RE: [PHP-DEV] Re: Object Overloading
-Original Message- From: Zeev Suraski [mailto:[EMAIL PROTECTED]] Sent: 23 August 2002 18:59 To: Sam Liddicott Cc: [EMAIL PROTECTED] Subject: RE: [PHP-DEV] Re: Object Overloading At 17:09 23/08/2002, Sam Liddicott wrote: Or am I missing something obvious? ZE2 objects are not at all like ZE1 objects. We're way beyond that point there now :) Is there any chance I can look at some zend 5 code to see the new model, and how it might fit with SWIG? Also, overloaded objects are good, what do you think about overloaded variables? (And with references these might just become overloaded zvals... Hmm...) Anyway... Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] help, C property get/set object overloading
swig wraps C++ objects as zend overloaded objects, the property getters and
setters fall back to storing things in the
Z_OBJPROP_P(property_reference-object) hash but I'm not doing it properly
and I can't work out what I'm doing wrong.
The propget functions are generated like this:
static pval _wrap_propget_ArrayStruct(zend_property_reference
*property_referenc
e) {
pval result;
pval **_result;
zend_llist_element *element = property_reference-elements_list-head;
zend_overloaded_element *property=(zend_overloaded_element
*)element-data;
/* Parent class hierachy dispatcher */
if (_propget_ArrayStruct(property_reference, result)==SUCCESS) return
result;
/* Nope? return it ourselves from property hash*/
if
(zend_hash_find(Z_OBJPROP_P(property_reference-object),Z_STRVAL_P((proper
ty-element)),1+Z_STRLEN_P((property-element)),(void**)_result)==SUCCESS)
ret
urn **_result;
result.type = IS_NULL;
return result;
}
No the trouble is I don't get reference counting/copying/whatever right,
and, if for example I store a string in the hash, then later, if I read the
string back a few times I find it's value changes, so I'm sure I'm passing
around a free'd zval.
The set function is generated like this:
static int _wrap_propset_ArrayStruct(zend_property_reference
*property_reference
, pval *value) {
zend_llist_element *element = property_reference-elements_list-head;
zend_overloaded_element *property=(zend_overloaded_element
*)element-data;
/* Parent class hierachy dispatcher */
if (_propset_ArrayStruct(property_reference, value)==SUCCESS) return
SUCCESS;
/* nope? set it ourselves */
return
add_property_zval_ex(property_reference-object,Z_STRVAL_P((property-
element)),1+Z_STRLEN_P((property-element)),value);
}
I've tried zval_copy_ctor, SEPERATE_ZVAL and the like both before storing
and after reading but I can't get any joy:
This sort of php script:
?php
// Sample test file
require arrays.php;
$ss=new simplestruct();
$ss-array_c=ab;
var_dump($ss);
var_dump($ss-array_c);
$x=$ss-array_c;
print 1 $x\n;
$x=$ss-array_c;
print 2 $x\n;
$x=$ss-array_c;
print 3 $x\n;
var_dump($x);
?
outputs this:
X-Powered-By: PHP/4.2.1
Content-type: text/html
object(simplestruct)(2) {
[_cPtr]=
resource(1) of type (_p_SimpleStruct)
[array_c]=
string(2) ab
}
string(2) ab
1 1
2 2
3 3
string(2) T
instead of
string(2) ab
1 ab
2 ab
3 ab
string(2) ab
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Object Overloading
Zeev I just wanted to make you aware that not just the Java module, but also all SWIG generated php modules also use the object overloading propget and propset way of accessing properties of the wrapped objects. I hear this was being revamped for zend 2.0 and thought I should let you know. SWIG-php works by generating a seperate overloaded class for each wrapped class, with methods as normal in the method table, but special behaviour to intercept the getting and setting of properties. Will the propget/propset system be lost in zend 2.0? Further for your information; (and I realise not for discussion); the wrapped class hierachy is mirrored in php (get_parent_class) except where the wrapped class hierachy uses mutiple inheritance in which case get_parent_class only returns the main parent class, but class properties and such from all classes remain available. I mention this in case zend 2.0 does support multiple inheritance and so you may know how we would use it. Thanks Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Only one CXX_SUFFIX
Is there a great reason why rules_common.mk can't be modified so the suffix rules work on .cxx as well as .cpp ? Sam _ Samuel Liddicott Support Consultant [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] Tel: +44 (0)113 367 4523 Fax: +44 (0)113 367 4680 Switchboard: +44 (0)113 367 4600 Orange Multimedia Operations Marshall Mill Marshall Street Leeds LS11 9YJ http://www.ananova.com/ www.ananova.com Ananova Limited Registered Office: St James Court Great Park Road Almondsbury Park Bradley Stoke Bristol BS32 4QJ Registered in England No.2858918 The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you receive this in error, please contact the sender and delete the material from any computer. I'm required to have this long signature so please don't pick on me about it, yes I know it's often longer than the email message it goes with.
[PHP-DEV] INTERNAL FEATURE request
Please can we have a magic cookie resource value or zval associated with PHP objects accessable via the C API? This would be used when the PHP class it wrapping a C++ class (such as one linked in via swig ( www.swig.org http://www.swig.org )) so there was a cheap way to find which C++ object the php object was wrapping. Currently we use _cPtr element of the wrapped object but it is open to freakery by user scripts. Thanks SAm _ Samuel Liddicott Support Consultant [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] Direct Dial: +44 (0)113 367 4523 Fax: +44 (0)113 367 4680 Switchboard: +44 (0)113 367 4600 Ananova Limited Marshall Mill Marshall Street Leeds LS11 9YJ http://www.ananova.com Registered Office: St James Court Great Park Road Almondsbury Park Bradley Stoke Bristol BS32 4QJ Registered in England No.2858918 The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you receive this in error, please contact the sender and delete the material from any computer.
