Closed dritter closed 4 years ago
Are you sure it fixes #1267, #1291 and #1306? Can you reproduce these bugs without this PR and then verify that bugs no longer trigger with the PR?
Are you sure it fixes #1267, #1291 and #1306? Can you reproduce these bugs without this PR and then verify that bugs no longer trigger with the PR?
issue | broken on next |
fixed on fix-completion-menu |
---|---|---|
#1267 | yes | yes |
#1291 | yes | yes |
#1306 | yes | kinda |
Are you sure it fixes #1267, #1291 and #1306? Can you reproduce these bugs without this PR and then verify that bugs no longer trigger with the PR?
issue broken on
next
fixed onfix-completion-menu
1267 yes yes
1291 yes yes
1306 yes kinda
I may not be reading the code right but doesn't it simply remove support for P9K_RPROMPT_ON_NEWLINE=false
? If this is the goal, there is no need to look at transient_rprompt
and to hook zle widgets -- zsh will do the right thing automatically.
I may not be reading the code right but doesn't it simply remove support for P9K_RPROMPT_ON_NEWLINE=false?
D0h! Fixed three Issues and brought back another one.. 🙄 Checking for a fix. Thanks for the hint!
Moving in the right direction :+1:
You might want to link to https://www.reddit.com/r/zsh/comments/cgbm24/multiline_prompt_the_missing_ingredient/.
Btw. This is still not finished, as transient_rprompt with multiline prompt eats up one line.. Still investigating this.
Btw. This is still not finished, as transient_rprompt with multiline prompt eats up one line.. Still investigating this.
Does the same thing work in p10k? If it doesn't, I'll join in debugging.
Nah. It works in p10k. It's just a bit hard to read the generated code.. ;)
Nah. It works in p10k. It's just a bit hard to read the generated code.. ;)
Tell me about it :grin:
It's also a good idea to check that reset-prompt
doesn't eat (or create) a line. Hit esc-x
and type reset-prompt
. Many kinds of bugs can do this sort of badness.
Another test is to hit ctrl-r
and to verify that your cursor doesn't move. Same thing -- it's easy to break this.
Thanks for the tips. I'll double check that, once I am this far..
I think this is now done. @romkatv The checks you suggested pass on my machine.
Additionally this resolves issues with ZLE_RPROMPT_INDENT
, which we handle as 1 when set to zero.
@Syphdias I even reactivated testUsingUnsetVariables
. :) It would be super nice if you could retest this.
This PR is a backport from P10K and fixes several problems around multi line prompts.
Fixes #1267 , #1291 and #1306 .
Test are not green right now, because I am waiting for #1329 to be merged..