Sascha Schumann [Sun, 5 May 2002 02:07:16 +0000 (02:07 +0000)]
the output buffer is flushed during request shutdown before it
reaches our rshutdown, so it is safe to delete the contents of
the url rewriter variables here.
Zeev Suraski [Sat, 4 May 2002 23:23:16 +0000 (23:23 +0000)]
We shouldn't force Content-length:, but much worse, it's wrong in many
cases (basically, unless you use output buffering to buffer all of your content,
it won't work; output buffering is used for lots of different things as a
functional interface, and such buffers have no relation to Content-type at
all).
Stig Venaas [Sat, 4 May 2002 14:27:48 +0000 (14:27 +0000)]
Reworked result resource handling so that result is not freed until all
its result entry resources are freed
@- Fixed resource bug in LDAP extension. (Stig Venaas)
Yasuo Ohgaki [Sat, 4 May 2002 03:02:51 +0000 (03:02 +0000)]
Fixed crash when buffer is over written in ob callback. (Bug $ 12227)
Added check current output function check for Centent-Length header.
@ Fixed output buffer crash when buffer is over written in callback. (Yasuo)
Stig Bakken [Fri, 3 May 2002 13:13:51 +0000 (13:13 +0000)]
* switched from static to instance model
* added matching method with glob support, some examples:
$os->matchSignature('linux');
$os->matchSignature('linux-2.4*);
$os->matchSignature('linux-*-i?86');
@ - Added output_add_rewrite_var() and output_remove_rewrite_var() to inject
@ and remove variables from the URL-Rewriter. (thies)
i have also modified the session module to use this - so it doesn't
need to fiddle with the output-system any more
Stig Bakken [Fri, 3 May 2002 05:42:55 +0000 (05:42 +0000)]
* fixed autoconf vs. pear installer replacement issue in scripts/pear.in
(pear installer now substitutes "@prefix@/bin" to bin_dir ;-)
* added skeleton for OS_Guess class
Andi Gutmans [Thu, 2 May 2002 17:20:48 +0000 (17:20 +0000)]
Initial support for built-in backtracing.
There are still a few problems such as includes and calling other functions
from internal functions which aren't seen (will have to think if and how to
fix this).
Also the main scripts filename isn't available. Need to think about that.
Yasuo Ohgaki [Thu, 2 May 2002 13:55:05 +0000 (13:55 +0000)]
Change nest level to send Content-Length again. It seems this is the
best setting for now.
PHP will not send Content-Length always. It only sends when it is
possible to send. output_buffer=0 is supposed to disable chunked
output, but it seems it does not disable. It also behaves a litte
strange way. This should be addressed someday.
It is possible Content-Length header is not set. If it happens, try to
increase chunk size for now. (i.e. output_buffer=40960)
I included a little debug code for me and other develpers to play with,
when (NestLevel==1 && ObStatus==5), PHP sends Content-Length.
Jon Parise [Thu, 2 May 2002 05:36:21 +0000 (05:36 +0000)]
Revert the previous commit (revision 1.40).
@bindir@ ended up becoming '${exec_prefix}' (literally) on my system
(FreeBSD 4.5), thus producing a broken shebang line.
@prefix@ works fine for me here. If @bindir@ is indeed preferred, the
substitution problem noted above needs to be resolved before the change
is reapplied.
Yasuo Ohgaki [Thu, 2 May 2002 01:12:01 +0000 (01:12 +0000)]
Move Content-Length: header handling from zlib.c to output.c
When output buffer is enabled and header can be sent, Content-Length:
header is added always from now on.
Markus Fischer [Wed, 1 May 2002 10:41:10 +0000 (10:41 +0000)]
- Fix couple of problems with socket_create_pair():
- Force fourth argument to be passed by reference
- Since the argument is modified there is no need to force it to be an array
since it's destroyed anyway
- Only modify the argument if socketpair() was successfully
- Fix string modified for error message message
- Set global last_error when socketpair() fails