fix/bars-disappear-on-zoom

[nstolpe]

Description

Addresses the issue where bar chart items more than 50% out of their chart area when zoomed disappear completely, as the bar middle point was being used to determine whether they should be culled from view.

Fixes # 2905

Type of Change

• [x] Bug fix (non-breaking change which fixes an issue)
• [ ] New feature (non-breaking change which adds functionality)
• [ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)
• [ ] This change requires a documentation update

How Has This Been Tested?

Checklist: (Feel free to delete this section upon completion)

• [ ] I have included a changeset if this change will require a version change to one of the packages.
• [x] I have performed a self-review of my own code
• [x] I have commented my code, particularly in hard-to-understand areas
• [ ] I have made corresponding changes to the documentation
• [x] I have run all builds, tests, and linting and all checks pass
• [ ] I have added tests that prove my fix is effective or that my feature works
• [x] My changes generate no new warnings
• [ ] Any dependent changes have been merged and published in downstream modules

Original video from issue:

https://github.com/user-attachments/assets/311b9725-00b7-4867-a539-90f9dc8e125e

Video after fix:

https://github.com/user-attachments/assets/965e120c-73ee-43bc-9923-33a666e56e85

[vercel[bot]]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Updated (UTC)
victory	✅ Ready (Inspect)	Visit Preview	Nov 18, 2024 9:14pm

[changeset-bot[bot]]

🦋 Changeset detected

Latest commit: 210b9b00a4cf20555b0e1c1b32c24fa665e3fd9b

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 30 packages

| Name | Type | | --------------------------- | ----- | | victory-bar | Patch | | victory-core | Patch | | victory-canvas | Patch | | victory-histogram | Patch | | victory | Patch | | victory-area | Patch | | victory-axis | Patch | | victory-box-plot | Patch | | victory-brush-container | Patch | | victory-brush-line | Patch | | victory-candlestick | Patch | | victory-chart | Patch | | victory-create-container | Patch | | victory-cursor-container | Patch | | victory-errorbar | Patch | | victory-group | Patch | | victory-legend | Patch | | victory-line | Patch | | victory-pie | Patch | | victory-polar-axis | Patch | | victory-scatter | Patch | | victory-selection-container | Patch | | victory-shared-events | Patch | | victory-stack | Patch | | victory-tooltip | Patch | | victory-voronoi-container | Patch | | victory-voronoi | Patch | | victory-zoom-container | Patch | | victory-native | Patch | | victory-vendor | Patch |

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[nstolpe]

@carbonrobot this fixes the issue in 2905, in a way that I don't think is going to conflict with anything else. There's a lot of jumping around in logic, from the actual component class, to its helper methods, into data, and into domain as well. I don't think this the best way to fix it in terms of code maintainability and understandability, but I think to have a better fix would require a larger refactor, which seems like something coming in the future. There might also be a slightly cleaner check than component.type.displayName, I'll take a look into that refactor once I get your thoughts and see if you'd definitely like to go this route.

[github-actions[bot]]

Folder/File	Previous size	New size	Difference
/packages/victory/dist/victory.min.js	385.59KB	385.76KB	+0.17KB (+0.04%)
TOTAL			+0.17KB

[nstolpe]

After seeing @nlkluth's solution here, I gave a try at subclassing VictoryBar and overriding renderData, getBaseProps and getCalculatedValues from the EventsMixin in the derived class to see if this could be accomplished without modifying internal packages (aside from exposing an export or 2). Unfortunately, the overrides got out of control, each additional one triggering an error that required another. I think what's here might be the best way to go at this point.