refreshVersions
refreshVersions copied to clipboard
Semantic versioning metadata
Hey,
- [x] I have read the guidelines for submitting issues
β οΈ Is your feature request related to a problem? Please describe
I push library versions following the semantic versioning described on https://semver.org/. When I add metadata, 0.0.0+sha.0000000
, the version isn't picked up by the plugin.
π‘ Describe the solution you'd like
Could the plugin allow versions with metadata?
π€ Do you want to develop this feature yourself?
- [x] Yes, if you can give a few hints :)
- [ ] No
Hello, can you give version examples instead of the generic 0.0.0+sha.0000000
?
We also need the current version, the versions that you expect to be picked up by the plugin but aren't, and whether you're using rejectVersionsIf { β¦ }
.
Hey @LouisCAD,
I just created a small repository to show that versions with a +
aren't listed: https://github.com/plippe/refreshVersions-metadata.
Hope it helps
foo, bar⦠can you rename with more meaningful names?
I am unsure if you are pulling my leg, but I am happy to rename them
Both projects are libraries, but one depends on the another. I can rename the dependency dependancy
and the one that depends dependant
. Would that work for you?
I am kinda thinking the same: You are following a versioning scheme that is unusual in the maven and Kotlin ecosystems, and asking us to support it, then, when I ask for a list of version examples, you link a multi-project repository instead, suggesting us to figure it out. π BTW, I was checking the repo out from my phone (because not always at a desk), navigating is not convenient, but the problem was supposed to not require navigating a project in the first place. Anyway.
dependency
and dependent
are alright, but I peeked again from my computer: I won't follow the steps, I'll stick to my initial request which is a list of versions, which can then be added as test data here: https://github.com/jmfayard/refreshVersions/tree/main/plugins/core/src/test/resources/versions-comparison
Example (fewer versions are fine):
1.0
10.0.0
10.0.0-whatever
10.0.1
10.0.1-whatever
11.0-alpha-4
11.0
11.1-alpha01
11.1-alpha02
11.1-beta01
11.1-beta02
11.1-beta02-whatever
11.1-rc01
11.1-rc01-whatever
11.1-rc02
11.1-rc02-whatever
11.1
11.1-whatever
Sorry,
I thought you were looking for clarification while it seems you were looking for test data.
1.0+123
10.0.0+234
10.0.0-whatever+234
10.0.1+341
10.0.1-whatever+412
11.0-alpha-4+123
11.0+234
11.1-alpha01+341
11.1-alpha02+412
11.1-beta01+123
11.1-beta02+234
11.1-beta02-whatever+341
11.1-rc01+412
11.1-rc01-whatever+123
11.1-rc02+234
11.1-rc02-whatever+341
11.1+412
11.1-whatever+123
Hopefully that data will help
Sorry, I'll try to be more clear next time π
Is that really conforming to the 0.0.0+sha.0000000
format that you describer though?
:)
Those are a few more.
1.0.0-alpha+001
1.0.0+20130313144700
1.0.0-beta+exp.sha.5114f85
1.0.0+21AF26D3β-117B344092BD
They come from the semver.org website, https://semver.org/#spec-item-10. It has a regex expression to match against, if that can help.
One of them has +exp.sha.5114f85
. It isn't exactly like my original example, but quite close.
There are not in order, right?
The test data requires ordered versions, and then, we check our sorting logic keeps the order intact.
In order, it should be
1.0.0-alpha+001
1.0.0-beta+exp.sha.5114f85
1.0.0+20130313144700
1.0.0+21AF26D3β-117B344092BD
Just FYI, this is very low priority for us as very few people in the community uses such versions, sticking to the simple side of semantic versioning. If you can, I recommend you to do the same, would likely be simpler for library users as well. You can at least not use the +
character
Are you sure about the last one? Honestly, it looks like it just exists to inflige pain, I see no reason to have a short and long dash, that follow each other, and what the hex chars are supposed to mean is equally cryptic.
Yes, the last one does seem weird to me too.
Currently we are using hyphen. So this isn't a priority for us either. Anyways, if you point me in the right direction, I can definitely try to do the work.
On Fri 30 Sept 2022, 18:53 Louis CAD, @.***> wrote:
Just FYI, this is very low priority for us as very few people in the community uses such versions, sticking to the simple side of semantic versioning. If you can, I recommend you to do the same, would likely be simpler for library users as well. You can at least not use the + character
Are you sure about the last one? Honestly, it looks like it just exists to inflige pain, I see no reason to have a short and long dash, that follow each other, and what the hex chars are supposed to mean is equally cryptic.
β Reply to this email directly, view it on GitHub https://github.com/jmfayard/refreshVersions/issues/630#issuecomment-1263855107, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAKRKO3DGAUWMZALUV7LTGTWA4SJ3ANCNFSM6AAAAAAQYV6X2A . You are receiving this because you were assigned.Message ID: @.***>