Bug #52373 [Opn]: XPath returns inconsistent results

2010-07-19 Thread david at sickmiller dot com
Edit report at http://bugs.php.net/bug.php?id=52373&edit=1

 ID:   52373
 User updated by:  david at sickmiller dot com
 Reported by:  david at sickmiller dot com
 Summary:  XPath returns inconsistent results
 Status:   Open
 Type: Bug
 Package:  DOM XML related
 Operating System: Cent OS 5
 PHP Version:  5.2.13

 New Comment:

I wonder if this may be related to the libxml2 library.  I am using
v2.6.26-2.1.2.8 from the CentOS distribution.


Previous Comments:

[2010-07-19 07:59:35] m dot philipp at coreto dot de

Cannot repeat with php 5.3.2.

All test runs resulted in: 

"In 1 instances, found 1 match(es) for
/root[1]/fie...@attr1='val1']"


[2010-07-19 01:37:51] david at sickmiller dot com

Description:

Sometimes the DOMXpath->evaluate() function returns unreliable results. 
This seems to happen when the document has at least a depth of 3 and
selectors are used at multiple levels.  There may be other conditions
that trigger it; I'm not sure.  The test script below is consistently
showing errors for me.

Test script:
---
';

$rawxpath = "/root[1]/fie...@attr1='val1']";



$results = array();

for ($i = 0; $i < 1; $i++) {

$xmlDocument = new DOMDocument();

$xmlDocument->loadXML($rawxml);

$xp = new DOMXPath($xmlDocument);

$xpr = $xp->evaluate($rawxpath, $xmlDocument->documentElement);



if (isset($results[(string) $xpr->length]))

$results[(string) $xpr->length]++;

else

$results[(string) $xpr->length] = 1;

}



foreach ($results as $matches => $count)

printf("In %5d instances, found $matches match(es) for
$rawxpath\n", $count);

Expected result:

In 1 instances, found 1 match(es) for /root[1]/fie...@attr1='val1']

Actual result:
--
The exact numbers seem to vary, but I get roughly this response:





In  9924 instances, found 1 match(es) for /root[1]/fie...@attr1='val1']

In76 instances, found 2 match(es) for /root[1]/fie...@attr1='val1']






-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52373&edit=1


Req #52362 [Opn->Wfx]: give str_ireplace an option to refer the original string

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52362&edit=1

 ID:  52362
 Updated by:  ahar...@php.net
 Reported by: emiobe at guehring dot de
 Summary: give str_ireplace an option to refer the original string
-Status:  Open
+Status:  Wont fix
 Type:Feature/Change Request
 Package: *General Issues
 PHP Version: 5.3.2

 New Comment:

The point of having the separate str_replace and str_ireplace functions

is that they're intended to be fast replacers for times when you don't

need any special features such as regular expression matching or back

references. This would introduce special case code that would only slow

them down and duplicate functionality the preg functions provide.


Previous Comments:

[2010-07-16 14:31:32] emiobe at guehring dot de

erm, just replace expected and actual result... I have mixed it up,
sorry


[2010-07-16 14:21:28] giorgio dot liscio at email dot it

use preg_replace with /i PCRE_CASELESS identifier



bogus


[2010-07-16 14:13:21] emiobe at guehring dot de

Description:

Currently I can search case-insensitive with str_ireplace. But when I
need the original string in the replacement I have to use
preg_replace//i. 



It would be nice to have the original text in a "magic variable" like $1
to avoid using the "expensive" preg-engine.

Test script:
---
currently:



print str_ireplace ("original", "original", "the ORIGINAL
text");



idea:



print str_ireplace ("original", "$1", "the ORIGINAL text");



Expected result:

the original text

Actual result:
--
the ORIGINAL text






-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52362&edit=1


Bug #52359 [Opn->Bgs]: http_build_query seem to encode array incorrectly

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52359&edit=1

 ID:   52359
 Updated by:   ahar...@php.net
 Reported by:  jj at danskscanning dot dk
 Summary:  http_build_query seem to encode array incorrectly
-Status:   Open
+Status:   Bogus
 Type: Bug
 Package:  URL related
 Operating System: Windows XP SP3
 PHP Version:  Irrelevant

 New Comment:

http_build_query() encodes the numeric keys in the output. This is

expected (and documented) behaviour.


Previous Comments:

[2010-07-17 23:26:03] k.schroe...@php.net

Automatic comment from SVN on behalf of k.schroeder
Revision: http://svn.php.net/viewvc/?view=revision&revision=301354
Log: Test for #52359


[2010-07-16 11:25:04] jj at danskscanning dot dk

Description:

I was working in VS.php with a CodeIgniter Project (5.2) and have not
made any 

changes to the standard setup.

Test script:
---
http://bugs.php.net/bug.php?id=52359&edit=1


[PHP-BUG] Bug #52376 [NEW]: opendir() cannot open UNC paths in IIS7 using passthrough auth.

2010-07-19 Thread ryan at kisc dot edu dot np
From: 
Operating system: win64 - W2008R2
PHP version:  5.3.2
Package:  *Directory/Filesystem functions
Bug Type: Bug
Bug description:opendir() cannot open UNC paths in IIS7 using passthrough auth.

Description:

This seems to be identical to #50542 but that one was closed and I couldn't
do anything to it.



Basically it was marked closed because with passthrough-authentication
essentially disabled by putting a username in "physical path credentials"
and selecting ClearText in the "Physical Path Credentials Logon Type" both
in advanced settings, this can be made to work so long as that
configuration is acceptable.



I too am having trouble with this and unfortunately the solution of
editing

physical path credentials breaks the pass through authentication to the

database, ldap, and other things. This all worked on 5.3.1 with IIS6 on
W2K3 but

on IIS7 with W2K8R2 no joy. So I upgraded to 5.3.2, still no joy. My
application is so reliant on the pass-through authentication that when I do
arrange things such that opendir works (clear-text + specific physical path
credentials) the system becomes a single user system!



So I can confirm the workaround mentioned for #50542 works if the
environment allows for it, but mine does not. Any other ideas? If I connect
to IIS using the environment requirement of network type auth and
pass-through authentication it does NOT work, though it really should. If I
use that exact same user in the "physical path credentials" field and set
"clear text" opendir suddenly works. Please don't close this as it really
is a bug and apparently specific to IIS7

Test script:
---


Expected result:

I expect to get no errors.





