Discourse2Jira

About

This project is built specifically to automate the creation of Jira issues for each new topic in the 'store-requests' category of forum.snapcraft.io

Some of it has been generalised, but there are many elements of it that are specific to that use case.

Included is code from lvoytek's dsctriage. Thanks Lena.

Install / Quickstart

1. Clone the repo:

    $ git clone git@github.com:dclane-code/Discourse2Jira.git

2. Create a virtual env and install the module (and its dependencies)

    $ cd Discourse2Jira
    $ python3 -m venv env
    $ . env/bin/activate
    $ pip3 install .

3. Create a config file

    $ cp example_config.json config.json
    [edit config.json in your favourite editor]

4. Insert a Jira API token into your config.

In order to be able to post to Jira, the script needs some credentials. Use an API token that you can generate here

    https://id.atlassian.com/manage-profile/security/api-tokens

5. Create an empty database

   $ touch forum_db.json

6. Run (see initial setup below before removing '-n')

    $ dsc2jira -n  # dry run, see what might happen with no databse
    $ dsc2jira -i  # initialise the database so that no initial issues are created
    (optional) edit the database as described below 
    $ dsc2jira     # all future runs :)

The database format

Each discourse topic is represented in a json 'database' file by default called forum_db.json.

The format for each topic is

{
    "id": 36045,
    "slug": "request-for-a-new-pc-gadget-snap-track-classic-23-10",
    "jira": "SEC-2457",
    "url": "https://forum.snapcraft.io/t/request-for-a-new-pc-gadget-snap-track-classic-23-10/36045/"
}

Important If a topic does not exist in the database, or has "jira": "0", then a Jira issue will be created for it.

Initial setup

(first time only, prevents excess Jira issue creation)

To avoid unnecessary Jira issue creation on first run (with an empty database), you should perform a database initialization with:

$ dsc2jira -i

• This will produce a forum_db.json that has "skip" for each Jira entry (instead of actual Jira issue numbers)
• Inspect the database file and the forum manually and for any forum topics that you would like Jira issues created for
  • Replace the value "skip" with a 0.
  • On subsequent runs, any database items that have a Jira value of 0, will also have issues created for them, and that field updated.
  • You can use dry-run -n to test

Search period

By default, Discourse2Jira searches for forum topics back 7-days from now(). This can be overriden with the -d argument, eg: dsc2jira -d 4w to create Jira cards for the last 4 weeks worth of forum topics