tr1::regex boilerplate

[Paolo Carlini]

Stephen M. Webb wrote:

>Hey all you all,
>
>I've managed to catch a little bit of breath during a Krunch Time That Never 
>Ends and put together a review patch of my original regex boilerplate and 
>some test cases for an initial review.  Please have at it.  Don't hold back.  
>Do your worst, I lost my pride years ago.
>  
>
:) I quickly applied the patch, fixed a couple of trivial issues (like 
"" in wstring constructors) and ran the regression tests on x86 and x86_64:

FAIL: tr1/7_regular_expressions/algorithms/search/char/1-ecma.cc 
execution test
FAIL: tr1/7_regular_expressions/algorithms/search/char/1-posix.cc 
execution test
FAIL: tr1/7_regular_expressions/algorithms/search/char/2-ecma.cc 
execution test
FAIL: tr1/7_regular_expressions/algorithms/search/char/2-posix.cc 
execution test
FAIL: tr1/7_regular_expressions/algorithms/search/char/3-ecma.cc 
execution test
FAIL: tr1/7_regular_expressions/algorithms/search/char/3-posix.cc 
execution test
FAIL: tr1/7_regular_expressions/regex/cons/char/c_string_extended_bad.cc 
execution test

Are those fails actually expected? I think so, I think unimplemented 
functions are involved, but wanted to be sure (maybe you wanted to test 
something else...) and in case xfail the tests for now and go ahead.

>Hopefully I'll be able to catch some free coding cycles to whip more of the 
>meat and potatoes of the parser and engine into a presentable form in the 
>next little while.
>
Please!

Thanks,
Paolo.