Actual result:
--
Warning:
opendir(//someserver/someshare/somefolder,//someserver/someshare/somefolder)
[function.opendir]: Access is denied. (code: 5) in
C:\inetpub\wwwroot\testdir.php on line 2



Warning: opendir(//someserver/someshare/somefolder) [function.opendir]:
failed to open dir: No such file or directory in
C:\inetpub\wwwroot\testdir.php on line 2

-- 
Edit bug report at http://bugs.php.net/bug.php?id=52376&edit=1
-- 
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52376&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52376&r=trysnapshot53
Try a snapshot (trunk):  
http://bugs.php.net/fix.php?id=52376&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52376&r=fixed
Fixed in SVN and need be documented: 
http://bugs.php.net/fix.php?id=52376&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52376&r=alreadyfixed
Need backtrace:  
http://bugs.php.net/fix.php?id=52376&r=needtrace
Need Reproduce Script:   
http://bugs.php.net/fix.php?id=52376&r=needscript
Try newer version:   
http://bugs.php.net/fix.php?id=52376&r=oldversion
Not developer issue: 
http://bugs.php.net/fix.php?id=52376&r=support
Expected behavior:   
http://bugs.php.net/fix.php?id=52376&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=52376&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=52376&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52376&r=globals
PHP 4 support discontinued:  http://bugs.php.net/fix.php?id=52376&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52376&r=dst
IIS Stability:   
http://bugs.php.net/fix.php?id=52376&r=isapi
Install GNU Sed: 
http://bugs.php.net/fix.php?id=52376&r=gnused
Floating point limitations:  
http://bugs.php.net/fix.php?id=52376&r=float
No Zend Extensions:  
http://bugs.php.net/fix.php?id=52376&r=nozend
MySQL Configuration Error:   
http://bugs.php.net/fix.php?id=52376&r=mysqlcfg



Bug #52317 [Asn->Sus]: Segmentation fault when using mail() on a rhel 4.x (only 64 bit)

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52317&edit=1

 ID:   52317
 Updated by:   ahar...@php.net
 Reported by:  om at wysiwyg dot de
 Summary:  Segmentation fault when using mail() on a rhel 4.x
   (only 64 bit)
-Status:   Assigned
+Status:   Suspended
 Type: Bug
 Package:  Reproducible crash
 Operating System: rhel4.8_x86_64
 PHP Version:  5.2SVN-2010-07-12 (snap)
 Assigned To:  aharvey

 New Comment:

Gah, missing header inclusion + older version of gcc = failure, hence

why this wasn't manifesting on RHEL 5 or recent Ubuntu versions. I do

find the fact it's 64-bit only interesting, 



Fix checked into trunk. Given we're deep into the RC cycle, I'll talk

to the RMs before committing to the 5.2 and 5.3 branches, although I

think this is a pretty good candidate for both.


Previous Comments:

[2010-07-19 12:46:46] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301392
Log: Fix for bug #52317 (Segmentation fault when using mail() on a rhel
4.x (only 64
bit)).


[2010-07-15 10:32:42] om at wysiwyg dot de

(gdb) bt

#0  0x003fc0571360 in strlen () from /lib64/tls/libc.so.6

#1  0x003fc0542e79 in vfprintf () from /lib64/tls/libc.so.6

#2  0x003fc0548626 in fprintf () from /lib64/tls/libc.so.6

#3  0x0060b902 in php_mail (to=0x2a9557f208 "o...@wysiwyg.de",
subject=0x2a9557efc0 "the subject", 

message=0x2a95580b28 "hello", headers=0x9557f388 , 

extra_cmd=0x0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:281

#4  0x0060b6dc in zif_mail (ht=4, return_value=0x2a9557f318,
return_value_ptr=0x0, this_ptr=0x0, 

return_value_used=0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:180

#5  0x006ee2cb in zend_do_fcall_common_helper_SPEC
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:200

#6  0x006f3bdb in ZEND_DO_FCALL_SPEC_CONST_HANDLER
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:1740

#7  0x006eddc1 in execute (op_array=0x2a9557e130)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:92

#8  0x006c747f in zend_execute_scripts (type=8, retval=0x0,
file_count=3)

at /usr/local/src/php5.2-201007131430/Zend/zend.c:1134

#9  0x00671979 in php_execute_script
(primary_file=0x7fb890)

at /usr/local/src/php5.2-201007131430/main/main.c:2036

#10 0x00741dfb in main (argc=2, argv=0x7fb9e8)

at /usr/local/src/php5.2-201007131430/sapi/cli/php_cli.c:1165

(gdb)


[2010-07-15 10:15:59] om at wysiwyg dot de

sorry I forgot the testscript:



 seg. fault

mail($to, $subject, $message, $headers);



?>


[2010-07-13 18:26:07] ahar...@php.net

Since I committed the offending patch, I'll look into it.



If you could provide both the test script (which doesn't seem to have
made it) and, if possible, a backtrace per the instructions [1], that
would be most helpful.



[1] http://bugs.php.net/bugs-generating-backtrace.php


[2010-07-12 16:52:53] om at wysiwyg dot de

Description:

Since the patch for "headers_trimmed" was added in the mail.c I always
get a segmentation fault when using mail(). This only affects rhel4.x 64
bit os. rhel 5.x 64 bit or rhel4.x 32 bit works fine.



Using mail.c from a snapshot before 2010-04-22 in the actual snapshot
source everything works fine.



--- php/php-src/branches/PHP_5_2/ext/standard/mail.c2010-04-22 01:07:48
UTC (rev 298290) -> works fine

+++ php/php-src/branches/PHP_5_2/ext/standard/mail.c2010-04-22 02:22:49
UTC (rev 298291) -> segmentation fault when using mail() in a php script
on a rhel4.x 64 bit host









Test script:
---
running "make test" after building php from source.



the provided test script also reports the problem.







-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52317&edit=1


Bug #52317 [Sus]: Segmentation fault when using mail() on a rhel 4.x (only 64 bit)

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52317&edit=1

 ID:   52317
 Updated by:   ahar...@php.net
 Reported by:  om at wysiwyg dot de
 Summary:  Segmentation fault when using mail() on a rhel 4.x
   (only 64 bit)
 Status:   Suspended
 Type: Bug
 Package:  Reproducible crash
 Operating System: rhel4.8_x86_64
 PHP Version:  5.2SVN-2010-07-12 (snap)
 Assigned To:  aharvey

 New Comment:

Editing fail: ignore "I do find the fact it's 64-bit only
interesting,".



It's not that interesting when you think about it for two seconds. :)


Previous Comments:

[2010-07-19 12:46:51] ahar...@php.net

Gah, missing header inclusion + older version of gcc = failure, hence

why this wasn't manifesting on RHEL 5 or recent Ubuntu versions. I do

find the fact it's 64-bit only interesting, 



Fix checked into trunk. Given we're deep into the RC cycle, I'll talk

to the RMs before committing to the 5.2 and 5.3 branches, although I

think this is a pretty good candidate for both.


[2010-07-19 12:46:46] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301392
Log: Fix for bug #52317 (Segmentation fault when using mail() on a rhel
4.x (only 64
bit)).


[2010-07-15 10:32:42] om at wysiwyg dot de

(gdb) bt

#0  0x003fc0571360 in strlen () from /lib64/tls/libc.so.6

#1  0x003fc0542e79 in vfprintf () from /lib64/tls/libc.so.6

#2  0x003fc0548626 in fprintf () from /lib64/tls/libc.so.6

#3  0x0060b902 in php_mail (to=0x2a9557f208 "o...@wysiwyg.de",
subject=0x2a9557efc0 "the subject", 

message=0x2a95580b28 "hello", headers=0x9557f388 , 

extra_cmd=0x0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:281

#4  0x0060b6dc in zif_mail (ht=4, return_value=0x2a9557f318,
return_value_ptr=0x0, this_ptr=0x0, 

return_value_used=0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:180

#5  0x006ee2cb in zend_do_fcall_common_helper_SPEC
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:200

#6  0x006f3bdb in ZEND_DO_FCALL_SPEC_CONST_HANDLER
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:1740

#7  0x006eddc1 in execute (op_array=0x2a9557e130)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:92

#8  0x006c747f in zend_execute_scripts (type=8, retval=0x0,
file_count=3)

at /usr/local/src/php5.2-201007131430/Zend/zend.c:1134

#9  0x00671979 in php_execute_script
(primary_file=0x7fb890)

at /usr/local/src/php5.2-201007131430/main/main.c:2036

#10 0x00741dfb in main (argc=2, argv=0x7fb9e8)

at /usr/local/src/php5.2-201007131430/sapi/cli/php_cli.c:1165

(gdb)


[2010-07-15 10:15:59] om at wysiwyg dot de

sorry I forgot the testscript:



 seg. fault

mail($to, $subject, $message, $headers);



?>


[2010-07-13 18:26:07] ahar...@php.net

Since I committed the offending patch, I'll look into it.



