Importer script for moving Unfuddle tickets to JIRA.

This includes some newer functionality JIRA enabled that aren't covered by existing Unfuddle/JIRA migration scripts: in particular, comments, milestones as proper epics, ticket linking, custom select-field value mappings, and attachments.

Follow the directions under the Unfuddle project settings to get your project's backups. Extract the backup.complete directory from the archives to the same directory as the unfuddle_to_jira.rb script. Edit the Ruby script to set PROJECT_NAME, ISSUE_NUMBER_OFFSET, and IMPORT_USER. Set CUSTOM_USER_MAPPINGS if you have existing JIRA users whose usernames are different than the ones they used in Unfuddle, and set BACKUP_FILE and ATTACHMENTS_DIR if your Unfuddle backup lives somewhere else. Since this script will rename attachment files from the backup, you should keep the original backup archive around locally in case you need to start over from scratch without re-downloading your backup.

Install the script's dependencies with bundle install, then run the script:

bundle exec ./unfuddle_to_jira.rb

Caveats:

• Milestones that are archived or completed are closed; all other milestones will be left open.
• Custom field value mappings are tested with single-selects, and nothing else at the moment.
• All ticket links are mapped to one type in the CSV, regardless of parent/child/duplicate relationship status in Unfuddle.
• Severity is pulled in, but currently not priority (for historical reasons specific to our project).
• We assume comments, descriptions, etc. are written in Markdown in Unfuddle. Only a few basic bits of Markdown from Unfuddle are converted to JIRA's proprietary syntax. JIRA OnDemand currently has no way to render Markdown directly, so more advanced formatting may be mangled.
• Attachment files aren't included in the CSV, but are moved/renamed to subfolders that can then be uploaded via JIRA's bulk attachment upload script.
• Comment attachments are converted to attachments on the associated tickets; JIRA's importer can't do comment-specific attachments.

Note that there are some invalid characters that can be passed through in Unfuddle's XML; you may need to remove those manually for this to process.

The included sample importer config, jira-importer-config.txt, can be uploaded to JIRA along with your CSV as a starting point. You'll need to select 'Map Field Value' for the status, resolution, severity, and issue-type fields to map to your particular JIRA setup's field values, as well as any custom field mappings you might have.

For more technical details, see JIRA's documentation onCSV imports and the file format.