RE: [PHP-DEV] Persistent overloaded class registration problem,
-Original Message- From: Andi Gutmans [mailto:[EMAIL PROTECTED]] Sent: 25 April 2002 23:14 To: Sam Liddicott; Sam Liddicott; 'Rasmus Lerdorf' Cc: 'PHP Developers Mailing List' Subject: Re: [PHP-DEV] Persistent overloaded class registration problem, At 11:11 24/04/2002 +0100, Sam Liddicott wrote: My overloaded classes are now registered in MINIT with: INIT_OVERLOADED_CLASS_ENTRY() zend_register_internal_class() ..zend_register_list_destructors_ex And all seems mostly fine; except the second time apache serves a request, one of the classes (!!) has been de-registered! As shown by get_declared_classes() in the script. I'm not sure what's happening there. Did you grep the directories under ext/ for examples of how to use zend_register_internal_class()? Yes, most of it I copied from java/java.c however it turned out to be another pointer fault (HashPosition * instead of HashPosition) which was clobbering zend memory, so it works fine now. Thanks v.much Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Proposal! Destructor notification
Some know I'm working to improve swig-php to make module generation easy and satisfying. The final hurdle remains in handling and generating callbacks from the module to PHP. Or rather from the library the module wraps to PHP. This requires some kind of callback wrapper that can convert from a flat function pointer the library expects to a complete object-method (or just function) reference that call_user_function expects. This problem is solved but it requires some per-callback-function allocation to hold this extra data. The new problem is that swig can't tell when the scope of the callback closes (when no more callbacks will occur) and when it can free this structure, so for longer lived scripts which make new objects often to handle callbacks we allocate and do not free ever more and more structures. If a module could request notification when all/specific objects are destroyed it would be able to free these structures when the object dies and thus avoid long term memory leaks. Are PHP developers willing to provide a mechanism where more than one module can request notification when objects generally, or specific objects are destroyed? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] RE: phpize writes config.m4
Sorry; wrong list, I get boths lists in the same mailbox and I just replied to a message already there. -Original Message- From: Sam Liddicott Sent: 26 April 2002 12:01 To: [EMAIL PROTECTED] Subject: phpize writes config.m4 I'm making some more changes to php-swig, where it writes out the sample config.m4 file. It's pretty complete now but it strikes me that this code should probably be part of phpize. Most modules have the same requirements for config.m4 (below). Perhaps we could have a phpize.in file where we provide the module name, required header and library files and an optional symbol to test compile each library. Does this make sense? I've already done it for swig but I think it belongs in phpize Sam 1) PHP_ARG_WITH 2) search paths for INCDIR looking for certain include files 3) search paths for LIBDIR looking for certain LIBS 4) PHP_REQUIRE_CXX, AC_CHECK_LIB(stdc++, cin) if it is a C++ module optional 5) AC_CHECK_LIB... to check the library links and a test symbol is present 6) PHP_SUBST(***_SHARED_LIBADD) AC_DEFINE(HAVE_***, 1, [ ]) AC_DEFINE_UNQUOTED(PHP_***_DIR, $***_DIR, [ ] ) for each library PHP_ADD_LIBRARY_WITH_PATH(***, $***_LIBDIR, ***_SHARED_LIBADD) for include path PHP_ADD_INCLUDE($***_INCDIR) 7) PHP_EXTENSION(***, $ext_shared) -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] ?= and %= both work, why not ?php=
-Original Message- From: Brinkman, Theodore [mailto:[EMAIL PROTECTED]] Sent: 26 April 2002 14:55 To: 'PHP Developers Mailing List' Subject: RE: [PHP-DEV] ?= and %= both work, why not ?php= Sure, and its only an extra 4 character, really. But that's not the issue at hand. The issue at hand is that the inconsistency of supporting ?= and %= but not ?php= encourages quite a few people to use the 'optional' short form tags, meaning that their code isn't portable. I guy here who till recently poo-poo'd asp tags is now using them because %=$VAR;% is emminently more readable than the alternative. For each person who says ?php= $variable ? is hard to read at least one other person says they find ?php echo $variable ? harder to read. I personally find the first easier to read when it is embedded in the middle of a long line of HTML (like an input tag for example). Yep. What possible harm comes from improving the internal consistency of the language? Why is a two-line patch that would completely remove an inconsistency so bitterly fought against? To emphasise; people here are adopting bad-old short tags in order to keep readability of code. It makes it easy to see the code is passive, echoing only. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Persistent overloaded class registration problem,
My overloaded classes are now registered in MINIT with: INIT_OVERLOADED_CLASS_ENTRY() zend_register_internal_class() ..zend_register_list_destructors_ex And all seems mostly fine; except the second time apache serves a request, one of the classes (!!) has been de-registered! As shown by get_declared_classes() in the script. Swig is generating all the classes the same way so I'm asking if anyone might know the cause for this. The class registration pattern I am following leaves no scope for me to estrdup or emalloc and the like as zend underlying layers do all this stuff properly for me. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Thanks! RE: [PHP-DEV] Persistent overloaded class registration problem,
