Trac Search

Bring the modernity of Elastic Search to the classical Trac.

Trac is a nice tool, but it search engine is confusing.

Install it

Create a nice virtualenv.

pip install -r requirements.txt
python setup.py install

The command line application is named tracsearch.

Try it

Launch a local elastic.

Configure

cp tracsearch.ini.sample tracsearch.ini
vim tracsearch.ini

Index some tickets

tracsearch ticket

Launch the web application

tracsearch web

Deploy it

For now, there is no automatic indexation.

Virtualenv and supervisord

Use supervisord to launch the web application.

Use the source folder as virtualenv folder.

pip install gunicorn

Edit /etc/supervisor/conf.d/tracsearch.conf

[program:tracsearch]
user=tracsearch
command=/opt/tracsearch/bin/gunicorn -b 127.0.0.1:5000 -w 4 tracsearch.web:app
directory=/opt/tracsearch
environment=PATH="/opt/tracsearch/bin"

Sentry

You can use Sentry to watch Tracsearch's errors.

Grab your sentry dsn and add it to tracsearch.ini

[sentry]
dsn=http://public_key:[email protected]/1

You can use Tracseerch without a sentry section in the conf file.

Todo

• √ Index ticket and comment
• √ Index wiki
• √ Web UI
• √ Time line
• √ Sentry integration
• √ Multiple Trac
• _ Plugins
• _ Angular UI
• _ Trac module to live index
• _ Redmine variant
• _ Blood hound integration

Licence

BSD © Mathieu Lecarme 2013