On Sat, May 22, 2010 at 5:15 PM, Ilia Alshanetsky wrote:
Felipe,
The idea behind the more complex logic was to allow people to have
Int/String/etc... class names, which if we make the types reserved keywords
will no longer be possible and break existing code.
Felipe,
The idea behind the more complex logic was to allow people to have
Int/String/etc... class names, which if we make the types reserved keywords
will no longer be possible and break existing code.
On Sat, May 22, 2010 at 11:03 AM, Felipe Pena wrote:
2010/5/22 Ilia Alshanetsky <ilia@prohost.org>
Felipe,
which are inclued in the {WHITESPACE}.
And I consider simpler to match just the type name, thus not needing the
yyless() logic. Hence I also suggested in the patch to remove the _HINT
suffix.
--
Regards,
Felipe Pena
2010/5/22 Ilia Alshanetsky <ilia@prohost.org>
Felipe,
Wouldn't the simpler change be to allow use of no spaces between scalar
hint and a variable? Basically change + to a *?
But it is not just about 0 or more tab/spaces, there is also \r, \n...hint and a variable? Basically change + to a *?
which are inclued in the {WHITESPACE}.
And I consider simpler to match just the type name, thus not needing the
yyless() logic. Hence I also suggested in the patch to remove the _HINT
suffix.
--
Regards,
Felipe Pena
--
Pierre
@pierrejoye | http://blog.thepimp.net | http://www.libgd.org