Actually, the return values throughout that function are all bogus...
Don't worry about reverting (for 4.3).
I'll add this to my TODO for PHP 5.
--Wez.
On Sat, 15 Feb 2003, Wez Furlong wrote:
> This is a bogus patch; the return value of the php_stream_copy_to_XXX
> functions are the number of bytes that were copied, and this patch
> breaks that.
>
> The correct fix for #22234 is to fix php_copy_file() to handle a 0
> return from this function.
>
> Please revert!
>
> --Wez.
>
> On Sat, 15 Feb 2003, Moriyoshi Koizumi wrote:
>
> > moriyoshi Sat Feb 15 14:56:13 2003 EDT
> >
> > Modified files:
> > /php4/main streams.c
> > Log:
> > Fixed bug #22234
> >
> >
> > Index: php4/main/streams.c
> > diff -u php4/main/streams.c:1.154 php4/main/streams.c:1.155
> > --- php4/main/streams.c:1.154 Thu Feb 13 16:08:04 2003
> > +++ php4/main/streams.c Sat Feb 15 14:56:12 2003
> > @@ -20,7 +20,7 @@
> > +----------------------------------------------------------------------+
> > */
> >
> > -/* $Id: streams.c,v 1.154 2003/02/13 21:08:04 wez Exp $ */
> > +/* $Id: streams.c,v 1.155 2003/02/15 19:56:12 moriyoshi Exp $ */
> >
> > #define _GNU_SOURCE
> > #include "php.h"
> > @@ -1206,6 +1206,17 @@
> > /* fall through - we might be able to copy in smaller chunks */
> > }
> > #endif
> > +
> > + {
> > + php_stream_statbuf sbuf;
> > + if (php_stream_stat(src, &sbuf TSRMLS_CC) == 0) {
> > + /* in the event that the source file is 0 bytes, return 1 to
>indicate success
> > + * because opening the file to write had already created a
>copy */
> > + if (sbuf.sb.st_size == 0) {
> > + return 1;
> > + }
> > + }
> > + }
> >
> > while(1) {
> > readchunk = sizeof(buf);
> >
> >
> >
> > --
> > PHP CVS Mailing List (http://www.php.net/)
> > To unsubscribe, visit: http://www.php.net/unsub.php
> >
> >
> >
>
>
> --
> PHP CVS Mailing List (http://www.php.net/)
> To unsubscribe, visit: http://www.php.net/unsub.php
>
>
>
--
PHP CVS Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
