Re: safer way to use gnulib

[Ben Abbott]

On Tuesday, March 16, 2010, at 05:39PM, "Ben Abbott" <address@hidden> wrote:
>On Tuesday, March 16, 2010, at 04:51PM, "John W. Eaton" <address@hidden> wrote:
>>On 16-Mar-2010, Ben Abbott wrote:
>>
>>| #! /bin/sh -ev
>>| ## Use the flex from Fink's flex-devel package
>>| ##   The MacOS X flex will not work properly due to the patch below.
>>| ##   
>>http://www.opensource.apple.com/source/flex/flex-24.1/patches/scanEOF.diff
>>| export PREFIX="/sw"
>>| export PATH="$PREFIX/lib/flex/bin:$PATH"
>>| make
>>| make check
>>| --------------------------
>>
>>OK, from looking at that patch I guess that if your src/lex.cc file
>>has the lines
>>
>>  case EOB_ACT_END_OF_FILE:
>>          {
>>          if ( octave_wrap( ) )
>>                  return 0;
>>
>>instead of
>>
>>  case EOB_ACT_END_OF_FILE:
>>          {
>>          if ( octave_wrap( ) )
>>                  return EOF;
>>
>>then it looks like the version of flex that you used to generate
>>lex.cc is the one that has been modified with the bad patch.
>>
>>jwe
>
>You are correct. My lex.cc has the lines below.
>
>3259         case EOB_ACT_END_OF_FILE:
>3260           {
>3261           if ( octave_wrap( ) )
>3262             return 0;
>
>By "bad patch", I assume you refer to Apple's patch? Which would mean that I 
>created the lex.cc using Apple's flex and not the proper one from Fink.
>
>I've deleted lex.cc and begun my build again. Now when I look at lex.cc, I see 
>..
>
>3273         case EOB_ACT_END_OF_FILE:
>3274           {
>3275           if ( octave_wrap( ) )
>3276             return EOF;
>
>Good catch on your part ... and a sloppy fumble on mine ;-)
>
>Thanks! ... I'll report back when the build is done.
>
>Ben

The dependencies of lex.cc aren't too many, so it finished quickly. Except for 
the one failure I always get with data.cc, all tests pass.

Thanks again!

Ben