If you could provide both the test script (which doesn't seem to have
made it) and, if possible, a backtrace per the instructions [1], that
would be most helpful.



[1] http://bugs.php.net/bugs-generating-backtrace.php




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=52317


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52317&edit=1


Bug #11058 [Com]: php_network_getaddresses: getaddrinfo failed

2010-07-19 Thread office at hermannseib dot com
Edit report at http://bugs.php.net/bug.php?id=11058&edit=1

 ID:   11058
 Comment by:   office at hermannseib dot com
 Reported by:  pat at mail dot rit dot edu
 Summary:  php_network_getaddresses: getaddrinfo failed
 Status:   Bogus
 Type: Bug
 Package:  Network related
 Operating System: OpenBSD 2.6
 PHP Version:  4.0.6

 New Comment:

My solution:

Check whether PHP may read /etc/resolv.conf.

First, do a



  ls -la /etc/resolv.conf



and if it says something like this



-rw-r- 1 root root   /etc/resolv.conf



(i.e, NOT readable to anyone but root and the members of group root,
which normally means a highly elitist bunch :-), 

then issue the following command as root:



  chmod a+r /etc/resolv.conf



This cured the pesky "getaddrinfo failed" for me.


Previous Comments:

[2010-06-02 15:53:43] elynets at gmail dot com

Hello,

I`ve got the same problem on my server while tried to execute some of my
scripts.

The only solution that works fine for me is to disable firewall.

When firewall is disabled I get no errors. 

In the firewall settings all traffic allowed on ports 80 and 53.


[2010-05-26 23:27:26] pipra133 at yahoo dot com

I am having the same problem .. I am using XAMPP 2.5 version ... it did
work for me on the last version 



I am using the youtube API to receive youtube videos ... 



there is a major change in my case 



Message: Unable to Connect to tcp://gdata.youtube.com:80. Error #0:
php_network_getaddresses: getaddrinfo failed: No such host is known. 



the return says it tries with TCP rather then http , but my actual stack
trace says it is HTTP . 



#0 C:\xampp\php\PEAR\Zend\Gdata.php(221):
Zend_Gdata_App->performHttpRequest('GET', 'http://gdata.yo...', Array,
NULL, NULL, NULL)



Does anyone have a clue what is wrong ???


[2010-03-25 06:28:07] cha0s at therealcha0s dot net

Still present in 5.3.


[2009-04-08 05:37:11] aven...@php.net

I have this problem on RHEL5 too.

My sloved is "chmod 644 /etc/hosts"

Simple,but useful for me.


[2004-09-27 19:06:52] rowaldru at gpi dot com

I am currently on 4.3.8. Has this issue been addressed and fixed in
5.0.1?




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=11058


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=11058&edit=1


[PHP-BUG] Req #52377 [NEW]: Add .user.ini support to FPM

2010-07-19 Thread pch at ordbogen dot com
From: 
Operating system: 
PHP version:  5.3.2
Package:  FPM related
Bug Type: Feature/Change Request
Bug description:Add .user.ini support to FPM

Description:

If would be nice if the FastCGI Process Manager SAPI supported .user.ini
files 

like the ordinary FastCGI SAPI does. Although you can specify per-pool
php.ini 

for FPM, it really doesn't give you the same flexibility as .htaccess and 

.user.ini does.


-- 
Edit bug report at http://bugs.php.net/bug.php?id=52377&edit=1
-- 
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52377&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52377&r=trysnapshot53
Try a snapshot (trunk):  
http://bugs.php.net/fix.php?id=52377&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52377&r=fixed
Fixed in SVN and need be documented: 
http://bugs.php.net/fix.php?id=52377&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52377&r=alreadyfixed
Need backtrace:  
http://bugs.php.net/fix.php?id=52377&r=needtrace
Need Reproduce Script:   
http://bugs.php.net/fix.php?id=52377&r=needscript
Try newer version:   
http://bugs.php.net/fix.php?id=52377&r=oldversion
Not developer issue: 
http://bugs.php.net/fix.php?id=52377&r=support
Expected behavior:   
http://bugs.php.net/fix.php?id=52377&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=52377&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=52377&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52377&r=globals
PHP 4 support discontinued:  http://bugs.php.net/fix.php?id=52377&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52377&r=dst
IIS Stability:   
http://bugs.php.net/fix.php?id=52377&r=isapi
Install GNU Sed: 
http://bugs.php.net/fix.php?id=52377&r=gnused
Floating point limitations:  
http://bugs.php.net/fix.php?id=52377&r=float
No Zend Extensions:  
http://bugs.php.net/fix.php?id=52377&r=nozend
MySQL Configuration Error:   
http://bugs.php.net/fix.php?id=52377&r=mysqlcfg



Bug #52317 [Sus->Asn]: Segmentation fault when using mail() on a rhel 4.x (only 64 bit)

2010-07-19 Thread johannes
Edit report at http://bugs.php.net/bug.php?id=52317&edit=1

 ID:   52317
 Updated by:   johan...@php.net
 Reported by:  om at wysiwyg dot de
 Summary:  Segmentation fault when using mail() on a rhel 4.x
   (only 64 bit)
-Status:   Suspended
+Status:   Assigned
 Type: Bug
 Package:  Reproducible crash
 Operating System: rhel4.8_x86_64
 PHP Version:  5.2SVN-2010-07-12 (snap)
 Assigned To:  aharvey

 New Comment:

Approved for 5.2 and 5.3.


Previous Comments:

[2010-07-19 12:47:40] ahar...@php.net

Editing fail: ignore "I do find the fact it's 64-bit only
interesting,".



It's not that interesting when you think about it for two seconds. :)


[2010-07-19 12:46:51] ahar...@php.net

Gah, missing header inclusion + older version of gcc = failure, hence

why this wasn't manifesting on RHEL 5 or recent Ubuntu versions. I do

find the fact it's 64-bit only interesting, 



Fix checked into trunk. Given we're deep into the RC cycle, I'll talk

to the RMs before committing to the 5.2 and 5.3 branches, although I

think this is a pretty good candidate for both.


[2010-07-19 12:46:46] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301392
Log: Fix for bug #52317 (Segmentation fault when using mail() on a rhel
4.x (only 64
bit)).


