[PHP-DEV] Re: CVS Account Request: wudicgi

2006-09-16 Thread Pierre
Hello, On 9/15/06, Wudicgi <[EMAIL PROTECTED]> wrote: > Managing the source code of PEAR Crypt_XXTEA. Confirmed. Wudi, please mail pear-group before asking for a CVS account. --Pierre -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php

[PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Pierre
Hello, Given the recent discussion about the filter API in the pecl-dev list (http://news.php.net/php.pecl.dev/4123), I summarize again what I proposed a while back. It is very important to agree on an API now or we will have to remove filter from 5.2.0. This proposal does not reflect what we hav

[PHP-DEV] leaks related with virtual_file_ex()

2006-09-16 Thread Nuno Lopes
Hello, There are some leaks in PHP associated with virtual_file_ex() usage. The problem is that that function doesn't free the new 'cwd' on error (path not found). Taking a look at http://lxr.php.net/ident?i=virtual_file_ex we can see the function usage. For example, expand_filepath() free()s

[PHP-DEV] Re: ext/filter, Final API proposal

2006-09-16 Thread Christian Schneider
Pierre wrote: Proposal: ** I. Availalbe functions: --- * input_get Gets variable from outside PHP or from a userland variable and optionally filters it using one filter and its options or flags. It accepts only scalar by default, array can be returned using ALLOW_ARR

[PHP-DEV] Re: ext/filter, Final API proposal

2006-09-16 Thread Pierre
Hello, On 9/16/06, Christian Schneider <[EMAIL PROTECTED]> wrote: Pierre wrote: > Proposal: > ** > > I. Availalbe functions: >--- > > * input_get > Gets variable from outside PHP or from a userland variable and > optionally filters it using one filter and its options

[PHP-DEV] Re: ext/filter, Final API proposal

2006-09-16 Thread Christian Schneider
Pierre wrote: It always returns an array when FILTER_FLAG_ARRAY is given. Even when the input value was a scalar. It spares yet another is_array/isset test. And it accepts only scalar by default or when FILTER_FLAG_SCALAR is given. Makes sense but the example shows: $args = array( ... 'vers

[PHP-DEV] Re: ext/filter, Final API proposal

2006-09-16 Thread Pierre
Hello, On 9/16/06, Christian Schneider <[EMAIL PROTECTED]> wrote: Pierre wrote: > It always returns an array when FILTER_FLAG_ARRAY is given. Even when > the input value was a scalar. It spares yet another is_array/isset > test. And it accepts only scalar by default or when FILTER_FLAG_SCALAR >

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Dan Scott
On 16/09/06, Pierre <[EMAIL PROTECTED]> wrote: Proposal: ** I. Availalbe functions: --- * input_get Gets variable from outside PHP or from a userland variable and optionally filters it using one filter and its options or flags. It accepts only scalar by default, ar

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Andrei Zmievski
* input_get Gets variable from outside PHP or from a userland variable and optionally filters it using one filter and its options or flags. It accepts only scalar by default, array can be returned using ALLOW_ARRAY. What about the charset parameter for input_get()? We'll need it for PHP 6 so h

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Ilia Alshanetsky
On 16-Sep-06, at 11:35 AM, Dan Scott wrote: Minor: the general structure of the function names are __, so it would be more consistent if this was 'input_list_filters'. Overall: +1 (with bonus points if you go with __ for the two identified cases) I for one think we should try to keep the nam

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Dan Scott
On 16/09/06, Ilia Alshanetsky <[EMAIL PROTECTED]> wrote: On 16-Sep-06, at 11:35 AM, Dan Scott wrote: > Minor: the general structure of the function names are > __, so it would be more consistent if this was > 'input_list_filters'. > > Overall: +1 (with bonus points if you go with __ > for the t

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Ilia Alshanetsky
On 16-Sep-06, at 1:28 PM, Dan Scott wrote: Agreed. input_id_filter() is shorter than input_name_to_filter(), and input_list_filters() is the same length as input_filters_list(). Why not change the function prefixes to filter_*, which in turn allows for MUCH shorter name, Ex: input_list_filt

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Pierre
Hello, On 9/16/06, Dan Scott <[EMAIL PROTECTED]> wrote: On 16/09/06, Ilia Alshanetsky <[EMAIL PROTECTED]> wrote: > > On 16-Sep-06, at 11:35 AM, Dan Scott wrote: > > > Minor: the general structure of the function names are > > __, so it would be more consistent if this was > > 'input_list_filters

RE: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Andi Gutmans
Yep I agree. I thought the idea was to go to filter_* which has been our standard for extensions. I'm sure we can find good names for the functions in this way. > -Original Message- > From: Ilia Alshanetsky [mailto:[EMAIL PROTECTED] On Behalf Of > Ilia Alshanetsky > Sent: Saturday, Septe

Re: [PHP-DEV] ext/filter, Final API proposal

2006-09-16 Thread Pierre
Hello, On 9/17/06, Andi Gutmans <[EMAIL PROTECTED]> wrote: Yep I agree. I thought the idea was to go to filter_* which has been our standard for extensions. I'm sure we can find good names for the functions in this way. old name > new name: input_get > filter_get input_get_args