Bugzilla Analysis Program is an analysis tool that uses some basic Natural Language Processing to extract descriptive information from bug reports.
Bugzap uses Distiller: Automated Keyword Extraction from Document Collections
###Scraping Bugzilla
To start scraping bug reports for a Bugzilla instance, run:
$ scrapy crawl query.bugzilla -a query=queryfile -o outfile -t json
The query file needs the domain of the Bugzilla instance in its first line, plus at least one search query URL to start the scrape from:
mybugzillainstance.com
https://issues.mybugzillainstance.com/buglist.cgi?email1=myemail%40email.com&emailassigned_to1=1&emailcc1=1&emailreporter1=1&emailtype1=exact&list_id=2581552
Scrapy will store all retrieved bug reports to outfile in json format.
###Processing Documents
To process the retrieved bug reports, run:
$ python bugzap/main.py -j outfile -r bugzilla -b mylacklist -n mybugs
All reports will be saved to the bugzap/visualization/data/mybugs folder.
For further info, see: $ python buzap/main.py -h
###Viewing Statistics
The reports can be viewed from Mozilla by opening bugzap/visualization/freqDist.html file and entering the name of your data set:
$ firefox bugzap/visualization/freqDist.html
