feat(version): add MANUAL_VERSION, --next and --patch to version command

[bearomorphism]

Closes #1679

Manually tested and added test cases, the result LGTM.

[bearomorphism]

@Lee-W I already ran pytest tests/commands/test_version_command.py -n auto --regen-all to regenerate the test files, but the pipeline still fails. (The tests passed on my machine)

Do you have any ideas why this happen

[woile]

I realize now that this creates problems with the (not)monotonic kind of versions (and possible non-semver). I'm not sure what to do about it.

I think for now it's fine that if you diverge too much from semver in your custom version scheme, then you won't get the full range of features.

[woile]

There was a way to run all the pre-commits, but I don't remember how it works 😅

[woile]

what about having a parametrize of the combinations that should fail:

@pytest.mark.parametrize(
    "args",
    [
        # incomplete list:
        {"next"},
        # ...
        {"--verbose", "next"}
    ]
)

[bearomorphism]

Sounds good.

[woile]

Nice to see this going forward 🎉

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.92%. Comparing base (193859b) to head (4a9a5da).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1724      +/-   ##
==========================================
- Coverage   97.93%   97.92%   -0.01%     
==========================================
  Files          60       61       +1     
  Lines        2611     2654      +43     
==========================================
+ Hits         2557     2599      +42     
- Misses         54       55       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[bearomorphism]

@Lee-W I already ran pytest tests/commands/test_version_command.py -n auto --regen-all to regenerate the test files, but the pipeline still fails. (The tests passed on my machine)

Do you have any ideas why this happen

I see why the test failure happens. Please see my other PR #1726 .

[bearomorphism]

I will rebase this branch after #1726 is merged. The test failure should be resolved then.

[noirbizarre]

That's an interesting feature, one that I would totally use!

Also, I never realized that the version command was doing totally different things depending on the flags 😅
So maybe it's the occasion to have a dedicated --version flag and have the version command focusing on the the project versioning and not the Commitizen version. It would be cleaner, easier to understand and more standard. WDYT ?

Note

I would not remove the Type aliases as it makes the PR bigger than it should and it introduce a breaking change (and reading the PR and issues, we know that it has been used by some users for some custom version schemes).

[noirbizarre]

Also, the enumn part seems to overlap with #1518, so maybe #1518 should go first so this PR can reuse the Increment enum

[bearomorphism]

Also, the enumn part seems to overlap with #1518, so maybe #1518 should go first so this PR can reuse the Increment enum

I would prefer to merge this before #1518 because it is a bit larger than this PR.

So maybe it's the occasion to have a dedicated --version flag and have the version command focusing on the the project versioning and not the Commitizen version. It would be cleaner, easier to understand and more standard. WDYT ?

+1. As a user, I often run cz --version and end up seeing an error. We can create another work item about adding the dedicated flag and then deprecate cz version --commitizen and other removable flags in the next major version.

[woile]

We had a --version in the past, but it was replaced by version, do you remember why @Lee-W ?
It would be a breaking change if we go that route. If agreed, I can add it as a different task on the parent issue tracking this work #1677

[bearomorphism]

@woile @noirbizarre @Lee-W
Could you review and merge this PR, so we can move on to other cz version related task? Thanks!

[woile]

There are some open conversations still