Update version script to remove bump2version dependency #41
Conversation
gigony
added
breaking
ci/release/update-version.sh
Outdated
| @@ -11,7 +11,7 @@ | |||
| NEXT_FULL_TAG=$1 | |||
|
|
|||
| # Get current version | |||
| CURRENT_TAG=$(git tag | grep -xE 'v[0-9\.]+' | sort --version-sort | tail -n 1 | tr -d 'v') | |||
| CURRENT_TAG=$(cat VERSION) | |||
There was a problem hiding this comment.
This change actually brought to my attention some issues with this script. I have some changes locally that I will put into a new PR that you can merge in here.
There was a problem hiding this comment.
yes, @jakirkham. you beat me to it 🙂. Ray and I had to fix some things in this script. We'll be pushing those changes to all the other repos as well.
There was a problem hiding this comment.
@ajschmidt8 cuCIM project depends on VERSION file that is used in building/packaging (cmake/python).
According to update-version.sh's use (executed for each branching), I think getting the current tag (without 'v' or 'a' in it) from VERSION file does not impact on existing workflow. Please let me know if there is an issue with this.
There was a problem hiding this comment.
I think using the tag should be okay. I will test the update-version.sh script locally first before approving this PR.
ci/release/update-version.sh
Outdated
| sed_runner "s/${CURRENT_TAG}/${NEXT_FULL_TAG}/g" python/cucim/VERSION | ||
| sed_runner "s/${CURRENT_TAG}/${NEXT_FULL_TAG}/g" cpp/plugins/cucim.kit.cuslide/VERSION | ||
| sed_runner "s#\[Version ${CURRENT_TAG}\](release_notes/v${CURRENT_TAG}.md)#\[Version ${NEXT_FULL_TAG}\](release_notes/v${NEXT_FULL_TAG}.md)#g" python/cucim/docs/index.md | ||
| sed_runner "s/v${CURRENT_TAG}/v${NEXT_FULL_TAG}/g" python/cucim/docs/getting_started/index.md | ||
| sed_runner "s#cucim.kit.cuslide@${CURRENT_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" python/cucim/docs/getting_started/index.md | ||
| sed_runner "s#cucim.kit.cuslide@${CURRENT_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" cucim.code-workspace |
There was a problem hiding this comment.
| sed_runner "s/${CURRENT_TAG}/${NEXT_FULL_TAG}/g" VERSION | |
| sed_runner "s/${CURRENT_TAG}/${NEXT_FULL_TAG}/g" python/cucim/VERSION | |
| sed_runner "s/${CURRENT_TAG}/${NEXT_FULL_TAG}/g" cpp/plugins/cucim.kit.cuslide/VERSION | |
| sed_runner "s#\[Version ${CURRENT_TAG}\](release_notes/v${CURRENT_TAG}.md)#\[Version ${NEXT_FULL_TAG}\](release_notes/v${NEXT_FULL_TAG}.md)#g" python/cucim/docs/index.md | |
| sed_runner "s/v${CURRENT_TAG}/v${NEXT_FULL_TAG}/g" python/cucim/docs/getting_started/index.md | |
| sed_runner "s#cucim.kit.cuslide@${CURRENT_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" python/cucim/docs/getting_started/index.md | |
| sed_runner "s#cucim.kit.cuslide@${CURRENT_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" cucim.code-workspace | |
| sed_runner "s/${CURRENT_LONG_TAG}/${NEXT_FULL_TAG}/g" VERSION | |
| sed_runner "s/${CURRENT_LONG_TAG}/${NEXT_FULL_TAG}/g" python/cucim/VERSION | |
| sed_runner "s/${CURRENT_LONG_TAG}/${NEXT_FULL_TAG}/g" cpp/plugins/cucim.kit.cuslide/VERSION | |
| sed_runner "s#\[Version ${CURRENT_LONG_TAG}\](release_notes/v${CURRENT_LONG_TAG}.md)#\[Version ${NEXT_FULL_TAG}\](release_notes/v${NEXT_FULL_TAG}.md)#g" python/cucim/docs/index.md | |
| sed_runner "s/v${CURRENT_LONG_TAG}/v${NEXT_FULL_TAG}/g" python/cucim/docs/getting_started/index.md | |
| sed_runner "s#cucim.kit.cuslide@${CURRENT_LONG_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" python/cucim/docs/getting_started/index.md | |
| sed_runner "s#cucim.kit.cuslide@${CURRENT_LONG_TAG}.so#cucim.kit.cuslide@${NEXT_FULL_TAG}.so#g" cucim.code-workspace |
@gigony, can you merge the changes from #42 into your PR and then switch all of these to CURRENT_TAG variables to CURRENT_LOG_TAG?
ci/release/update-version.sh
Outdated
| @@ -11,7 +11,7 @@ | |||
| NEXT_FULL_TAG=$1 | |||
|
|
|||
| # Get current version | |||
| CURRENT_TAG=$(git tag | grep -xE 'v[0-9\.]+' | sort --version-sort | tail -n 1 | tr -d 'v') | |||
| CURRENT_TAG=$(cat VERSION) | |||
There was a problem hiding this comment.
can you revert the cat VERSION change here and make sure this line matches the change that was done in #42 after you merge those changes into your PR?
There was a problem hiding this comment.
Actually, to make the statement more robust to tag changes from other branches (e.g., adding a patch tag[21.06.01] to branch-21.06 after creating tag[21.08.00]),
CURRENT_TAG=$(git tag | grep -xE 'v[0-9\.]+' | sort --version-sort | tail -n 1 | tr -d 'v')to
CURRENT_TAG=$(git tag --merged HEAD | grep -xE 'v[0-9\.]+' | sort --version-sort | tail -n 1 | tr -d 'v')
what do you think?
There was a problem hiding this comment.
sure, that seems like a good idea. feel free to add it to this PR.
|
@gigony, looks like you've got some merge conflicts. can you pull the latest from |
Remove scripts/docs related to bump2version because we don't need it anymore since versioning scheme is changed to CalVer.
gigony
force-pushed
the
update_version_script
branch
from
1df0187
to
46ca77b
Compare
|
We've been seeing some installation conflicts (also in CI currently). Unfortunately CI stops short of showing the conflict. Initially suspected these were related to version discrepancies between cuCIM and the rest of RAPIDS. Maybe that is still true and we are just missing something here? Though maybe we are missing something else that is going on? |
|
rerun tests |
Can't say I have any ideas about this. I just reran CI, so let's see if we run into the issue again. |
|
@gigony, any reason you used the |
|
Yeah it's still failing. Unfortunately the builds are being terminated before we get the conflict report. Would it be possible to extend the build time so we could get this information? |
sure. I just bumped it from 40 minutes to 2 hours. rerun tests |
gigony
added
non-breaking
Ah, I thought the 'breaking' label needs to be added if we touch some code but looks like we add it if it breaks things for 'user'? In either case, it doesn't touch internal code and I removed the label. |
|
@jakirkham @ajschmidt8 Do you have any idea on the following build error messages? And shows bunch of conflicts. |
|
@gigony, I think the first error log you posted was a random conda server error. the real issue here is likely with the conflicts in the second log you posted. |
|
from quickly looking at the logs, I noticed this discrepancy in Python versions. this might warrant further investigation.
|
Refactor
update-version.shto not depend onbump2versionutility.