[BUG] Various inconsistencies in evennia/utils/text2html

text2html.py has various inconsistencies. They are demonstrated by skipped tests.

• When an ansi string contains a HILITE effect with a subsequent UNHILITE effect, the UNHILITE effect char seq is removed but the effect is applied until the end of the string. https://github.com/evennia/evennia/pull/1969#discussion_r338683863
• When an ansi string contains an effect followed by a NORMAL effect, the effect terminates at the correct location but the NORMAL effect char sequence is not removed from the string. https://github.com/evennia/evennia/pull/1969#discussion_r338905658, https://github.com/evennia/evennia/pull/1969#discussion_r338906034
• convert_urls doesn’t perform URL encoding correctly.

[Griatch]

No, we are not spending time on backporting to master at this time, except if it’s something really breaking. The develop branch will merge into master when 1.0 is releasing.

If we can no longer reproduce the issues reported here in develop branch, we’ll either close this, or mark it as Implemented so that it will close once it merges.

That said, I’m not sure the URL one in particular is addressed; but I was never sure exactly what @Henddher suggests need to change here.

[InspectorCaracal]

Do fixes like that generally get back-ported from develop into master?

Some of them! I’m not in charge of anything, but simple fixes for major issues often do. It would be pretty easy to just copy the current 1.0 parser into master, I think; it didn’t cause any breaking changes.

I think the project focus at the moment is more getting 1.0 ready to be moved to master wholesale so backporting fixes might be lower priority, but a PR to backport fixes like these would probably get merged quickly.