-Original Message- From: Markus Fischer [mailto:[EMAIL PROTECTED]] Sent: 24 April 2002 11:24 To: Sam Liddicott Cc: 'Rasmus Lerdorf'; 'PHP Developers Mailing List' Subject: Re: [PHP-DEV] Persistent overloaded class registration problem, Hi, it might help if you give a link to some self-contained generated source example so we get the full picture. Good tip and in doing this I found it was a typemap wrapper using HasPosition *pos instead of HashPositionPos so I was clobbering random memory. So thanks to Easmus and Markus for this great help in getting swig-php4 improved. All works fine now! The release (1.3.12) is anticipated somewhere around June time according to the swig maintainer David Beezley. I anticipate that swig could replace ext_skel as the preferred way to create php modules; who wants to discuss making this an official position? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] NOTICE: Wise volunteer needed for PHP module review
I'm working with Richard Palmer on the php module of SWIG (http://www.swig.org) Swig is used to automatically write php, perl, python, guile, java, ruby, etc etc modules to link 3rd part libraries into these scripting systems. This means if I have a fancy library libfoo and I want to use it for php, I just write a single simple interface file (or even use the supplied .h header file); run swig on it and taraa!!! I have a php module ready to compile in to php. Richard and myself are making some _heavy_changes_ to the php module lately and we need some eagle/gimlet eye to review the output code of some simple cases to see that we are generating the code correctly - with regards use of ZVALs, destructors, resource handling, CG() vars and so forth. My involvement with swig is working on xapian (www.xapian.org) which is the open muscat descendant; so once we get this sorted we will also have a xapian module for php; in fact php will be providing the scripting for the main xapian search results page supplied with xapian. I've asked quite a few questions lately here but had no response and so have had to guess at the answers. Volunteers to review swig generated code would be appreciated. regards Sam Liddicott -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] MINIT or RINIT
Swig generated code registers some overloaded classes to wrap C classes. I think it should do this in PHP_MINIT but when I do it here they are only available as registered for the first apache request. If I register them in PHP_RINIT then they are always available. Why would this be so? Which should I use? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] MINIT or RINIT
-Original Message- From: Rasmus Lerdorf [mailto:[EMAIL PROTECTED]] Sent: 23 April 2002 11:30 To: Sam Liddicott Cc: PHP Developers Mailing List Subject: Re: [PHP-DEV] MINIT or RINIT Well, where are you storing them? Make sure they are not in memory that gets cleaned on a per-request basis. ie. do your own malloc() in MINIT and free() in MSHUTDOWN. Stupid questions follow: So... does my zend_class_entry have to be wrapped in CG()? Right now swig generates lines like: CG(class_entry).type=... Can I just do: zend_class_entry _swig_wrap_; _swig_wrap_X.type=... And it also currently does: CG(class_entry).refcount=(int *)emalloc(sizerof(int)); *CG(class_entry).refcount=1; I have no idea why it does this; do we need to init refcount? To anything simpler? Thanks Sam -Rasmus On Tue, 23 Apr 2002, Sam Liddicott wrote: Swig generated code registers some overloaded classes to wrap C classes. I think it should do this in PHP_MINIT but when I do it here they are only available as registered for the first apache request. If I register them in PHP_RINIT then they are always available. Why would this be so? Which should I use? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] MINIT or RINIT
-Original Message- From: Rasmus Lerdorf [mailto:[EMAIL PROTECTED]] Sent: 23 April 2002 11:30 To: Sam Liddicott Cc: PHP Developers Mailing List Subject: Re: [PHP-DEV] MINIT or RINIT Well, where are you storing them? Make sure they are not in memory that gets cleaned on a per-request basis. ie. do your own malloc() in MINIT and free() in MSHUTDOWN. This means my class must be declared as ZEND_INTERNAL_CLASS then? Are there any other side effects to doing this? Must I un-register my classes during shutdown? It seems zend does that for me. I've made these changes, to use ZEND_INTERNAL_CLASS and malloc and strdup instead of emalloc and estrdup but now zend hangs on shutdown. It seems to come round (under gdb) somewhere in _efree, under zend_hash_destroy under destroy_zend_class So it sounds like something I did wrong... Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] MINIT or RINIT
It seems like if I use zend_register_internal_class then I can't make use of the .function table of my class entry because zend_register_internal class null's this out. Is there an entirely good reason for this? I want the internal class to have its functions registered seperatly with zend_hash_add(class_entry.function_table, But I don't want the class damaged at RSHUTDOWN. Do these desires conflict? At least I will have to reproduce much of zend_register_internal_class but preserve the function table. Are there any other side effects of being ZEND_INTERNAL_CLASS instead of ZEND_USER_CLASS? Perhaps it will be safer to register as ZEND_USER_CLASS in RINIT instead of ZEND_INTERNAL_CLASS in MINIT? Sam -Original Message- From: Sam Liddicott Sent: 23 April 2002 15:50 To: Rasmus Lerdorf; Sam Liddicott Cc: PHP Developers Mailing List Subject: RE: [PHP-DEV] MINIT or RINIT -Original Message- From: Rasmus Lerdorf [mailto:[EMAIL PROTECTED]] Sent: 23 April 2002 11:30 To: Sam Liddicott Cc: PHP Developers Mailing List Subject: Re: [PHP-DEV] MINIT or RINIT Well, where are you storing them? Make sure they are not in memory that gets cleaned on a per-request basis. ie. do your own malloc() in MINIT and free() in MSHUTDOWN. This means my class must be declared as ZEND_INTERNAL_CLASS then? Are there any other side effects to doing this? Must I un-register my classes during shutdown? It seems zend does that for me. I've made these changes, to use ZEND_INTERNAL_CLASS and malloc and strdup instead of emalloc and estrdup but now zend hangs on shutdown. It seems to come round (under gdb) somewhere in _efree, under zend_hash_destroy under destroy_zend_class So it sounds like something I did wrong... Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] MINIT or RINIT