[2010-07-15 10:32:42] om at wysiwyg dot de

(gdb) bt

#0  0x003fc0571360 in strlen () from /lib64/tls/libc.so.6

#1  0x003fc0542e79 in vfprintf () from /lib64/tls/libc.so.6

#2  0x003fc0548626 in fprintf () from /lib64/tls/libc.so.6

#3  0x0060b902 in php_mail (to=0x2a9557f208 "o...@wysiwyg.de",
subject=0x2a9557efc0 "the subject", 

message=0x2a95580b28 "hello", headers=0x9557f388 , 

extra_cmd=0x0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:281

#4  0x0060b6dc in zif_mail (ht=4, return_value=0x2a9557f318,
return_value_ptr=0x0, this_ptr=0x0, 

return_value_used=0) at
/usr/local/src/php5.2-201007131430/ext/standard/mail.c:180

#5  0x006ee2cb in zend_do_fcall_common_helper_SPEC
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:200

#6  0x006f3bdb in ZEND_DO_FCALL_SPEC_CONST_HANDLER
(execute_data=0x7fbfffd1c0)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:1740

#7  0x006eddc1 in execute (op_array=0x2a9557e130)

at /usr/local/src/php5.2-201007131430/Zend/zend_vm_execute.h:92

#8  0x006c747f in zend_execute_scripts (type=8, retval=0x0,
file_count=3)

at /usr/local/src/php5.2-201007131430/Zend/zend.c:1134

#9  0x00671979 in php_execute_script
(primary_file=0x7fb890)

at /usr/local/src/php5.2-201007131430/main/main.c:2036

#10 0x00741dfb in main (argc=2, argv=0x7fb9e8)

at /usr/local/src/php5.2-201007131430/sapi/cli/php_cli.c:1165

(gdb)


[2010-07-15 10:15:59] om at wysiwyg dot de

sorry I forgot the testscript:



 seg. fault

mail($to, $subject, $message, $headers);



?>




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=52317


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52317&edit=1


Bug #52317 [Asn->Csd]: Segmentation fault when using mail() on a rhel 4.x (only 64 bit)

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52317&edit=1

 ID:   52317
 Updated by:   ahar...@php.net
 Reported by:  om at wysiwyg dot de
 Summary:  Segmentation fault when using mail() on a rhel 4.x
   (only 64 bit)
-Status:   Assigned
+Status:   Closed
 Type: Bug
 Package:  Reproducible crash
 Operating System: rhel4.8_x86_64
 PHP Version:  5.2SVN-2010-07-12 (snap)
 Assigned To:  aharvey



Previous Comments:

[2010-07-19 15:38:55] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301396
Log: MFH: fix for bug #52317 (Segmentation fault when using mail() on a
rhel 4.x
(only 64 bit)).


[2010-07-19 15:08:15] johan...@php.net

Approved for 5.2 and 5.3.


[2010-07-19 12:47:40] ahar...@php.net

Editing fail: ignore "I do find the fact it's 64-bit only
interesting,".



It's not that interesting when you think about it for two seconds. :)


[2010-07-19 12:46:51] ahar...@php.net

Gah, missing header inclusion + older version of gcc = failure, hence

why this wasn't manifesting on RHEL 5 or recent Ubuntu versions. I do

find the fact it's 64-bit only interesting, 



Fix checked into trunk. Given we're deep into the RC cycle, I'll talk

to the RMs before committing to the 5.2 and 5.3 branches, although I

think this is a pretty good candidate for both.


[2010-07-19 12:46:46] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301392
Log: Fix for bug #52317 (Segmentation fault when using mail() on a rhel
4.x (only 64
bit)).




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=52317


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52317&edit=1


Bug #52317 [Csd]: Segmentation fault when using mail() on a rhel 4.x (only 64 bit)

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52317&edit=1

 ID:   52317
 Updated by:   ahar...@php.net
 Reported by:  om at wysiwyg dot de
 Summary:  Segmentation fault when using mail() on a rhel 4.x
   (only 64 bit)
 Status:   Closed
 Type: Bug
 Package:  Reproducible crash
 Operating System: rhel4.8_x86_64
 PHP Version:  5.2SVN-2010-07-12 (snap)
 Assigned To:  aharvey

 New Comment:

This bug has been fixed in SVN.

Snapshots of the sources are packaged every three hours; this change
will be in the next snapshot. You can grab the snapshot at
http://snaps.php.net/.
 
Thank you for the report, and for helping us make PHP better.




Previous Comments:

[2010-07-19 15:38:55] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301396
Log: MFH: fix for bug #52317 (Segmentation fault when using mail() on a
rhel 4.x
(only 64 bit)).


[2010-07-19 15:08:15] johan...@php.net

Approved for 5.2 and 5.3.


[2010-07-19 12:47:40] ahar...@php.net

Editing fail: ignore "I do find the fact it's 64-bit only
interesting,".



It's not that interesting when you think about it for two seconds. :)


[2010-07-19 12:46:51] ahar...@php.net

Gah, missing header inclusion + older version of gcc = failure, hence

why this wasn't manifesting on RHEL 5 or recent Ubuntu versions. I do

find the fact it's 64-bit only interesting, 



Fix checked into trunk. Given we're deep into the RC cycle, I'll talk

to the RMs before committing to the 5.2 and 5.3 branches, although I

think this is a pretty good candidate for both.


[2010-07-19 12:46:46] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301392
Log: Fix for bug #52317 (Segmentation fault when using mail() on a rhel
4.x (only 64
bit)).




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=52317


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52317&edit=1


[PHP-BUG] Bug #52378 [NEW]: new DateTime(time()) failed and should not

2010-07-19 Thread odou...@php.net
From: 
Operating system: 
PHP version:  5.3.2
Package:  Date/time related
Bug Type: Bug
Bug description:new DateTime(time()) failed and should not

Description:

Creating a new DateTime object with a unix timestamp failed with error  

Uncaught exception 'Exception' with message 'DateTime::__construct():
Failed to parse time string (1279547855) at position 8 (5): Unexpected
character'





If this is expected behaviour, this should be written in documentation and
DateTime::setTimestamp() should be fixed:

http://www.php.net/manual/en/datetime.settimestamp.php#Notes

"Passing a Unix timestamp to DateTime::__construct()  is an alternative
when using PHP 5.2."





Test script:
---
http://bugs.php.net/bug.php?id=52378&edit=1
-- 
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52378&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52378&r=trysnapshot53
Try a snapshot (trunk):  
http://bugs.php.net/fix.php?id=52378&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52378&r=fixed
Fixed in SVN and need be documented: 
http://bugs.php.net/fix.php?id=52378&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52378&r=alreadyfixed
Need backtrace:  
http://bugs.php.net/fix.php?id=52378&r=needtrace
Need Reproduce Script:   
http://bugs.php.net/fix.php?id=52378&r=needscript
Try newer version:   
http://bugs.php.net/fix.php?id=52378&r=oldversion
Not developer issue: 
http://bugs.php.net/fix.php?id=52378&r=support
Expected behavior:   
http://bugs.php.net/fix.php?id=52378&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=52378&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=52378&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52378&r=globals
PHP 4 support discontinued:  http://bugs.php.net/fix.php?id=52378&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52378&r=dst
IIS Stability:   
http://bugs.php.net/fix.php?id=52378&r=isapi
Install GNU Sed: 
http://bugs.php.net/fix.php?id=52378&r=gnused
Floating point limitations:  
http://bugs.php.net/fix.php?id=52378&r=float
No Zend Extensions:  
http://bugs.php.net/fix.php?id=52378&r=nozend
MySQL Configuration Error:   
http://bugs.php.net/fix.php?id=52378&r=mysqlcfg



[PHP-BUG] Bug #52379 [NEW]: php.ini not loaded

2010-07-19 Thread circus2 at freenet dot de
From: 
Operating system: WinXP
PHP version:  5.3.3RC3
Package:  CGI related
Bug Type: Bug
Bug description:php.ini not loaded

Description:

Load the Windows php-533r3-VC9-ts.zip. Unzip the package. Rename the
php.ini-development to php.ini. Run the cli command: php --ini in the
unzipped directory.



