Re: [CEDET-devel] seems a bug

[Richard K. <em...@gm...>]

Peng,

I do not consider special handling of "new" word a bug.
Rather I consider it a feature.

"new" is a special word in C++ along with "delete", "class", "try",
"catch", "throw" and others.  Even though these may have no special
meaning in C, they should still not be used in C in my opinion,
primary because many C programs are either converted to C++ later or
many C programs are written such that they compile and work exactly
the same whether you throw C or C++ compiler at them.

I understand that some C programs will probably never have to deal
with C++. However in my personal experience over the past 15 to 20
years, I have converted almost all C programs that I had to modify and
maintain into C++ first.  First reason is that C++ compilers provide
much better checks than C compilers.  I have found that taking a C
program and merely trying to resolve all compile warnings and errors
that result when compiled as c++ results in finding several bugs not to
mention tightening up loose end.

Hence I would recommend that as much C programs be written to avoid
conflict with C++ especially if it is easy to do so.  In summary I
would ask that cedet NOT BE MODIFIED to allow "new" or any other C++
keywords to be treated as user defined names.



-----Original Message-----
From: peng yu <yup...@gm...>
To: ced...@li...
Subject: [CEDET-devel] seems a bug
Date: Fri, 21 May 2010 21:55:11 +0800

void test_func ( void )
{
    char new;
    new = 0;
}

When I move the cursor to "new" on the line "new = 0", and type M-x
semantic-ia-fast-jump,
semantic can not find the define of variable "new".
If I change the variable "new" to another name, it can work well.
"new" is a valid variable in C program, so it seems a bug.
------------------------------------------------------------------------------

_______________________________________________
Cedet-devel mailing list
Ced...@li...
https://lists.sourceforge.net/lists/listinfo/cedet-devel