OK, I just noticed class_entry-builtin_functions which I can probably use. Sam -Original Message- From: Sam Liddicott Sent: 23 April 2002 16:19 To: Sam Liddicott; 'Rasmus Lerdorf' Cc: 'PHP Developers Mailing List' Subject: RE: [PHP-DEV] MINIT or RINIT It seems like if I use zend_register_internal_class then I can't make use of the .function table of my class entry because zend_register_internal class null's this out. Is there an entirely good reason for this? I want the internal class to have its functions registered seperatly with zend_hash_add(class_entry.function_table, But I don't want the class damaged at RSHUTDOWN. Do these desires conflict? At least I will have to reproduce much of zend_register_internal_class but preserve the function table. Are there any other side effects of being ZEND_INTERNAL_CLASS instead of ZEND_USER_CLASS? Perhaps it will be safer to register as ZEND_USER_CLASS in RINIT instead of ZEND_INTERNAL_CLASS in MINIT? Sam -Original Message- From: Sam Liddicott Sent: 23 April 2002 15:50 To: Rasmus Lerdorf; Sam Liddicott Cc: PHP Developers Mailing List Subject: RE: [PHP-DEV] MINIT or RINIT -Original Message- From: Rasmus Lerdorf [mailto:[EMAIL PROTECTED]] Sent: 23 April 2002 11:30 To: Sam Liddicott Cc: PHP Developers Mailing List Subject: Re: [PHP-DEV] MINIT or RINIT Well, where are you storing them? Make sure they are not in memory that gets cleaned on a per-request basis. ie. do your own malloc() in MINIT and free() in MSHUTDOWN. This means my class must be declared as ZEND_INTERNAL_CLASS then? Are there any other side effects to doing this? Must I un-register my classes during shutdown? It seems zend does that for me. I've made these changes, to use ZEND_INTERNAL_CLASS and malloc and strdup instead of emalloc and estrdup but now zend hangs on shutdown. It seems to come round (under gdb) somewhere in _efree, under zend_hash_destroy under destroy_zend_class So it sounds like something I did wrong... Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Refcount question
If I'm writing a C++ extension module and I want return_value to be the same as this_ptr do I need to increase the ref count of the zval this_ptr? Or is it safe just to do as the last line: return_Value=this_ptr; or *return_value=*this_ptr; or what ?? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] call_user_function()
What happened with this in the end.
I have a case where call_user_function needs to call a class function
without me having to instantiate a dummy object first.
Sam
-Original Message-
From: Andi Gutmans [mailto:[EMAIL PROTECTED]]
Sent: 26 February 2002 17:13
To: brad lafountain; [EMAIL PROTECTED]
Subject: Re: [PHP-DEV] call_user_function()
Can you send a unified diff of what call_user_function()
would look like
with your changes?
Andi
At 06:04 AM 2/25/2002 -0800, brad lafountain wrote:
Ok,
I know ive already posted this but no responded. So I'll
do it again.
Currently call_user_function() doesn't call overloaded class methods.
I hacked together a function to get around this. Does anyone
want to look
at it
and see if it's gunna mess anything up as far as zend
execution. if not put it
in as part as the normal call_user_function.
int my_call_user_function(HashTable *function_table, zval
**object_pp, zval
*function_name, zval *retval_ptr, int param_count, zval
*params[] TSRMLS_DC)
{
if(call_user_function(function_table, object_pp,
function_name,
retval_ptr,
param_count, params TSRMLS_CC) == FAILURE)
{
if(Z_OBJCE_PP(object_pp)-handle_function_call != NULL)
{
zend_overloaded_element overloaded_element;
zend_property_reference property_reference;
zend_function_state function_state;
zend_function_state
*original_function_state_ptr;
int i;
overloaded_element.element = *function_name;
overloaded_element.type = OE_IS_METHOD;
function_state.function = (zend_function *)
emalloc(sizeof(zend_function));
function_state.function-type =
ZEND_OVERLOADED_FUNCTION;
function_state.function-common.arg_types = NULL;
function_state.function-overloaded_function.function_name =
Z_STRVAL_P(function_name);
property_reference.object = *object_pp;
property_reference.type = BP_VAR_NA;
property_reference.elements_list =
(zend_llist
*)emalloc(sizeof(zend_llist));
zend_llist_init(property_reference.elements_list,
sizeof(zend_overloaded_element), NULL, 0);
zend_llist_add_element(property_reference.elements_list,
overloaded_element);
//Build argument stack
for(i = 0;i param_count;i++)
zend_ptr_stack_push(EG(argument_stack),
params[i]);
zend_ptr_stack_n_push(EG(argument_stack), 2,
(void *)param_count, NULL);
EG(function_state_ptr) = function_state;
Z_OBJCE_PP(object_pp)-handle_function_call(param_count, retval_ptr,
*object_pp, 1 TSRMLS_CC, property_reference);
EG(function_state_ptr) =
original_function_state_ptr;
zend_llist_destroy(property_reference.elements_list);
efree(property_reference.elements_list);
efree(function_state.function);
zend_ptr_stack_clear_multiple(TSRMLS_C);
return SUCCESS;
}
}
}
__
Do You Yahoo!?
