Re: [PHP-DEV] PATCH: SimpleXML asXML()

2004-01-19 Thread Sterling Hughes
Change committed - thanks. -Sterling > SimpleXML's asXML() method always returns the entire document > regardless of the node. I believe (Sterling?) we decided the correct > behavior here is to only return the XML data for the current node and > its children. > > A patch to fix the behavior is a

[PHP-DEV] PATCH: SimpleXML asXML()

2004-01-19 Thread Adam Maccabee Trachtenberg
SimpleXML's asXML() method always returns the entire document regardless of the node. I believe (Sterling?) we decided the correct behavior here is to only return the XML data for the current node and its children. A patch to fix the behavior is available here: http://www.trachtenberg.com/patches

Re: [PHP-DEV] mktime problem...

2004-01-19 Thread Jonathan Hernandez
Well, I see that is a normal behavior of libc/mktime (it returns -1 if an error was detected), I see others have the same problem and implements their own php/mktime function (see the first comment on http://www.php.net/mktime)... I wonder if it can be implemented directly on the php function...

Re: [PHP-DEV] The __clone Wars

2004-01-19 Thread George Schlossnagle
On Jan 19, 2004, at 3:52 PM, [EMAIL PROTECTED] wrote: In other words, should it be like this: Yes. P.S. Perhaps the $this object should already be set using the default implementation, and then the user could override certain properties if he so wishes; I believe the preliminary documentation

[PHP-DEV] The __clone Wars

2004-01-19 Thread lingwitt
(Sorry for the subject, I couldn't resist). It seems that the parent class's variables are not being copied per bit by the default implementation, so how should a superclasses __clone() method be dealt with? In other words, should it be like this: public function __clone() { $this = par

[PHP-DEV] CLI segfault, in sapi_cli_single_write

2004-01-19 Thread Pierre-Alain Joye
Hello, I got a segfault in sapi_cli_single_write. For an (yet) unknown reason, str becomes empty. The attached patch fixes it. sidenote: found it by trying to resolve huge segfault in dom_xpath_query pierre Index: sapi/cli/php_cli.c ===

Re: [PHP-DEV] mktime problem...

2004-01-19 Thread Wez Furlong
Your libc doesn't support negative time_t values. --Wez. - Original Message - From: "Jonathan Hernandez" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, January 19, 2004 7:20 PM Subject: [PHP-DEV] mktime problem... > Hello, > > I have a problem with mktime, when I try to run

[PHP-DEV] mktime problem...

2004-01-19 Thread Jonathan Hernandez
Hello, I have a problem with mktime, when I try to run this: "; $date = getdate(-347090400); echo $date[0]." ".$date['mday']."-".$date['mon']."-".$date['year'].""; ?> Output: -62 31-12-1969 -347090400 1-1-1959 I compile php5.0b3 on Fedora(gcc 3.3.2/ kernel 2.4.22), if I try to run it on Debian

[PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c

2004-01-19 Thread Marcus Boerger
Hello Wez, Monday, January 19, 2004, 12:15:24 PM, you wrote: > The win32 build will correctly order extensions for internal_functions.c, > so that solves static compiles, but that doesn't help dynamic loading > (nor the unix build). > Would it be ok to keep Marcus' function so that a wise extens

[PHP-DEV] Re: [PHP-CVS] cvs: php-src /ext/xsl php_xsl.c php_xsl.h xsl_fe.h xsltprocessor.c

2004-01-19 Thread Martin Jansen
On Mon Jan 19, 2004 at 03:0709PM -, Christian Stocker wrote: > chreguMon Jan 19 10:07:09 2004 EDT > > Modified files: > /php-src/ext/xsl php_xsl.c php_xsl.h xsl_fe.h xsltprocessor.c > Log: > @- added XsltProcessor->registerPhpFunctions() (Christian) S

Re: [PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c

2004-01-19 Thread Christian Schneider
Wez Furlong wrote: This is quite difficult to implement correctly, as the extension name != module file name, and that filename is platform dependent Hmm. Three passes should do the job: First DL_LOAD, then DL_FETCH_SYM to determine order of init then initialize the modules. Or am I missing somet

Fw: [PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c

2004-01-19 Thread Wez Furlong
Bah, stupid OE. - Original Message - From: "Wez Furlong" <[EMAIL PROTECTED]> To: "Christian Schneider" <[EMAIL PROTECTED]> Sent: Monday, January 19, 2004 11:53 AM Subject: Re: [PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c > This is quite difficult to implement correctly,

[PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c

2004-01-19 Thread Christian Schneider
Wez Furlong wrote: Even better would be to add an array of module names into the module structure and have the ZE apply this check when it loads the module. I think that it is reasonably important to warn the user that things will break if we can detect it. If there was an array of dependencies in

[PHP-DEV] Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 / zend_API.c

2004-01-19 Thread Wez Furlong
The win32 build will correctly order extensions for internal_functions.c, so that solves static compiles, but that doesn't help dynamic loading (nor the unix build). Would it be ok to keep Marcus' function so that a wise extension can check if things were loaded in the correct order? Maybe adjust

Re: [PHP-DEV] SPL

2004-01-19 Thread Derick Rethans
On Mon, 19 Jan 2004, Andi Gutmans wrote: > Fine with me. It is done then. Derick -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php

Re: [PHP-DEV] SPL

2004-01-19 Thread Andi Gutmans
Fine with me. At 11:22 PM 1/18/2004 +0100, Marcus Boerger wrote: Hello internals, i would like to see ext/SPL from PECL as a default built-in PHP5 module. It is stable enough and for 5.0 i do not plan any more features. Since some earlier SimpleXML features are moved into SPL this makes even mo