develooper Front page | perl.perl5.porters | Postings from November 2011

Re: On dual-living Unicode::GCString in v5.16

Thread Previous | Thread Next
From:
Karl Williamson
Date:
November 1, 2011 21:43
Subject:
Re: On dual-living Unicode::GCString in v5.16
Message ID:
4EB0C9E4.2030805@khwilliamson.com
On 11/01/2011 06:27 PM, Craig A. Berry wrote:
> On Tue, Nov 1, 2011 at 4:24 PM, Eric Brine<ikegami@adaelis.com>  wrote:
>> On Mon, Oct 31, 2011 at 8:37 PM, Brian Fraser<fraserbn@gmail.com>  wrote:
>>>
>>> Sorry, kind of having to skip reading this since my internet will be down
>>> soonish, but something important I want to bring up is that
>>> Unicode::LineBreak (and thus GCString) fails to build on non-cygwin Windows.
>>> Please please someone fix that before including it in the core.
>>
>> About that, the author says
>> [https://rt.cpan.org/Public/Bug/Display.html?id=65693]:
>>
>> I don't have any plan to support:
>> * NMAKE + CL + LIB
>> * CMD.EXE
>>
>> So people trying to build and to install this module on Windows should, of
>> my limited knowledge, install:
>> * Cygwin, or
>> * MinGW and MSYS.
>>
>> Note that mingw alone doesn't help. The problem relates to the installation
>> of the "Sombok" library, a dependency included in the distro.
>
> But who needs bash when you have Perl, and who needs autotools when
> you have %Config?  The config.h for sambok has 34 defines, most of
> which are the usual suspects from Perl's configuration:
>
> <http://cpansearch.perl.org/src/NEZUMI/Unicode-LineBreak-2011.010_26/sombok/config.h.in>
>
> The Makefile.PL does not look very portable:
>
> <http://cpansearch.perl.org/src/NEZUMI/Unicode-LineBreak-2011.010_26/Makefile.PL>
>
> but nor does it look hugely evil, and --- perhaps more importantly --
> not huge.  Revising it to generate a config.h for sambok based on what
> Perl already knows about the current system may not be that bad.
>
> Of course the proof is in the pudding and lots of problems could come
> to light after creating an alternate configuration method.  But if
> it's not doing anything wacky with IPC, permissions, filenames, and
> other OS-specific nonsense, we shouldn't assume it's difficult to port
> without trying.
>

I have not looked at the code nor the test suite.  But, Unicode does 
furnish test suites for things like Line Breaking and Graphemes.  The 
latter is part of unicore/TestProp.pl to test \X regex matching.  (The 
relevant subroutine (and calls) is Test_X.

Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About