python: extract only 'bug fixed' desc

bug_topic.py

import json
import cld

input_json_file = 'apps_public_m_t_desc.json'
output_bugs_desc_file = 'bugs_desc.txt'

with open(input_json_file) as fp:
    json_contents = []
    bugs_desc = []
    for line in fp.readlines():
        contents = json.loads(line)
        desc = contents['desc'].lower()
        if ("bug fixes" in desc) or \
           ("bug fixed" in desc) or \
           ("fixed bug" in desc) or \
           ("known bug" in desc):
            lang = cld.detect(desc.encode('utf-8'))
            if lang[1] == 'en' and len(lang[4]) == 1:
                json_contents.append(line)
                bugs_desc.append(desc)

print "total:", len(json_contents)
with open(output_bugs_desc_file, "wb") as fp:
    for line in bugs_desc:
        fp.write("%s\n" % line.encode('utf8'))

mogoexport

$ mongoexport -h guitarxx.cs.xxx.xxx -d apps -c public -f m,t,desc -o apps_public_m_t__desc.json

Number of description

a. total: 201,830
b. filter(bug fixes, bug fixed, fixed bug, know bug): 791
c. b by english only: 639