Yahoo! Sports - Coverage of the 2002 Olympic Games
http://sports.yahoo.com
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] call_user_func_array bug
call_user_func_array and family support methods and functions but not class methods. I'm using swig to generate the php module for xapian (was OpenMuscat) and want to patch swig so that methods shadow classes can take variable numbers of arguments. The shadow methods need to pass on the same number of arguments using get_args() and call_user_func_array, but can't because call_user_func_array can't call a class method. Is there a good reason why call_user_func_array doesn support class methods? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
[PHP-DEV] Autoconf 2.52 and Open Muscat
Yes!! The good news is that what was open muscat but that is now xapian (www.xapian.org) is having it's php module finished this week! [The muscat module will expire and be replaced by Xapian, a much more comprehensive module] The trouble is Xapian prefers autoconf 2.52 as it provides large-file-support test macros for all platforms; but php chokes on this version of autoconf for some reason; the same reason many other packages choke on it; e.g.: http://archive.linuxfromscratch.org/mail-archives/blfs-support/2001/11/0014. html So I thought I would bring the pendig autoconf 2.52 changes to the attention of those concerned and then leave them to address it at their leisure. Thanks guys. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again)
-Original Message- From: Hartmut Holzgraefe [mailto:[EMAIL PROTECTED]] Sent: 01 February 2002 11:20 To: Sam Liddicott Cc: Stig S. Bakken; Manuel Lemos; [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again) Sam Liddicott wrote: have you had a look on php4/Zend/LICENSE lately ? ;) Yes, I checked it minutes before sending out my mail. Have you? so you've noticed the difference between CVS revision 1.5 and 1.6 of that file? No I checked the one on the web site, url provided at the time, but you could stop playing games and get to the point. [goes off to check] I'm not sure which file you meant by that file. http://cvs.php.net/cvs.php/php4 LICENSE 1.17 imajes 3 daysupdating license to 2002 (hasn't been done since 2000) Not version 1.5 or 1.6 and no significant changes. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again)
-Original Message- From: Hartmut Holzgraefe [mailto:[EMAIL PROTECTED]] Sent: 31 January 2002 10:15 To: Sam Liddicott Cc: Stig S. Bakken; Manuel Lemos; [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again) Sam Liddicott wrote: The Zend Open Source license http://www.php.net/license/ZendGrant/ZendGrant3.png and http://www.php.net/license/ZendGrant/ZendGrant4.png (see 5a) you MUST provide the source to software which links with it. It will take lawyers to decide if compiled php which requires the runtime to run is included. have you had a look on php4/Zend/LICENSE lately ? ;) Yes, I checked it minutes before sending out my mail. Have you? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again)
-Original Message- From: Stig S. Bakken [mailto:[EMAIL PROTECTED]] Sent: 31 January 2002 08:49 To: Manuel Lemos Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: [PHP-DEV] NEW PHP standalone compiler (was Re: [PHP-DEV] Here we try again) Unless I'm mistaken, Java doesn't have eval. Any language with eval needs a runtime environment when executing, even if the rest of the code is converted to C. Which has licensing implications due to the terms of the PHP License. The Zend Open Source license http://www.php.net/license/ZendGrant/ZendGrant3.png and http://www.php.net/license/ZendGrant/ZendGrant4.png (see 5a) you MUST provide the source to software which links with it. It will take lawyers to decide if compiled php which requires the runtime to run is included. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] maybe serious error in RC3 memory manager
-Original Message- From: Stig Venaas [mailto:[EMAIL PROTECTED]] Sent: 26 November 2001 10:22 To: Oleg Khimich Cc: [EMAIL PROTECTED] Subject: Re: [PHP-DEV] maybe serious error in RC3 memory manager On Mon, Nov 26, 2001 at 11:38:52AM +0200, Oleg Khimich wrote: On my FreeBSD 4.3 php4.1.0rc3 not failing, but message maximum execution time... appeared. PHP4.2.0-dev core dumped after some time. Good to see I'm not the only one. I see now that I should have read the CGI quick cleanup thread. I think this should be fixed, but actually when running as CGI/standalone it would be good to just exit without freeing all the memory the nice way. In that thread it's suggested to turn off fast cache, but couldn't we by default exit quickly when run as standalone? I think it should; turning off fast-cache was just a quick fix which I haven't tried in case it slows down access to all those array items; not being sure what fast cache is actually for. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] CGI quick cleanup
I am using PHP for a system script to import TV listings to a database. It works well but creates 20,000 to 30,000 hash elements with various relationships and it *seems* to take exponentially longer for the script to actually stop running after it reaches the exit statement the more items there are. I also note that even though max execution time is set to zero and the script runs for a long time, I still get a max-execution error at 30 seconds after the script has been trying to exit (although the script runs for many minutes). 30 seconds is the timeout in the php.ini which I override with ini_set. I think this long shutdown time is PHP's per-page cleanup freeing all the objects, which is not so important if running as a CGI. Currently I am quitting with posix_kill(posix_getpid(),5) as it is pretty quick but I wondered if the per-page cleanup which is only important when running as an apache module could be disabled in cgi mode. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] CGI quick cleanup