It will not detect the php.ini unless it is stored in C:\Windows. My path
contains spaces which may cause the problems, but with PHP 5.2.x it works
like expected.

Expected result:

It should detect the php.ini in the unzipped php directory. Works like
expected with php 5.2.14r3.


-- 
Edit bug report at http://bugs.php.net/bug.php?id=52379&edit=1
-- 
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52379&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52379&r=trysnapshot53
Try a snapshot (trunk):  
http://bugs.php.net/fix.php?id=52379&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52379&r=fixed
Fixed in SVN and need be documented: 
http://bugs.php.net/fix.php?id=52379&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52379&r=alreadyfixed
Need backtrace:  
http://bugs.php.net/fix.php?id=52379&r=needtrace
Need Reproduce Script:   
http://bugs.php.net/fix.php?id=52379&r=needscript
Try newer version:   
http://bugs.php.net/fix.php?id=52379&r=oldversion
Not developer issue: 
http://bugs.php.net/fix.php?id=52379&r=support
Expected behavior:   
http://bugs.php.net/fix.php?id=52379&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=52379&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=52379&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52379&r=globals
PHP 4 support discontinued:  http://bugs.php.net/fix.php?id=52379&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52379&r=dst
IIS Stability:   
http://bugs.php.net/fix.php?id=52379&r=isapi
Install GNU Sed: 
http://bugs.php.net/fix.php?id=52379&r=gnused
Floating point limitations:  
http://bugs.php.net/fix.php?id=52379&r=float
No Zend Extensions:  
http://bugs.php.net/fix.php?id=52379&r=nozend
MySQL Configuration Error:   
http://bugs.php.net/fix.php?id=52379&r=mysqlcfg



Bug #52376 [Opn->Fbk]: opendir() cannot open UNC paths in IIS7 using passthrough auth.

2010-07-19 Thread pajoye
Edit report at http://bugs.php.net/bug.php?id=52376&edit=1

 ID:   52376
 Updated by:   paj...@php.net
 Reported by:  ryan at kisc dot edu dot np
 Summary:  opendir() cannot open UNC paths in IIS7 using
   passthrough auth.
-Status:   Open
+Status:   Feedback
 Type: Bug
 Package:  *Directory/Filesystem functions
 Operating System: win64 - W2008R2
 PHP Version:  5.3.2
-Assigned To:  
+Assigned To:  pajoye

 New Comment:

Can you try to follow the advice here please:



http://support.microsoft.com/kb/214806


Previous Comments:

[2010-07-19 12:38:30] ryan at kisc dot edu dot np

Description:

This seems to be identical to #50542 but that one was closed and I
couldn't do anything to it.



Basically it was marked closed because with passthrough-authentication
essentially disabled by putting a username in "physical path
credentials" and selecting ClearText in the "Physical Path Credentials
Logon Type" both in advanced settings, this can be made to work so long
as that configuration is acceptable.



I too am having trouble with this and unfortunately the solution of
editing

physical path credentials breaks the pass through authentication to the

database, ldap, and other things. This all worked on 5.3.1 with IIS6 on
W2K3 but

on IIS7 with W2K8R2 no joy. So I upgraded to 5.3.2, still no joy. My
application is so reliant on the pass-through authentication that when I
do arrange things such that opendir works (clear-text + specific
physical path credentials) the system becomes a single user system!



So I can confirm the workaround mentioned for #50542 works if the
environment allows for it, but mine does not. Any other ideas? If I
connect to IIS using the environment requirement of network type auth
and pass-through authentication it does NOT work, though it really
should. If I use that exact same user in the "physical path credentials"
field and set "clear text" opendir suddenly works. Please don't close
this as it really is a bug and apparently specific to IIS7

Test script:
---


Expected result:

I expect to get no errors.





Actual result:
--
Warning:
opendir(//someserver/someshare/somefolder,//someserver/someshare/somefolder)
[function.opendir]: Access is denied. (code: 5) in
C:\inetpub\wwwroot\testdir.php on line 2



Warning: opendir(//someserver/someshare/somefolder) [function.opendir]:
failed to open dir: No such file or directory in
C:\inetpub\wwwroot\testdir.php on line 2






-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52376&edit=1


Bug #52369 [Com]: trying to delete nodes while looping domnodelist do not delete them

2010-07-19 Thread giorgio dot liscio at email dot it
Edit report at http://bugs.php.net/bug.php?id=52369&edit=1

 ID:   52369
 Comment by:   giorgio dot liscio at email dot it
 Reported by:  giorgio dot liscio at email dot it
 Summary:  trying to delete nodes while looping domnodelist do
   not delete them
 Status:   Open
 Type: Bug
 Package:  DOM XML related
 Operating System: all
 PHP Version:  5.3.2

 New Comment:

not a bug?


Previous Comments:

[2010-07-18 13:14:54] giorgio dot liscio at email dot it

Description:

hi, as summary says

here is the test code



';

$a = new DOMDocument(); $a->loadXML($s);

$els = $a->getElementsByTagName("abc");





// uncomment those three lines to fix

//foreach($els as $b)

//  $els2[]=$b; 

//$els=$els2;





foreach($els as $el)

$el->parentNode->removeChild($el);



echo "";

echo "document element () should be empty:\n";

echo $a->saveXML();



?>



uncomment lines to see the expected behavior







-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52369&edit=1


[PHP-BUG] Req #52381 [NEW]: You're doing the wrong thing with string accessors

2010-07-19 Thread mbt at gator dot net
From: 
Operating system: Linux
PHP version:  5.3.2
Package:  Unknown/Other Function
Bug Type: Feature/Change Request
Bug description:You're doing the wrong thing with string accessors

Description:

I should have made more fuss about this earlier, but here it is now.



I think you are taking exactly the wrong approach to addressing single

characters in strings.  I am of the very strong opinion that the direct

addressing of characters in strings have an obvious syntactic difference
from

the addressing of array elements.  This means you should keep {} and
deprecate

[] for addressing characters in strings.



PHP is not C!!  In C, strings are arrays, so [] is the preferred C operator
to

address single bytes (or character codes) in a string.  In PHP, strings are
a

primitive type; they are not arrays.  The notation $foo{0} makes it plain
that

$foo is a string, not an array and that you are accessing a byte, not an
array

element.  Making this obvious through the syntax is more important than
you

might think.



Imagine $bar to be a 3x3 array of strings.  In this case, what are you to

assume when you see in the middle of a script, far away from the assignment
to

$bar, when you read

   $baz = $bar[$a][$b][$c];



If you have instead

   $baz = $bar[$a][$b]{$c};

you know you're not doing ordinary array addressing.



These issues become important for the people who have to maintain the code.


Consider two more ways confusion can arise because strings are not arrays.



If $abc is a string,

   $d = $abc[2];

works but

   $abc[2] = array('boo');

fails.  The elements of a string "array" do not work like true array
elements.



