sorting.
service=enginetype, # GOOGLE, YAHOO, BING, ...
license=None, # You should supply your own API license key
# for the given service.
strict=True, # Wraps the query in quotes, i.e.'mac sweet'.
reverse=True, # Reverses term and context: 'sweet mac' <=>
# 'mac sweet'.
cached=True)
for weight, term in results:
results.append("%5.2f" % (weight * 100) + "%", term)
return {
'data': results_list,
'summary': None
}
import os, sys; sys.path.insert(0, os.path.join("..", ".."))
from pattern.web import GOOGLE, YAHOO, BING, sort
# The pattern.web module includes an interesting sort() algorithm.
# Ir classifies search terms according to a search engine's total results count.
# When a context is defined, it sorts according to relevancy to the context:
# sort(terms=["black", "green", "red"], context="Darth Vader") =>
# yields "black" as the best candidate,
# because "black Darth Vader" yields more search results.
results = sort(
terms = [
"arnold schwarzenegger",
"chuck norris",
"dolph lundgren",
"steven seagal",
"sylvester stallone",
"mickey mouse",
],
context = "dangerous", # Term used for sorting.
service = BING, # GOOGLE, YAHOO, BING, ...
license = None, # You should supply your own API license key for the given service.
strict = True, # Wraps the query in quotes, i.e. 'mac sweet'.
reverse = True, # Reverses term and context: 'sweet mac' instead of 'mac sweet'.
cached = True)
for weight, term in results:
print "%5.2f" % (weight * 100) + "%", term
with open(fname) as f:
diction = f.readlines()
for term in diction:
if len(term) > lengthmin:
subset.append(term.strip('\n'))
# function to get a random term from the minlength dictionary in subset list
def rando(listofterms, num):
i = 0
while i < num:
randomed = random.choice(listofterms)
#print randomed
searchlist.append(randomed)
i = i + 1
return
searchlist = [
] # the list of terms that will be generated in the rando function
# setup the default search terms
rando(
subset, numterms
) # get total list of terms based on numterms set in the globals section above
from pattern.web import sort
results = sort(terms=searchlist, context=contexter, prefix=True)
for weight, term in results:
print "%.2f" % (weight * 100) + '%', term
exit()
sys.path.insert(0, os.path.join("..", ".."))
from pattern.web import GOOGLE, YAHOO, BING, sort
# The pattern.web module includes an interesting sort() algorithm.
# Ir classifies search terms according to a search engine's total results count.
# When a context is defined, it sorts according to relevancy to the context:
# sort(terms=["black", "green", "red"], context="Darth Vader") =>
# yields "black" as the best candidate,
# because "black Darth Vader" yields more search results.
results = sort(
terms=[
"arnold schwarzenegger",
"chuck norris",
"dolph lundgren",
"steven seagal",
"sylvester stallone",
"mickey mouse",
],
context="dangerous", # Term used for sorting.
service=BING, # GOOGLE, YAHOO, BING, ...
license=
None, # You should supply your own API license key for the given service.
strict=True, # Wraps the query in quotes, i.e. 'mac sweet'.
reverse=
True, # Reverses term and context: 'sweet mac' instead of 'mac sweet'.
cached=True)
for weight, term in results:
print "%5.2f" % (weight * 100) + "%", term
from pattern.web import sort
results = sort(terms=[
'sunset', 'puppies', 'kittens', 'babies', 'hedgehogs', 'birthday',
'bunnies'
],
context='joy',
prefix=True)
for weight, term in results:
print "%.2f" % (weight * 100) + '%', term
diction = []
subset = []
lengthmin = 6
numterms = 10
fname = 'assets/dictionary-list.html'
with open(fname) as f:
diction = f.readlines()
for term in diction:
if len(term) > lengthmin:
subset.append(term.strip('\n'))
# function to get a random term from the minlength dictionary in subset list
def rando(listofterms,num):
i = 0
while i < num:
randomed = random.choice(listofterms)
#print randomed
searchlist.append(randomed)
i = i + 1
return
searchlist = [] # the list of terms that will be generated in the rando function
# setup the default search terms
rando(subset,numterms) # get total list of terms based on numterms set in the globals section above
from pattern.web import sort
results = sort(terms=searchlist,context=contexter,prefix=True)
for weight, term in results:
print "%.2f" % (weight * 100) + '%', term
exit()
from pattern.web import Google, plaintext, sort
from optparse import OptionParser
parser = OptionParser()
parser.add_option("-q", "--query", dest="query",
help="txt file with searchwords")
parser.add_option("-f", "--filename", dest="filename",
help="specify filename to save data too")
parser.add_option("-c", "--context", dest="context",
help="specify context for search")
(options, args) = parser.parse_args()
# open the list of query words
f = codecs.open(options.query, encoding='utf-8')
# build results in context of the "Energiewende"
results = sort(terms=[x.strip() for x in f], context = options.context)
# create weights
if results.count() > 0:
for weight, term in results:
#term = term.encode('utf-8') <-- needed?
with codecs.open(options.filename, "a", "utf-8") as f:
try:
f.write(term.encode("utf-8"))
f.write(',')
f.write("%5.2f" % (weight * 100))
f.write('\n')
except UnicodeDecodeError as e:
print e
else:
print "No results. Sorry."