I should add that without the kill -5 it takes longer for the script to finally exit than it took to execute! Sam -Original Message- From: Sam Liddicott [mailto:[EMAIL PROTECTED]] Sent: 23 November 2001 10:35 To: [EMAIL PROTECTED] Subject: [PHP-DEV] CGI quick cleanup I am using PHP for a system script to import TV listings to a database. It works well but creates 20,000 to 30,000 hash elements with various relationships and it *seems* to take exponentially longer for the script to actually stop running after it reaches the exit statement the more items there are. I also note that even though max execution time is set to zero and the script runs for a long time, I still get a max-execution error at 30 seconds after the script has been trying to exit (although the script runs for many minutes). 30 seconds is the timeout in the php.ini which I override with ini_set. I think this long shutdown time is PHP's per-page cleanup freeing all the objects, which is not so important if running as a CGI. Currently I am quitting with posix_kill(posix_getpid(),5) as it is pretty quick but I wondered if the per-page cleanup which is only important when running as an apache module could be disabled in cgi mode. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED] -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] CGI quick cleanup
-Original Message- From: Teodor Cimpoesu [mailto:[EMAIL PROTECTED]] Sent: 23 November 2001 10:54 To: [EMAIL PROTECTED] Subject: Re: [PHP-DEV] CGI quick cleanup Hi Edin! On Fri, 23 Nov 2001, Edin Kadribasic wrote: I have noticed the same problem with a scipt that used a very large array (~160 MB). The script run time was around 35 seconds, while it took over 4 minutes to shut down! Same amount of time was used in trying to unset() the array. [wild guess] probably the memory deallocation is the one expensive here. BTW, don't use large arrays in PHP as Apache module cause when the process grows requesting more memory (brk(2)) it cannot release it back to the system (because the httpd process is still running). To solve THIS problem I added ap_child_terminate support to php so you could request apache to terminate after the page, though this seems to have been dropped in later releases. The patch was accepted and did appear in at least one release, I don't know why it went. But our problem is not as an apache module but as a system script, using php -q Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] CGI quick cleanup
-Original Message- From: Markus Fischer [mailto:[EMAIL PROTECTED]] Sent: 23 November 2001 11:16 To: Sam Liddicott Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: [PHP-DEV] CGI quick cleanup On Fri, Nov 23, 2001 at 11:01:03AM -, Sam Liddicott wrote : To solve THIS problem I added ap_child_terminate support to php so you could request apache to terminate after the page, though this seems to have been dropped in later releases. The patch was accepted and did appear in at least one release, I don't know why it went. I found this in php4/sapi/apache/php_apache.c : [apache_child_terminate function deleted] heh! Thanks. However, it doesn't seem to be documented (couldn't find it in the manual). But yes, not really relevant because you're using a system script. Hmm. What do I need to do to get this documented then? Thanks Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] CGI quick cleanup
Here's a sample script that does most of the sorts of stuff I was doing
apart from database work.
Note how long it takes to exit after finishing.
#! /usr/bin/php -q
?php
ini_Set(max_execution_time,0);
ini_Set(memory_limit,500M);
class thingy {
function thingy($c) {
if ($c0) $this-ref=new thingy($c-1);
}
}
$stash=array();
$max=50;
$start=time();
for($i=0;$i$max;$i++) {
$r=rand(0,300);
$stash[$r][]=new thingy(rand(0,10));
echo \rUse: .floor($i/$max*100).% ;
}
echo \n;
$mid=time();
$max=count($stash);
$c=0;
foreach(array_keys($stash) as $key) {
unset($stash[$key]);
$c++;
echo \rFree: .floor($c/$max*100).% ;
}
unset($stash);
echo \n;
$done=time();
print Use: .($mid-$start).\n;
print Free: .($done-$mid).\n;
?
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] german umlauts!
-Original Message- From: Andre Lacour [mailto:[EMAIL PROTECTED]] Sent: 03 September 2001 09:28 To: Php-Dev Subject: [PHP-DEV] german umlauts! strtoupper() strtoulower() does not work with german umlauts (äöü/ÄÖÜ). Perhaps it depends on what your php locale is (as opposed to your web page characterset). Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] SLEEP and MAXEXECUTIONTIME and NAUGHTY modules
-Original Message-
From: Andi Gutmans [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, February 27, 2001 08:03
To: Sam Liddicott; [EMAIL PROTECTED]
Subject: Re: [PHP-DEV] SLEEP and MAXEXECUTIONTIME and NAUGHTY modules
At 12:11 PM 2/26/2001 +, Sam Liddicott wrote:
I notice that usleep and sleep both clash on solaris with
Max-execution-time.
1) I notice various modules using sleep:
db, hyperwave, mysql, oci8, satellite
which seems to break max execution time.
Any idea why and where they are using sleep()?
Further investigation seems to reveal that use of alarm in OCI8 (oracle
supplied libs actually) on solaris combined with lwp_alarm and oracle
threading issue are breaking max execution time. However to answer the
question:
db: db.c in php_dbm_open() for an NFS locking loop:
if (lock) {
...
#if NFS_HACK
...
php_sleep(1)
hyperwave: hg_comm.c in hg_read()/hg_write(), php_sleep used with
non-blocking socket to allow up to 5 seconds to try and read. IMHO they
should either block or return what bytes are available and let the user deal
with it. 5 seconds is pretty arbitrary.
With mysql one use is a BSD threading broken select hack so I guess its no
issue for now.
The other mysql use is in my_write.c function my_write and is used to allow
a full file system to empty somewhat so the write can succeed.
oci8: oci8.c calls php_sleep(3) in oci_failover_callback()
satellite was a read herring.
standard: uniqid.c in PHPFUNCTION(uniqid) uses usleep(1) to wait for more
entropy, usleep also uses sigalarm and may break other alarms on some
systems.
It appears that all these uses of sleep could be replaced by the #define
sleep with the select function below
2) Could we not use:
#include unistd.h
#define sleep(n) \
{ struct timeval tv; \
tv.tv_sec=n; \
tv.tv_usec=0; \
select(0,NULL,NULL,NULL,tv); \
}
Where did you see that?
I just made it up; left over from my perl days where
select(undef,undef,undef,$msecs); was traditionally used for small sleeps.
It compiles into PHP fine that way on solaris, but #include unistd.h is
needed early because it (on Solaris) declares sleep and sleep being #defined
first breaks the declaration.
I'm not entirely happy with it, and other uses of ALARM will break max
execution time anyway...
Maybe the above should be PHP_SLEEP which I note some modules use, then we
can persuade developers to use PHP_SLEEP instead.
Not ot mention, I put it into main/php_config.h (after running configure)
but I'm not sure if this is included by all units which use sleep. Hmmm...
What do you think?
