fixinc/inclhack.def (avoid_wchar_t_type): New fix

Bruce Korb bkorb@cruzio.com
Sat Nov 11 16:51:00 GMT 2000


Mark Mitchell wrote:
> 
> >>>>> "Bruce" == Bruce Korb <bkorb@cruzio.com> writes:
> 
>     Bruce> Hi Mark,
> 
>     Bruce> Most likely, somebody, somewhere is going to end the
>     Bruce> wchar_t typedef line with an incomplete comment.  The
>     Bruce> result will be that the ``#endif'' will be inside that
>     Bruce> comment.  I recommend that you delete the second
>     Bruce> `c_fix_arg' and let fixincl default to using the select
>     Bruce> clause.  That will put the comment on the end of the #endif
>     Bruce> line where it will do no harm.  :-)
> 
> I don't quite understand the full wonders of fixincludes. :-(

You're pretty close, though :-)

> Perhaps you would submit a patch?  Since I simply ripped off the bool
> fix above, I assume that fix has the same problem?

Yes, it does.  (Or, rather, would if someone had such a comment :)
Needless to say, I have stubbed my toe on this before....

--- inclhack.ori	Sat Nov 11 17:47:05 2000
+++ inclhack.def	Sat Nov 11 17:48:27 2000
@@ -764,9 +764,8 @@
 
     c_fix     = format;
     c_fix_arg = "#ifndef __cplusplus\n%0\n#endif";
-    c_fix_arg = "^[ \t]*typedef[ \t].*[ \t]bool[ \t]*;.*";
 
-    test_text = "typedef unsigned int\tbool \t; /* bool type */";
+    test_text = "typedef unsigned int\tbool \t; /* bool\n type */";
 };
 
 /*
@@ -781,9 +780,8 @@
 
     c_fix     = format;
     c_fix_arg = "#ifndef __cplusplus\n%0\n#endif";
-    c_fix_arg = "^[ \t]*typedef[ \t].*[ \t]wchar_t[ \t]*;.*";
 
-    test_text = "typedef unsigned short\twchar_t \t; /* wchar_t type */";
+    test_text = "typedef unsigned short\twchar_t \t; /* wchar_t\n type
*/";
 };
 
 /*


More information about the Gcc-patches mailing list