[PATCH] Make sure terminfo color tests work the same on all platforms

[Mads Kiilerich]

Danek Duvall wrote, On 06/26/2011 02:30 AM:
> # HG changeset patch
> # User Danek Duvall<duvall at comfychair.org>
> # Date 1309048141 25200
> # Node ID 03a164145bcb6b99aff61a14b563ad53ae52a2df
> # Parent  13011330885637d05262f2abceb8098fe4b8884b
> Make sure terminfo color tests work the same on all platforms.

(btw: The subject line should according to ContributingChanges start 
with 'tests:' and not end with '.'. A short description of what problem 
this solves and why it was solved this way would be nice too.)

Yes, this works - at least on one solaris and one linux machine. Can 
someone confirm that it works on other platforms too, such as OS/X and 
other BSDs and AIX?

Is 'tic' always available on all (test-)supported platforms?

I am a bit worried whether it is relevant to run the test in such an 
artificial environment if we know that the environment users have is 
different and not good enough. I guess the answer is that terminfo 
ensures that all environments are good enough even though they are 
different. But then why do we issue a warning?

> diff --git a/tests/test-status-color.t b/tests/test-status-color.t
> --- a/tests/test-status-color.t
> +++ b/tests/test-status-color.t
> @@ -167,7 +167,9 @@ hg status -A:
>
>   hg status -A (with terminfo color):
>
> -  $ TERM=xterm hg status --config color.mode=terminfo --color=always -A
> +  $ mkdir $TESTTMP/terminfo
> +  $ TERMINFO=$TESTTMP/terminfo tic $TESTDIR/xterm-color.ti
> +  $ TERM=xterm-color TERMINFO=$TESTTMP/terminfo hg status --config color.mode=terminfo --color=always -A

Would it make sense to rename xterm-color to something else so it isn't 
confused with the 'real' xterm-color that is available on most platforms?

/Mads