Indices to strings must be integers; unlike arrays, strings do not allow

associative indices.  This fails:

   $abc['key'] = 'def';



I believe that indicating single-byte access via [] is such a bad idea that
if

you're dead set on removing {}, I would prefer to deprecate that kind of
access

altogether in favor of substr() and friends.


-- 
Edit bug report at http://bugs.php.net/bug.php?id=52381&edit=1
-- 
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52381&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52381&r=trysnapshot53
Try a snapshot (trunk):  
http://bugs.php.net/fix.php?id=52381&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52381&r=fixed
Fixed in SVN and need be documented: 
http://bugs.php.net/fix.php?id=52381&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52381&r=alreadyfixed
Need backtrace:  
http://bugs.php.net/fix.php?id=52381&r=needtrace
Need Reproduce Script:   
http://bugs.php.net/fix.php?id=52381&r=needscript
Try newer version:   
http://bugs.php.net/fix.php?id=52381&r=oldversion
Not developer issue: 
http://bugs.php.net/fix.php?id=52381&r=support
Expected behavior:   
http://bugs.php.net/fix.php?id=52381&r=notwrong
Not enough info: 
http://bugs.php.net/fix.php?id=52381&r=notenoughinfo
Submitted twice: 
http://bugs.php.net/fix.php?id=52381&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52381&r=globals
PHP 4 support discontinued:  http://bugs.php.net/fix.php?id=52381&r=php4
Daylight Savings:http://bugs.php.net/fix.php?id=52381&r=dst
IIS Stability:   
http://bugs.php.net/fix.php?id=52381&r=isapi
Install GNU Sed: 
http://bugs.php.net/fix.php?id=52381&r=gnused
Floating point limitations:  
http://bugs.php.net/fix.php?id=52381&r=float
No Zend Extensions:  
http://bugs.php.net/fix.php?id=52381&r=nozend
MySQL Configuration Error:   
http://bugs.php.net/fix.php?id=52381&r=mysqlcfg



Bug #52376 [Fbk->Asn]: opendir() cannot open UNC paths in IIS7 using passthrough auth.

2010-07-19 Thread ryan at kisc dot edu dot np
Edit report at http://bugs.php.net/bug.php?id=52376&edit=1

 ID:   52376
 User updated by:  ryan at kisc dot edu dot np
 Reported by:  ryan at kisc dot edu dot np
 Summary:  opendir() cannot open UNC paths in IIS7 using
   passthrough auth.
-Status:   Feedback
+Status:   Assigned
 Type: Bug
 Package:  *Directory/Filesystem functions
 Operating System: win64 - W2008R2
 PHP Version:  5.3.2
 Assigned To:  pajoye

 New Comment:

Well, that allows me to browse to the folder via chrome and essentially
does the 

same thing as the 

workaround from #50542 but only on the one folder which could work. What
it does 

not appear to do 

is give me programmatic access to the folder. The instructions in that
KB are 

outdated as it uses 

adsutil.vbs and the setting is the same as the "Physical Path
Credentials" from 

the IIS manager. 

This could work, in a much less than ideal way, if there is some way to
run 

"opendir" on the 

virtual directory without specifying the unc path (since the UNC path
itself 

still does not work). 

I could just be unaware of how, since it seems to use the system paths
and not 

honor or even 

acknowledge virtual directories. I tried lots of different formats
against my 

better judgement but 

none worked. I even tried using the http path reference. Still no joy.
I've used 

PHP since 1998 but 

realize something could have changed at any one of the releases, however
I'm not 

sure I see how a 

virtual directory could solve the ability to use opendir. Fair enough if
this 

was just a try. I'm 

happy to keep trying if it can help the community. This worked
fabulously in 

IIS6. I also confess 

that it appears to be Microsoft's fault. Very frustrating.



I wonder if I can create a symbolic link equivalent. I've done this
before in 

older versions of 

windows, but I don't think I've ever done it to a network share. In my
case I 

may be able to just 

move the share to the IIS server and be done with it, but I'm willing to
stick 

this out if it will 

help find a solution to this issue.


Previous Comments:

[2010-07-19 23:50:21] paj...@php.net

Can you try to follow the advice here please:



http://support.microsoft.com/kb/214806


[2010-07-19 12:38:30] ryan at kisc dot edu dot np

Description:

This seems to be identical to #50542 but that one was closed and I
couldn't do anything to it.



Basically it was marked closed because with passthrough-authentication
essentially disabled by putting a username in "physical path
credentials" and selecting ClearText in the "Physical Path Credentials
Logon Type" both in advanced settings, this can be made to work so long
as that configuration is acceptable.



I too am having trouble with this and unfortunately the solution of
editing

physical path credentials breaks the pass through authentication to the

database, ldap, and other things. This all worked on 5.3.1 with IIS6 on
W2K3 but

on IIS7 with W2K8R2 no joy. So I upgraded to 5.3.2, still no joy. My
application is so reliant on the pass-through authentication that when I
do arrange things such that opendir works (clear-text + specific
physical path credentials) the system becomes a single user system!



So I can confirm the workaround mentioned for #50542 works if the
environment allows for it, but mine does not. Any other ideas? If I
connect to IIS using the environment requirement of network type auth
and pass-through authentication it does NOT work, though it really
should. If I use that exact same user in the "physical path credentials"
field and set "clear text" opendir suddenly works. Please don't close
this as it really is a bug and apparently specific to IIS7

Test script:
---


Expected result:

I expect to get no errors.





Actual result:
--
Warning:
opendir(//someserver/someshare/somefolder,//someserver/someshare/somefolder)
[function.opendir]: Access is denied. (code: 5) in
C:\inetpub\wwwroot\testdir.php on line 2



Warning: opendir(//someserver/someshare/somefolder) [function.opendir]:
failed to open dir: No such file or directory in
C:\inetpub\wwwroot\testdir.php on line 2






-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52376&edit=1


Bug #52376 [Asn->Opn]: opendir() cannot open UNC paths in IIS7 using passthrough auth.

2010-07-19 Thread ryan at kisc dot edu dot np
Edit report at http://bugs.php.net/bug.php?id=52376&edit=1

 ID:   52376
 User updated by:  ryan at kisc dot edu dot np
 Reported by:  ryan at kisc dot edu dot np
 Summary:  opendir() cannot open UNC paths in IIS7 using
   passthrough auth.
-Status:   Assigned
+Status:   Open
 Type: Bug
 Package:  *Directory/Filesystem functions
 Operating System: win64 - W2008R2
 PHP Version:  5.3.2
 Assigned To:  pajoye

 New Comment:

sorry about the line breaks, apparently this site doesn't like what
Chrome does when I resize the text box. I'll be more careful in the
future. Actually this site seems to hate Chrome altogether. 

I keep getting "incorrect username" constantly. The bug site is buggy,
at least in Chrome.


Previous Comments:

[2010-07-20 04:04:31] ryan at kisc dot edu dot np

Well, that allows me to browse to the folder via chrome and essentially
does the 

same thing as the 

workaround from #50542 but only on the one folder which could work. What
it does 

not appear to do 

is give me programmatic access to the folder. The instructions in that
KB are 

outdated as it uses 

adsutil.vbs and the setting is the same as the "Physical Path
Credentials" from 

the IIS manager. 

