ocd-mkhwds-terms

#!/usr/bin/python

import urllib, urllib2, json, sys

base_url = 'https://api.sketchengine.co.uk/corpus/'
method = 'wordlist'

corp = sys.argv[1]
refcorp = sys.argv[2]
size = int(sys.argv[3])
usr = sys.argv[4]
apikey = sys.argv[5]


attrs1 = dict(corpname=corp, format='json', api_key=apikey,
              wlattr='ws_terms', wltype='keywords', ref_corpname=refcorp)
attrs2 = dict(corpname=corp, format='json', api_key=apikey,
              wlattr='lempos', wltype='keywords', ref_corpname=refcorp,
              wlpat='\w*-[nvja]')

encoded_attrs1 = urllib.quote(json.dumps(attrs1))
encoded_attrs2 = urllib.quote(json.dumps(attrs2))
url1 = base_url + method + '?username=%s;json=%s' % (usr, encoded_attrs1)
url2 = base_url + method + '?username=%s;json=%s' % (usr, encoded_attrs2)
request1 = urllib2.Request(url1)
request2 = urllib2.Request(url2)
file1 = urllib2.urlopen(request1)
file2 = urllib2.urlopen(request2)
data1 = file1.read()
data2 = file2.read()
file1.close()
file2.close()
data1 = json.loads(data1)
data2 = json.loads(data2)

result = []
for item in data1['Keywords'][:(size/2)]:
    q = ''
    term = item['str'].split()[2].split('_')
    for w in term:
        q += '[lc=="%s"|lemma_lc=="%s"]' % (w, w)
    result.append((' '.join(term), q))

if 'Keywords' in data2:
    for item in data2['Keywords'][:(size-size/2)]:
        result.append((item['str'], '[lempos=="%s"]' % item['str']))
else:
    for item in data1['Keywords'][(size/2):size]:
        q = ''
        term = item['str'].split()[2].split('_')
        for w in term:
            q += '[lc=="%s"|lemma_lc=="%s"]' % (w, w)
        result.append((' '.join(term), q))

for term, query in result:
    entry = {"hw": term.encode('utf8'), "conclink": query.encode('utf8')}
    print json.dumps(entry)