Sam
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] SLEEP and MAXEXECUTIONTIME and NAUGHTY modules
I notice that usleep and sleep both clash on solaris with
Max-execution-time.
1) I notice various modules using sleep:
db, hyperwave, mysql, oci8, satellite
which seems to break max execution time.
2) Could we not use:
#include unistd.h
#define sleep (n) \
{ struct timeval tv; \
tv.tv_sec=n; \
tv.tv_usec=0; \
select(0,NULL,NULL,NULL,tv); \
}
as well as being more accurate it would not break max-execution-time.
And while we are at it define php_msleep for milliseconds?
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] RE: PHP 4.0 Bug #9064 Updated: Functions inside functions cause errors if the outer function is called twice.
-Original Message-
From: Lars Torben Wilson [mailto:[EMAIL PROTECTED]]
Sent: Friday, February 02, 2001 09:49
To: Christopher Kings-Lynne
Cc: Bug Database
Subject: Re: [PHP-DEV] RE: PHP 4.0 Bug #9064 Updated: Functions inside
functions cause errors if the outer function is called twice.
What does that buy you that this doesn't?
function test2() {
echo "hi";
}
function test() {
test2();
}
...other than a new scoping rule?
Don't underestimate scoping, it means you can have multiple test2()'s [lest
call it "sort" - common to use different sort routines in different
circumstances]
a) stops namespace polution
b) keeps access private
while
c) retaining modularisation.
Sam
Christopher Kings-Lynne writes:
Shouldn't you at least be able to register the "nested
function" as 'static?
This causes syntax errors in 4.0.4pl1.
ie:
?php
function test() {
static function test2() {
echo "hi";
}
test2();
}
test();
test();
test();
?
-Original Message-
From: Bug Database [mailto:[EMAIL PROTECTED]]
Sent: Friday, February 02, 2001 5:14 PM
To: [EMAIL PROTECTED]
Subject: PHP 4.0 Bug #9064 Updated: Functions inside
functions cause
errors if the outer function is called twice.
ID: 9064
Updated by: hholzgra
Reported By: [EMAIL PROTECTED]
Old-Status: Open
Status: Closed
Bug Type: Scripting Engine problem
Assigned To:
Comments:
php has no "nested function" feature like you'll find in PASCAL
or in gcc if you enable the feature
what happens here is that test2 is registered on execution of
test in the global namespace, and when you call test a second time
it cannot register test2 again
(the engine shouldn't accept nested functions at all i think)
Previous Comments:
--
-
[2001-02-01 22:18:32] [EMAIL PROTECTED]
This script fails with:
bFatal error/b: Cannot redeclare test2() in b-/b on line
b4/bbr
?php
function test() {
function test2() {
echo "hi";
}
test2();
}
test();
test();
test();
?
--
-
Full Bug description available at: http://bugs.php.net/?id=9064
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail:
[EMAIL PROTECTED]
--
++
|Torben Wilson [EMAIL PROTECTED]Adcore Finland|
|http://www.coastnet.com/~torbenhttp://www.adcore.com|
|Ph: 1.604.709.0506 [EMAIL PROTECTED]|
++
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail:
[EMAIL PROTECTED]
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] gawk at awk
./configure selects gawk over awk if present (very nice) but buildconf does not, it uses awk over gawk, but on solaris awk has serious line-length limits which prevent buildconf from working. Is there any reason not to try and use gawk first for buildconf? Has anybody replaced solaris awk with gawk - any reason why I should not do the same? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] Destruct order/What is reference counted?
-Original Message- From: Sam Liddicott [mailto:[EMAIL PROTECTED]] Sent: Monday, January 29, 2001 05:22 To: [EMAIL PROTECTED] Subject: [PHP-DEV] Destruct order/What is reference counted? Are resources refence counted, or variables? I have a muscat_module method I wrote which is of the form: $new_resource=make_new_object($old_resource); each resource is a C++ object, and I have to stop $old_resource being disposed of until $new_resource has also been disposed of. To be more accurate, when old_resource is disposed of it will dispose of any remaining underlying new_resource objects, leaving a load of invalid new_resource's which will then be free'd later - unsuccessfully. So I either need to preserve the life of the underlying old_resource until all visible new_resources are finished, or find some way to mark the remaining new_resources as invalid. It seems like there is no destroy order at end-of-page time? Whats the best way to enforce it? I think I am going to have to wrap all the C++ objects in more C++ to implement reference counting- aggh. Surely there is another way? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED] -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] Destruct order/What is reference counted?
Are resources refence counted, or variables? I have a muscat_module method I wrote which is of the form: $new_resource=make_new_object($old_resource); each resource is a C++ object, and I have to stop $old_resource being disposed of until $new_resource has also been disposed of. This is a strict condition, even at end-of-php-page-mass-destruction time. It seems like there is no destroy order at end-of-page time? Whats the best way to enforce it? I think I am going to have to wrap all the C++ objects in more C++ to implement reference counting- aggh. Surely there is another way? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] php eating carriage returns?
-Original Message- From: Monte Ohrt [mailto:[EMAIL PROTECTED]] Sent: Friday, January 26, 2001 03:17 To: [EMAIL PROTECTED] Subject: [PHP-DEV] php eating carriage returns? example: This is a test ? print "testing1"; ? ? print "testing2"; ? ? print "testing3"; ? When I run this from the command line (4.0.5dev) I get the following output: This is a test testing1testing2testing3 done. Why are there not carriage returns between the print statements? Is this an intended behavior? Yes. 1) Its not carriage returns but br (and others) which start new lines for html 2) If you want a carriage return (or linefeed as is usual) do print "Blah\n"; to start a new line in html and text, print "blahbr\n"; Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] le_ why static?
