How do I pull time estimates via a script?

[Anthony Newnam]

I see several issues against JIRA on pulling time estimate information with conflicting details. Is there any way through any mechanism except scraping HTML and parsing it off the site (which I guess I'll do next) to pull this data?

[Bob Swift (personal)]

JIRA Command Line Interface can be used with getIssue or getIssueList actions. Use --outputFormat 2 for the single issue action or --outputFormat 5 for the list action to get the time fields.

[Anthony Newnam]

Thanks, that's perfect. I had tried that without the outputFormat. Is there any documentation that lists the options to the outputFormat?

[Bob Swift (personal)]

getIssueList far right column although I just noticed getIssue needs it documented.

[Anthony Newnam]

Thanks, that's very helpful! I've changed my scripts around to use the CLI instead of manually saved files from JIRA. This seems to do everything I need now except give me information about parent tasks or subtasks. I would like to be able to filter either for issues that have no parent task (top level issues) or issues that have no subtasks (bottom level issues). I don't think this is possible in JQL, and I don't get parent or subtask information back from getIssueList.

Any ideas?

Thanks again for your help, Bob.

[Bob Swift (personal)]

Look at https://studio.plugins.atlassian.com/wiki/display/JQLT/Subtask+Functions - might be something to help there. You are right there is little support for parent/subtask info in the cli at this time, although I think with the newer JIRA releases there is more information available that should be incorporated.

[Anthony Newnam]

Do you see significantly slower query times using the CL client versus the API or web?

To query for "project = something and resolution = unresolved" on the web, it takes a few seconds. On the command line, it takes me 17 minutes.

[Bob Swift (personal)]

No, not normally. Did the query return a large number of issues? If so all (by default) are retrieved via the CLI and only the first page is retrieved on the GUI. Use --count 50 to limit the number returned. And a tip for larger queries or other memory intensive actions, increase your default java memory.

[Anthony Newnam]

It returned several hundred (I think about 600). The XML view on the web returns all records in my query too, and it seems to be very fast.