Add option to skip ordering of issues so ORDER BY in JQL can be used for more complex sorting

[20manas]

For example, I want to ORDER BY resolution DESC, rank but I can't do it with --order-by. If I pass ORDER BY to --jql, it doesn't work because the cli is always passing its own ORDER BY. So if there is a functionality to stop automatic passing of ORDER BY, it will allow the users to pass their own order in the jql string.

I can see a few ways to implement this:

• Add --no-order or --skip-order
• Add --order-by 'somevalue' which will skip sorting. somevalue can be null, nil, empty or something else
• If ORDER BY is present in the jql string, the cli should not pass its own ORDER BY

Thanks!

Edit: Actually, for my specific case, I was able to do it with --order-by 'resolution DESC,rank' --reverse. Although this functionality can still be useful for other queries/scenarios, so leaving this discussion open for now.

[HarrisonMc555]

Alternatively, we could also allow users to provide a finer-grained control of the ordering via the existing --order-by flag. I'm not sure of the best syntax, but something like this would be feasible:

jira issue list --order-by 'priority DESC, status ASC'

...in fact, this is already kind of supported...in a very janky way.

I'm guessing that the argument to the --order-by argument is simply interpolated into a raw JQL query string. I'm guessing that a DESC or ASC flag is appended to the end (as determined by the presence or absence of the --reverse flag).

So, you can actually specify arbitrarily complex ORDER BY clauses by simply adding in the ASC or DESC clauses to your --order-by argument, as long as the final argument doesn't have a ASC or DESC clause.

For example, this works:

jira issue list --order-by 'priority DESC,status ASC,duedate ASC,created'

(note that the final created column does not have an ASC or DESC modifier)

This is a weird workaround. It would be nice if this were supported in a better way. I think the best way would probably be to either recognize if the final "word" was ASC or DESC (case-insensitively) and refrain from adding an additional ASC/DESC if so, or to deprecate the --reverse flag and require users to specify ASC or DESC themselves.