Kevin Ballard [off-list ref] writes:
Enhance the test_decode_color function to handle all common color codes,
including background colors and escapes that contain multiple codes.
This change necessitates changing <WHITE> to <BOLD>, so update t4034
as well.
This change is necessary for the next commit in order to test
background colors properly.
Signed-off-by: Kevin Ballard <redacted>
---
I turned sed into awk. Looks like awk is already used elsehwere in git,
so I'm assuming this is safe, but please let me know if it's not.
I think calling BOLD BOLD is the right thing to do (who came up with the
bogus WHITE in the first place anyway---my terminal is black letters on
white background, thank you).
Even though some scripts seem to already use awk, they are all used for
very small and trivial processing without exercising anything remotely
fancy e.g. hexadecimal \xXX quoting or match() function, so I wouldn't be
surprised if we see breakage reports from minority platforms.
But I do not think of a trivial way to express combination of attributes
by extending the existing sed script (we can write loops and do the same
computation as your awk script does, but it does not reduce the complexity
nor risk of portability issues), so let's see what happens. We already
use Perl everywhere, which we might end up using for this if there are
platforms that have issues with your awk script.
Thanks.
On Oct 20, 2010, at 4:40 PM, Junio C Hamano wrote:
Kevin Ballard [off-list ref] writes:
quoted
Enhance the test_decode_color function to handle all common color codes,
including background colors and escapes that contain multiple codes.
This change necessitates changing <WHITE> to <BOLD>, so update t4034
as well.
This change is necessary for the next commit in order to test
background colors properly.
Signed-off-by: Kevin Ballard <redacted>
---
I turned sed into awk. Looks like awk is already used elsehwere in git,
so I'm assuming this is safe, but please let me know if it's not.
I think calling BOLD BOLD is the right thing to do (who came up with the
bogus WHITE in the first place anyway---my terminal is black letters on
white background, thank you).
Even though some scripts seem to already use awk, they are all used for
very small and trivial processing without exercising anything remotely
fancy e.g. hexadecimal \xXX quoting or match() function, so I wouldn't be
surprised if we see breakage reports from minority platforms.
Entirely possible, but I'm hoping that's not the case. I used this against
BSD awk version 20070501, so there's nothing remotely fancy there, and the
regular expression conforms to the subset of BRE's mentioned in
CodingGuidelines. But as you said, it's possible that minority platforms
don't handle this correctly.
But I do not think of a trivial way to express combination of attributes
by extending the existing sed script (we can write loops and do the same
computation as your awk script does, but it does not reduce the complexity
nor risk of portability issues), so let's see what happens. We already
use Perl everywhere, which we might end up using for this if there are
platforms that have issues with your awk script.
I considered writing this in Perl, but my complete lack of knowledge of Perl
made that a non-starter. However if there are any problems with the awk
script then I would welcome a Perl rewrite by someone who actually knows the
language.
-Kevin Ballard