test failues with terminfo on solaris

[Mads Kiilerich]

Danek Duvall wrote, On 06/25/2011 01:33 AM:
> On Sat, Jun 25, 2011 at 01:14:34AM +0200, Mads Kiilerich wrote
>> I see some terminfo failures on solaris / SunOS 5.10.
>>
>> changeset:   13987:e0f07847f8de
>> user:        Danek Duvall<duvall at comfychair.org>
>> date:        Thu Apr 21 13:47:45 2011 -0700
>> summary:     color: add support for terminfo-based attributes and color
>>
>> caused
>>
>> --- /home/mads/hg/tests/test-status-color.t
>> +++ /home/mads/hg/tests/test-status-color.t.err
>> @@ -166,15 +166,16 @@
>>   hg status -A (with terminfo color):
>>
>>     $ TERM=xterm hg status --config color.mode=terminfo --color=always -A
>> -  \x1b(B\x1b[m\x1b[32m\x1b[1mA added\x1b(B\x1b[m (esc)
>> -  [ ...]
>> +  no terminfo entry for setab/setaf: reverting to ECMA-48 color
>> +  \x1b[0;32;1mA added\x1b[0m (esc)
>> +  [ ... ]
>>
>> Can this be fixed in the test or code or does it work as intended so
>> the test be disabled if not curses.tigetstr('setab') ?
> The fall-back to ECMA-48 color is intentional.  The hard part is finding a
> terminal that has the expected codes for setab/setaf.  The Solaris
> terminfo entry for xterm has no color.  I'd suggest xterm-color instead,
> but I don't know if the Linux and Solaris terminfo entries for xterm-color
> match, either (in both setab/setaf and sgr0).

I don't speak curses/terminfo, but setting TERM to xterm-color makes no 
difference here, both on solaris and linux. Does it work for you?

> Now, if the testsuite shipped its own terminfo entry and used that
> directly, it could be sure to get the same results on all platforms.

Fine with me if you think that is the best solution and can make it work ;-)

Can you explain why test-ui-color.py started failing too?

/Mads