Replace CSS Variables in favor of theme variable for css styling

[AllanOXDi]

Summary

This PR refactors the SafeHTML rendering to compute and apply theme colors styles instead of using CSS variables (var()).
closes #13851

References

#13851

Reviewer guidance

• Verify table rendering with proper theme colors (borders, caption, thead/tfoot backgrounds)
• Verify image rendering with proper border color
• Test in browsers with limited CSS variable support
• Confirm no visual regressions

[rtibbles]

One small bit of cleanup, but this is looking correct otherwise!

[github-actions]

Build Artifacts

Asset type	Download link
PEX file	kolibri-0.19.1rc0.dev0_git.9.g64807fb8.pex
Windows Installer (EXE)	kolibri-0.19.1rc0.dev0+git.9.g64807fb8-windows-setup-unsigned.exe
Debian Package	kolibri_0.19.1rc0.dev0+git.9.g64807fb8-0ubuntu1_all.deb
Mac Installer (DMG)	kolibri-0.19.1rc0.dev0+git.9.g64807fb8.dmg
Android Package (APK)	kolibri-0.19.1rc0.dev0+git.9.g64807fb8-0.1.6-debug.apk
Raspberry Pi Image	kolibri-pi-image-0.19.1rc0.dev0+git.9.g64807fb8.zip
TAR file	kolibri-0.19.1rc0.dev0+git.9.g64807fb8.tar.gz
WHL file	kolibri-0.19.1rc0.dev0+git.9.g64807fb8-py2.py3-none-any.whl

[rtibbles]

This looks like the right way to do it, without drilling deeper into the component tree. Just manual QA to confirm no regressions and this will be good to go.

Changes have been made!

[pcenov]

Hi @AllanOXDi - the only issue I noticed in Chrome, Firefox and Safari is in 'All About Hummingbirds (long article)' where under the images there's the following text: tag was not properly removed by the sanitizer.

See the following video where I'm comparing the current version against the official 0.19.1 Kolibri build:

sanitizer.mp4

Note that I'm not seeing this issue in Microsoft's Edge browser where everything is working fine.

[rtibbles]

Interesting - sanitization should be happening before it gets passed to the code that has been changed here. Do you see this behaviour in 0.19.1 on Chrome as well?

[pcenov]

Hmmm, after additional investigation it turned out that I had been comparing it against the official Kolibri 0.19 build. When I updated the VM to Kolibri 0.19.1 I can see the same behaviour there as well so it has nothing to do with the changes made in this PR:

0.19.vs.0.19.1.mp4

[rtibbles]

Hrm... could you open a follow up issue for it @pcenov as this feels like a regression (albeit a minor one) from 0.19.0?

[rtibbles]

Nothing blocking from QA and code changes look good.