This could work, in a much less than ideal way, if there is some way to
run 

"opendir" on the 

virtual directory without specifying the unc path (since the UNC path
itself 

still does not work). 

I could just be unaware of how, since it seems to use the system paths
and not 

honor or even 

acknowledge virtual directories. I tried lots of different formats
against my 

better judgement but 

none worked. I even tried using the http path reference. Still no joy.
I've used 

PHP since 1998 but 

realize something could have changed at any one of the releases, however
I'm not 

sure I see how a 

virtual directory could solve the ability to use opendir. Fair enough if
this 

was just a try. I'm 

happy to keep trying if it can help the community. This worked
fabulously in 

IIS6. I also confess 

that it appears to be Microsoft's fault. Very frustrating.



I wonder if I can create a symbolic link equivalent. I've done this
before in 

older versions of 

windows, but I don't think I've ever done it to a network share. In my
case I 

may be able to just 

move the share to the IIS server and be done with it, but I'm willing to
stick 

this out if it will 

help find a solution to this issue.


[2010-07-19 23:50:21] paj...@php.net

Can you try to follow the advice here please:



http://support.microsoft.com/kb/214806


[2010-07-19 12:38:30] ryan at kisc dot edu dot np

Description:

This seems to be identical to #50542 but that one was closed and I
couldn't do anything to it.



Basically it was marked closed because with passthrough-authentication
essentially disabled by putting a username in "physical path
credentials" and selecting ClearText in the "Physical Path Credentials
Logon Type" both in advanced settings, this can be made to work so long
as that configuration is acceptable.



I too am having trouble with this and unfortunately the solution of
editing

physical path credentials breaks the pass through authentication to the

database, ldap, and other things. This all worked on 5.3.1 with IIS6 on
W2K3 but

on IIS7 with W2K8R2 no joy. So I upgraded to 5.3.2, still no joy. My
application is so reliant on the pass-through authentication that when I
do arrange things such that opendir works (clear-text + specific
physical path credentials) the system becomes a single user system!



So I can confirm the workaround mentioned for #50542 works if the
environment allows for it, but mine does not. Any other ideas? If I
connect to IIS using the environment requirement of network type auth
and pass-through authentication it does NOT work, though it really
should. If I use that exact same user in the "physical path credentials"
field and set "clear text" opendir suddenly works. Please don't close
this as it really is a bug and apparently specific to IIS7

Test script:
---


Expected result:

I expect to get no errors.





Actual result:
--
Warning:
opendir(//someserver/someshare/somefolder,//someserver/someshare/somefolder)
[function.opendir]: Access is denied. (code: 5) in
C:\inetpub\wwwroot\testdir.php on line 2



Warning: opendir(//someserver/someshare/somefolder) [function.opendir]:
failed to open dir: No such file or directory in
C:\inetpub\wwwroot\testdir.php on line 2






-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52376&edit=1


Bug #52376 [Opn->Asn]: opendir() cannot open UNC paths in IIS7 using passthrough auth.

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52376&edit=1

 ID:   52376
 Updated by:   ahar...@php.net
 Reported by:  ryan at kisc dot edu dot np
 Summary:  opendir() cannot open UNC paths in IIS7 using
   passthrough auth.
-Status:   Open
+Status:   Assigned
 Type: Bug
 Package:  *Directory/Filesystem functions
 Operating System: win64 - W2008R2
 PHP Version:  5.3.2
 Assigned To:  pajoye

 New Comment:

(Restoring status.)



As a fellow Chrome user, I feel your pain. :)


Previous Comments:

[2010-07-20 04:08:58] ryan at kisc dot edu dot np

sorry about the line breaks, apparently this site doesn't like what
Chrome does when I resize the text box. I'll be more careful in the
future. Actually this site seems to hate Chrome altogether. 

I keep getting "incorrect username" constantly. The bug site is buggy,
at least in Chrome.


[2010-07-20 04:04:31] ryan at kisc dot edu dot np

Well, that allows me to browse to the folder via chrome and essentially
does the 

same thing as the 

workaround from #50542 but only on the one folder which could work. What
it does 

not appear to do 

is give me programmatic access to the folder. The instructions in that
KB are 

outdated as it uses 

adsutil.vbs and the setting is the same as the "Physical Path
Credentials" from 

the IIS manager. 

This could work, in a much less than ideal way, if there is some way to
run 

"opendir" on the 

virtual directory without specifying the unc path (since the UNC path
itself 

still does not work). 

I could just be unaware of how, since it seems to use the system paths
and not 

honor or even 

acknowledge virtual directories. I tried lots of different formats
against my 

better judgement but 

none worked. I even tried using the http path reference. Still no joy.
I've used 

PHP since 1998 but 

realize something could have changed at any one of the releases, however
I'm not 

sure I see how a 

virtual directory could solve the ability to use opendir. Fair enough if
this 

was just a try. I'm 

happy to keep trying if it can help the community. This worked
fabulously in 

IIS6. I also confess 

that it appears to be Microsoft's fault. Very frustrating.



I wonder if I can create a symbolic link equivalent. I've done this
before in 

older versions of 

windows, but I don't think I've ever done it to a network share. In my
case I 

may be able to just 

move the share to the IIS server and be done with it, but I'm willing to
stick 

this out if it will 

help find a solution to this issue.


[2010-07-19 23:50:21] paj...@php.net

Can you try to follow the advice here please:



http://support.microsoft.com/kb/214806


[2010-07-19 12:38:30] ryan at kisc dot edu dot np

Description:

This seems to be identical to #50542 but that one was closed and I
couldn't do anything to it.



Basically it was marked closed because with passthrough-authentication
essentially disabled by putting a username in "physical path
credentials" and selecting ClearText in the "Physical Path Credentials
Logon Type" both in advanced settings, this can be made to work so long
as that configuration is acceptable.



I too am having trouble with this and unfortunately the solution of
editing

physical path credentials breaks the pass through authentication to the

database, ldap, and other things. This all worked on 5.3.1 with IIS6 on
W2K3 but

on IIS7 with W2K8R2 no joy. So I upgraded to 5.3.2, still no joy. My
application is so reliant on the pass-through authentication that when I
do arrange things such that opendir works (clear-text + specific
physical path credentials) the system becomes a single user system!



So I can confirm the workaround mentioned for #50542 works if the
environment allows for it, but mine does not. Any other ideas? If I
connect to IIS using the environment requirement of network type auth
and pass-through authentication it does NOT work, though it really
should. If I use that exact same user in the "physical path credentials"
field and set "clear text" opendir suddenly works. Please don't close
this as it really is a bug and apparently specific to IIS7

Test script:
---


Expected result:

I expect to get no errors.





Actual result:
--
Warning:
opendir(//someserver/someshare/somefolder,//someserver/someshare/somefolder)
[function.opendir]: Access is denied. (code: 5) in
C:\inetpub\wwwroot\testdir.php on line 2



Warning: opendir(//someserver/someshare/somefolder) [function.opendir]:
failed to open dir: No such file or directory in
C:\inetpub\wwwroot\testdir.php on line 2


-

Bug #51527 [Com]: is_callable() returning true for non-static callbacks

2010-07-19 Thread hnzksq at gmail dot com
Edit report at http://bugs.php.net/bug.php?id=51527&edit=1

 ID:   51527
 Comment by:   hnzksq at gmail dot com
 Reported by:  weierophin...@php.net
 Summary:  is_callable() returning true for non-static callbacks
 Status:   Open
 Type: Bug
 Package:  Class/Object related
 Operating System: Linux
 PHP Version:  5.3.2

 New Comment:





我测试可以用的。


Previous Comments:

[2010-05-07 16:39:56] crrodriguez at opensuse dot org

philduby at phriendly dot net : what you mention seems to be  another 

bug/misfeature.


