[Bug] Ignoring a SHA in the configuration does not seem to exclude it from the version calculation

[andysbolton]

Describe the bug
I may be confused about how this feature works, but when trying to ignore a given SHA (or a commit-befores date) from the version calculation, it seems to be included anyway.

Expected Behavior

Given this line from /showconfig:

...
ignore:
  sha:
  - 619b00106f14ccff9406aa07391eb42433b4cd4d

I would expect this commit to be excluded from the versioning calculation.

Actual Behavior

When I look into the diagnostic logs, I see the commit is still being considered for version calculation:
INFO [03/06/23 8:38:22:48] Merge commit 619b001 Merge pull request #43 from repo/OPS-698 incremented base versions Major, now 2.0.0

Possible Fix

Steps to Reproduce

When I have a bit more time, I will try to recreate this, but I wanted to rule out that this is due to a misunderstanding of this feature on my part.

Context

An accidental commit has caused the versioning to increment a Major version.

Your Environment

• Version Used: 5.12.0+Branch.support-5.x.Sha.3f75764963eb3d7956dcd5a40488c074dd9faf9e
• Operating System and version (Windows 10, Ubuntu 18.04): Windows 11
• Link to your project:
• Link to your CI build (if appropriate):

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. After 30 days from now, it will be closed if no further activity occurs.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. After 30 days from now, it will be closed if no further activity occurs.

[corentinvds]

Hello,

I have the same issue: A commit message accidentally contains "+semver: major" and adding its Sha to the "ignore:sha" list in the config does not prevent Git Version to increment the major version.

Do you have a workaround other than rewriting the git history?

Thanks

[corentinvds]

I had a look at the code and found out that the "ignore" list is only used to filter out "base version" candidates but not to prevent bad commits from incrementing the version. To do that, the "ignore" list should be use near IncrementStrategyFinder.FindCommitMessageIncrement.

Would it be possible to update the code to implement this filter too ?

If not, what would be the recommended workaround for a bad commit message causing a major version update ?

Thanks.

[asbjornu]

@corentinvds, we could make such backwards-incompatible changes in v6. I think it makes sense that all ignore properties lead to the ignored commits to be excluded from all version calculation. Pull requests are welcome!