Hi all;
I'm working on "const-ifying" my GNU package, now that I've finally
taken the leap and abandoned support for K&R compilers.
One thing I've discovered is that there are a number of places in my
code where strings are written to ONLY to stick a nul character to end a
string, so that the code can then use the normal string functions like
strchr() etc.
In other words, a buffer and a length are passed to a function, and the
function will save the character at buf[length] then replace it with
'\0', then proceed to parse the buffer using str*() functions. After
it's done the function replaces the saved character.
I'd really like to change this so that it's not necessary to set '\0',
preferably without making a local copy of the buffer (for efficiency's
sake) and doing that means abandoning the standard str*() functions.
I could use the length, with strn*() functions (although many are
missing obviously). But the problem there is I need to keep updating
the length as I move through the buffer, which is a drag.
I'd much prefer to keep a pointer to the end of the buffer: this won't
change so it doesn't need to be managed. For example, something like:
char *strpchr(const char *str, const char *endp, int c);
or whatever.
Of course, there aren't any standard str*() or mem*() functions that I'm
aware of that use an end pointer instead of a length.
Are there any such functions in GLIB or similar? Do people know of
similar functions in other OS's etc.? If not, and I wanted to think
about creating some, what kind of naming convention would be
appropriate? I thought of strp*() but I don't know.
Just curious...
--
-------------------------------------------------------------------------------
Paul D. Smith <psmith@xxxxxxx> Find some GNU make tips at:
http://www.gnu.org http://make.paulandlesley.org
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
_______________________________________________
Autoconf mailing list
Autoconf@xxxxxxx
http://lists.gnu.org/mailman/listinfo/autoconf
