PartyCrasher

PartyCrasher is a framework for automatically grouping crash reports based on the contents of stack traces and other data available at the time of a crash. You can read the paper about it.

Install

Note that the provided Dockerfile is out of date.

PartyCrasher requires Elastic (ElasticSearch).

To install the PartyCrasher REST client, install ElasticSearch; then install the Python dependencies (you may want to install these within a virtualenv if that's your taste).

pip install -r requirements.txt

To install the UI, you must have NPM.

cd ui
npm install

Usage

Run the REST service:

make start

The HTTP service will now be listening on port 5000. Refer to the [API Docs][] for further usage.

Uploading Test Data

Note: This will delete all data in the existing PartyCrasher instance.

To download test data and upload it to PartyCrasher, simply type:

$ make buckettest

Licensing

Assume that PartyCrasher is licensed under the GPL3+ unless otherwise specified.

Citation

If you use this code we would appreciate if you cited the paper!

@inproceedings{Campbell:2016:UET:2901739.2901766,
   author = {Campbell, Joshua Charles and Santos, Eddie Antonio and Hindle, Abram},
   title = {The Unreasonable Effectiveness of Traditional Information Retrieval in Crash Report Deduplication},
   booktitle = {Proceedings of the 13th International Workshop on Mining Software Repositories},
   series = {MSR '16},
   year = {2016},
   isbn = {978-1-4503-4186-8},
   location = {Austin, Texas},
   pages = {269--280},
   numpages = {12},
   url = {http://doi.acm.org/10.1145/2901739.2901766},
   doi = {10.1145/2901739.2901766},
   acmid = {2901766},
   publisher = {ACM},
   address = {New York, NY, USA},
   keywords = {automatic crash reporting, call stack trace, contextual information, deduplication, duplicate bug reports, duplicate crash report, free/open source software, information retrieval, software engineering},
}

naturalness / partycrasher