Why are le_* garbage destructor handles declared static? As far as I can tell this means other c files in the same module can't "extern" it but as it is static in file scope it has none of the other benefits of being static? Is it to do with threading? What? Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] c++ files as module
-Original Message- From: Sascha Schumann [mailto:[EMAIL PROTECTED]] Sent: Monday, January 22, 2001 03:13 To: Sam Liddicott Cc: [EMAIL PROTECTED] Subject: RE: [PHP-DEV] c++ files as module Is: AC_ADD_LIBRARY(stdc++) also my responsibility in config.m4? [My associated 3rd party .a's seem to require it] Not necessarily. Do you need to perform any linkage tests in your config.m4 with that third party library? No, Idon't do any linkage tests; but if I don't do it, at php link time I get errors of various istream, ostream etc unresolved links. If I do it, it links fine. Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] c++ files as module
Yes, the 3rd example worked - but then so did taking it all out altogther.
I've been having various linking problems all day as I try and combine c and
cpp modules with global variables and such, and can only think that various
use of extern "C" { and such have been causing problems.
So for now, I don't seem to explicilty mention stdc++ at all and it all
links fine
I'll get back if I discover something else.
THanks
Sam
-Original Message-
From: Sascha Schumann [mailto:[EMAIL PROTECTED]]
Sent: Monday, January 22, 2001 03:52
To: Sam Liddicott
Cc: [EMAIL PROTECTED]
Subject: RE: [PHP-DEV] c++ files as module
No, Idon't do any linkage tests; but if I don't do it, at
php link time I
get errors of various istream, ostream etc unresolved
links. If I do it, it
links fine.
Perhaps we should provide a macro for that purpose, as the
standard C++ library is available under various names:
AC_CHECK_LIB(C, cin)
AC_CHECK_LIB(g++, cin)
AC_CHECK_LIB(stdc++, cin)
AC_CHECK_LIB(cxx, cin)
I think it would be fine, if you would use the third test
until we decide to improve this area.
- Sascha
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail:
[EMAIL PROTECTED]
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] linux c++ __rethrow link error
I get this error when compiling my muscat module for php on linux but no solaris: /tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/muscat/muscatapi.cpp:107: undefined reference to `__rethrow' Needless to say I don't refer to __rethrow anywhere; the actual line is: muscatapi = new MuscatAPIlocal(muscatHome,memSize); and for the purposes of this test was the last line of the function. Has anyone come accross this with c++ module before? Thanks Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
[PHP-DEV] c++ files as module
I'm trying to add some .cpp files to the muscat module, the hard part is getting them compiled! My Makefile.in is: LTLIBRARY_NAME= libmuscat.la LTLIBRARY_SOURCES = muscat.c muscat_net.c LTLIBRARY_SOURCES_CPP = muscatapi.cpp LTLIBRARY_SHARED_NAME = muscat.la LTLIBRARY_SHARED_LIBADD = $(MUSCAT_SHARED_LIBADD) LTLIBRARY_OBJECTS_X = $(LTLIBRARY_SOURCES_CPP:.cpp=.lo) include $(top_srcdir)/build/dynlib.mk I've tried to rip bits off qtdom module, one of the few others that uses C++; but no joy. Without the LTLIBRARY_OBJECTS_X it doesn't build muscatapi.o from muscatapi.cpp at all; with it I get: make[4]: Entering directory `/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/muscat' I. -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/muscat -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/main -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4 -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/Zend -I/usr/local/empower/include -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/mysql/libmysql -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/xml/expat/xmltok -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/ext/xml/expat/xmlparse -I/tmp/muscat-php.1.1/1.1.C010/10/php-4.0.4/TSRM -DXML_BYTE_ORDER=12 -c muscatapi.cpp touch muscatapi.lo /bin/sh: I.: command not found make[4]: [muscatapi.lo] Error 127 (ignored) Whats the official rules on Makefile.in to have .cpp files build, AND where SHOULD I have looked to find this. Thanks Sam -- PHP Development Mailing List http://www.php.net/ To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] To contact the list administrators, e-mail: [EMAIL PROTECTED]
RE: [PHP-DEV] Octal numbers and automatic conversions
I remember this showing up years ago in an online java mortgage calculator
giving embarrasingly far-out results.
Only there the bug was the "other way round".
In this case my opinion is that it is behaving as expected and no error.
If the programmer writes 070 he probably means octal, but that doesn't mean
every user and zero-padded string means octal.
I hope the behaviour is not changed; or we may get bad mortage calculators
again.
Sam
-Original Message-
From: Sergio Bruder [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, January 16, 2001 02:01
To: [EMAIL PROTECTED]
Subject: [PHP-DEV] Octal numbers and automatic conversions
This snippet shows what I consider a problem: The automatic conversion
string-integer doesnt interprets '070' as an octal, giving
57 in the first
echo (correct) and 71 in the second (wrong).
?
echo (070 +1). "BR\n";
echo ('070'+1). "BR\n";
?
Sergio Bruder
--
(http://distro.conectiva.com.br
)) (tm) http://sergio.bruder.net
|""|-.http://pontobr.org
|__|-'[EMAIL PROTECTED], [EMAIL PROTECTED]
--
pub 1024D/0C7D9F49 2000-05-26 Sergio Devojno Bruder
[EMAIL PROTECTED]
Key fingerprint = 983F DBDF FB53 FE55 87DF 71CA 6B01
5E44 0C7D 9F49
sub 1024g/138DF93D 2000-05-26
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail:
[EMAIL PROTECTED]
--
PHP Development Mailing List http://www.php.net/
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
