Re: [PATCH] Add support for mkostemp, mkstemps and mkostemps

["Anthony G. Basile" <basile@opensource.dyc.edu>]

Okay a few points about my last patch.

1. I removed lots of unnecessary headers.

2. I tested using:

	strcpy(t, "aaaXXXXXX");
	mktemp(t);
	printf("mktemp = %s\n", t);

	strcpy(t, "aaaXXXXXX");
	fd = mkstemp(t);
	printf("mkstemp = %s\n", t);
	report_close(fd);
	//unlink(t);

	strcpy(t, "aaaXXXXXX");
	fd = mkostemp(t, O_WRONLY|O_CLOEXEC);
	printf("mkostemp = %s\n", t);
	report_close(fd);
	//unlink(t);

	strcpy(t, "aaaXXXXXXzzz");
	fd = mkstemps(t, 3);
	printf("mkstemps = %s\n", t);
	report_close(fd);
	//unlink(t);

	strcpy(t, "aaaXXXXXXzzz");
	fd = mkostemps(t, 3, O_WRONLY|O_CLOEXEC);
	printf("mkostemps = %s\n", t);
	report_close(fd);
	//unlink(t);


3. I went with Szabolcs' function to generate entropy:

	for (i=0; i<6; i++, r>>=5)
		template[i] = 'A'+(r&15)+(r&16)*2;

I think we can commit this (baring any other problems) and think about 
improving it in the future.  As I said before, uclibc/glibc give 
filenames which include:

static const char letters[] =
"abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789";

while 'A'+(r&15)+(r&16)*2 gives A-P a-p. So the space of names for the 
former is 62^6 while for the latter is 32^6.



-- 
Anthony G. Basile, Ph. D.
Chair of Information Technology
D'Youville College
Buffalo, NY 14201
(716) 829-8197