[2010-05-05 07:03:50] philduby at phriendly dot net

Another variation that actually (unexpectedly) works:

Calling is_callable and call_user_func from inside an instance
(non-static) method using any of: 'self::otherInstanceMethod',
array('self','otherInstanceMethod'), array(self,'otherInstanceMethod')
succeed.  It appears that (the context for) '$this' is carried over from
the original method, even though the calls are being done statically.



Calling self::otherInstanceMethod() directly also succeeds.  It appears
that methods called from an instance method 'inherit' the context for
$this.  A bit unexpected, but *reasonable*.

Win XP SP3

PHP 5.3.1 (xampp)


[2010-04-12 11:50:33] paj...@php.net

I don't like the optional flag idea. You can disable strictness using
error_reporting already.



However I would like to see this change in trunk, can you come up with a
RFC pls?


[2010-04-10 20:19:45] weierophin...@php.net

@johannes Perhaps an optional "strict" flag to is_callable() would
address this? 

That would keep BC, while allowing for better checking for >= PHP-5
apps. Usage 

would be:



$callback = array('Foo', 'bar');



if (is_callable($callback, true)) {

// Passed strict check

} else {

// failed strict check

}



Thoughts?


[2010-04-10 17:55:25] johan...@php.net

The method can always be called statically. The access to $this might be
forbidden but the method itself can be called. is_callable() doesn't
mean it will execute properly. (the $this error is basically the same as
a call to an undefined funciton in there or such)



To change this the only way would be to forbid calling non-static
methods statically. Maybe this can be done nowadays (we needed the
behavior for PHP 4 compatibility) but that's no bug but requires a RFC
and discussion on internals as this might break quite a few applications
(many PEAR-based things, many legacy applications not fully "converted"
to PHP 5 )




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at

http://bugs.php.net/bug.php?id=51527


-- 
Edit this bug report at http://bugs.php.net/bug.php?id=51527&edit=1


Req #52381 [Opn->Bgs]: You're doing the wrong thing with string accessors

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52381&edit=1

 ID:   52381
 Updated by:   ahar...@php.net
 Reported by:  mbt at gator dot net
 Summary:  You're doing the wrong thing with string accessors
-Status:   Open
+Status:   Bogus
 Type: Feature/Change Request
 Package:  Unknown/Other Function
 Operating System: Linux
 PHP Version:  5.3.2

 New Comment:

This decision dates back the best part of five years at this stage to

the developer meeting in November 2005. [0] It's come up a couple of

times on the Internals mailing list since then -- both immediately

after the developer meeting [1], and then to confirm the 5.3 migration

document was correct. [2]



To be blunt, the horse has already bolted on this. {} is deprecated,

and [] is the supported way to access strings by offset both now and

in the future.



[0] http://www.php.net/~derick/meeting-notes.html#cleanup-for-vs

[1] http://www.mail-archive.com/intern...@lists.php.net/msg18638.html

[2] http://www.mail-archive.com/intern...@lists.php.net/msg42883.html


Previous Comments:

[2010-07-20 01:53:12] mbt at gator dot net

Description:

I should have made more fuss about this earlier, but here it is now.



I think you are taking exactly the wrong approach to addressing single

characters in strings.  I am of the very strong opinion that the direct

addressing of characters in strings have an obvious syntactic difference
from

the addressing of array elements.  This means you should keep {} and
deprecate

[] for addressing characters in strings.



PHP is not C!!  In C, strings are arrays, so [] is the preferred C
operator to

address single bytes (or character codes) in a string.  In PHP, strings
are a

primitive type; they are not arrays.  The notation $foo{0} makes it
plain that

$foo is a string, not an array and that you are accessing a byte, not an
array

element.  Making this obvious through the syntax is more important than
you

might think.



Imagine $bar to be a 3x3 array of strings.  In this case, what are you
to

assume when you see in the middle of a script, far away from the
assignment to

$bar, when you read

   $baz = $bar[$a][$b][$c];



If you have instead

   $baz = $bar[$a][$b]{$c};

you know you're not doing ordinary array addressing.



These issues become important for the people who have to maintain the
code. 

Consider two more ways confusion can arise because strings are not
arrays.



If $abc is a string,

   $d = $abc[2];

works but

   $abc[2] = array('boo');

fails.  The elements of a string "array" do not work like true array
elements.



Indices to strings must be integers; unlike arrays, strings do not
allow

associative indices.  This fails:

   $abc['key'] = 'def';



I believe that indicating single-byte access via [] is such a bad idea
that if

you're dead set on removing {}, I would prefer to deprecate that kind of
access

altogether in favor of substr() and friends.







-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52381&edit=1


Doc->Bug #52380 [Opn->Asn]: Typo in lcfirst comment docs

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52380&edit=1

 ID:  52380
 Updated by:  ahar...@php.net
 Reported by: toddw at activestate dot com
 Summary: Typo in lcfirst comment docs
-Status:  Open
+Status:  Assigned
-Type:Documentation Problem
+Type:Bug
-Package: Documentation problem
+Package: Strings related
 PHP Version: trunk-SVN-2010-07-19 (SVN)
-Assigned To: 
+Assigned To: aharvey



Previous Comments:

[2010-07-19 21:46:24] toddw at activestate dot com

Description:

The lcfirst function has the wrong proto comment name - being "ucfirst"
instead of "lcfirst", see snippet from "ext/standard/string.c" below:



  /* {{{ proto string ucfirst(string str)

 Make a string's first character lowercase */

  PHP_FUNCTION(lcfirst)

  {









-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52380&edit=1


Bug #52380 [Asn->Sus]: Typo in lcfirst comment docs

2010-07-19 Thread aharvey
Edit report at http://bugs.php.net/bug.php?id=52380&edit=1

 ID:  52380
 Updated by:  ahar...@php.net
 Reported by: toddw at activestate dot com
 Summary: Typo in lcfirst comment docs
-Status:  Assigned
+Status:  Suspended
 Type:Bug
 Package: Strings related
 PHP Version: trunk-SVN-2010-07-19 (SVN)
 Assigned To: aharvey

 New Comment:

Fixed on trunk. Will MFH once the 5.2.14 and 5.3.3 releases are out.


Previous Comments:

[2010-07-20 06:26:59] ahar...@php.net

Automatic comment from SVN on behalf of aharvey
Revision: http://svn.php.net/viewvc/?view=revision&revision=301412
Log: Fix bug #52380 (Typo in lcfirst comment docs) by updating the
commented
prototype for lcfirst() to actually read lcfirst().


[2010-07-19 21:46:24] toddw at activestate dot com

Description:

The lcfirst function has the wrong proto comment name - being "ucfirst"
instead of "lcfirst", see snippet from "ext/standard/string.c" below:



  /* {{{ proto string ucfirst(string str)

 Make a string's first character lowercase */

  PHP_FUNCTION(lcfirst)

  {









-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52380&edit=1