galaxy icon indicating copy to clipboard operation
galaxy copied to clipboard

Extend regex groups in stdio regex matches

Open bernt-matthias opened this issue 2 years ago • 5 comments

Allow to extend groups in the matching regex in the description.

Wondering where to description is shown? Is this broken? The screenshot shows the job info page which shows that the Job Messages are shown as dict and the job details in the history show An error has occurred with this dataset.

Screenshot from 2023-11-13 15-30-29

How to test the changes?

(Select all options that apply)

  • [x] I've included appropriate automated tests.
  • [ ] This is a refactoring of components with existing test coverage.
  • [ ] Instructions for manual testing are as follows:
    1. [add testing steps and prerequisites here if you didn't write automated tests covering all your changes]

License

  • [x] I agree to license these and all my past contributions to the core galaxy codebase under the MIT license.

bernt-matthias avatar Nov 13 '23 14:11 bernt-matthias

The failing test seems relevant. Very exciting work though - this is awesome.

jmchilton avatar Nov 27 '23 21:11 jmchilton

Should we use CDTA in the examples?

bgruening avatar Nov 28 '23 19:11 bgruening

Thanks @mvdbeek :)

@bgruening CDTA?

bernt-matthias avatar Nov 29 '23 07:11 bernt-matthias

I think we can't use CTATA in attributes, isn't it?

bernt-matthias avatar Nov 29 '23 08:11 bernt-matthias

@hujambo-dunia could you have a look at the failing client unit test?

bernt-matthias avatar Feb 15 '24 15:02 bernt-matthias

@hujambo-dunia can you, please, take a look at the failing test?

jdavcs avatar Feb 26 '24 20:02 jdavcs

Sure - apologies for missing this email earlier.

On Mon, Feb 26, 2024, 3:03 PM John Davis @.***> wrote:

@hujambo-dunia https://github.com/hujambo-dunia can you, please, take a look at the failing test?

— Reply to this email directly, view it on GitHub https://github.com/galaxyproject/galaxy/pull/17016#issuecomment-1965153309, or unsubscribe https://github.com/notifications/unsubscribe-auth/AA4AVSYHPM4C23GYVCIO2PLYVTS7XAVCNFSM6AAAAAA7JH5HKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNRVGE2TGMZQHE . You are receiving this because you were mentioned.Message ID: @.***>

hujambo-dunia avatar Feb 26 '24 20:02 hujambo-dunia

@bernt-matthias This PR should fix the broken CI test. Please let me know if have any questions. https://github.com/bernt-matthias/galaxy/pull/8

hujambo-dunia avatar Feb 27 '24 03:02 hujambo-dunia

https://github.com/galaxyproject/galaxy/actions/runs/8061293163/job/22018792003?pr=17016#step:17:3606 is surely related

mvdbeek avatar Feb 27 '24 16:02 mvdbeek

https://github.com/galaxyproject/galaxy/pull/17016/commits/da0dd991e8101791109d1d292774ca1d8240c312 fixes this problems. Type annotation or introducing a type for such error messages might be a better long term solution, but I do not know where to start (and I guess it will be beyond the scope of this PR).

bernt-matthias avatar Feb 28 '24 16:02 bernt-matthias

Can you, please, resolve the conflict so we can merge this? @bernt-matthias @hujambo-dunia

jdavcs avatar Mar 01 '24 15:03 jdavcs

Can you, please, resolve the conflict so we can merge this? @bernt-matthias @hujambo-dunia

Could you take a look @hujambo-dunia .. I have no clue about this part of the code :)

bernt-matthias avatar Mar 01 '24 15:03 bernt-matthias

Can you, please, resolve the conflict so we can merge this? @bernt-matthias @hujambo-dunia

Could you take a look @hujambo-dunia .. I have no clue about this part of the code :)

Sure thing - taking a look now.

hujambo-dunia avatar Mar 01 '24 15:03 hujambo-dunia

I am sorry, but this will have to be rebased, starting with 5c238616a6f557e63f3a32a576acf632a6308aa8 That commit introduced a merge conflict (see these lines >>>/<<</===), which should have been resolved before committing. Otherwise we'll have several broken commits in the commit history. My suggestion would be to squash all subsequent commits except 4140efdad9b6af64b21d6f25acb20faf512edd84 and the last one: once the conflict is fixed, the last commit won't be necessary.

jdavcs avatar Mar 01 '24 18:03 jdavcs

I am sorry, but this will have to be rebased, starting with 5c238616a6f557e63f3a32a576acf632a6308aa8 That commit introduced a merge conflict (see these lines >>>/<<</===), which should have been resolved before committing. Otherwise we'll have several broken commits in the commit history. My suggestion would be to squash all subsequent commits except 4140efdad9b6af64b21d6f25acb20faf512edd84 and the last one: once the conflict is fixed, the last commit won't be necessary.

Sorry about that! I'll do as you recommend.

hujambo-dunia avatar Mar 01 '24 19:03 hujambo-dunia

I think the 2 last commits that touch the JobInformation.vue file are incorrect (e0039f76e2a58b357214f7951d18630851c972af and 91eb77390ac289cb85bde078a9a967e02cfeb93d): they override the changes introduced in #17566. Also, the last commit which is empty can be dropped.

jdavcs avatar Mar 04 '24 15:03 jdavcs

I think the 2 last commits that touch the JobInformation.vue file are incorrect (e0039f7 and 91eb773): they override the changes introduced in #17566. Also, the last commit which is empty can be dropped.

Agreed - I see what you're saying, including JohnC's commits from 5 days ago getting dropped in my last merge, and will make those adjustments shortly.

hujambo-dunia avatar Mar 04 '24 16:03 hujambo-dunia

I'm sorry for being the annoying one here! Here's the thing: those 2 commits I mentioned are still present, and they take the code into a broken state. Even though it's for a short span of 3 commits, fixed in the forth, it is not ideal: the history doesn't tell a coherent story, and if anyone happens to end up on one of those commits (a checkout, a pause in a rebase, etc.), they may be confused.

The fix is simple: reorder the commits, so that those two are followed by the fix. Then squash the three into one. Then you'll have one commit that handles those changes, without any do/undo parts.

jdavcs avatar Mar 04 '24 19:03 jdavcs

@hujambo-dunia The rebase look correct - thank you!

jdavcs avatar Mar 05 '24 15:03 jdavcs

Looks good to me. Waiting for the tests before merging (there may will be failures that are unrelated). Thanks again for addressing the comments and handling the rebase, @hujambo-dunia!

@jdavcs it helps to have such a great teacher!

hujambo-dunia avatar Mar 05 '24 16:03 hujambo-dunia

Can you, please, take a look at the client linting error?

jdavcs avatar Mar 05 '24 17:03 jdavcs