Hello Rui,
many thanks!!!
Sunday, March 23, 2008, 12:40:59 PM, you wrote:
> Hello Marcus,
> I try to prepare some short test scripts for zend-multibyte stuff.
> It is like,
> <?php
> declare(encoding='Shift_JIS');
> $s = "表"; // 0x95+0x5c
> printf("%x %x\n",ord($s{0}),ord($s{1})); // expected: 95 5c
?>>
> The problems to prepare the test scripts are,
> 1. php binary should be compiled with --enable-zend-multibyte option.
You can provide a --SKIPIF-- section to detect MB support
(http://qa.php.net/write-test.php ).
--SKIPIF--
<?php
if (!in_array("detect_unicode", array_keys(ini_get_all()))) {
die "skip Requires --enable-zend-multibyte option";
}
?>
> 2. the output of script is depends on mbstring.internal_encoding option.
> To obtain the expected output (95 5c in the example),
> mbstring.internal_encoding=SJIS should be defined in php.ini or
> a runtime option like,
> 'php -d mbstring.internal_encoding=SJIS test.phpt'
> mbstring.internal_encoding cannot be specified in the script by
> ini_set().
There is the --INI-- section in .phpt's (http://qa.php.net/write-test.php ).
--INI--
mbstring.internal_encoding=SJIS
I have created a new directory Zend/tests/multibyte and added a first test
based on your script above. Please provide more of those.
I am CC'ing internals@ so that more people see how to write tests for it
in the hope of getting more tests.
Marcus
> On Sat, 22 Mar 2008 15:12:58 +0100
> Marcus Boerger <[EMAIL PROTECTED]> wrote:
>> Hello Rui,
>>
>> what we really need is tests. If we do not get tests we will keep
>> multibyte support out of PHP. Maybe you can mail some Asian user groups who
>> probably use this feature to provide some .phpt files?
>>
>> marcus
>>
>> Tuesday, March 4, 2008, 12:06:58 AM, you wrote:
>>
>> > Johannes,
>>
>> > I send a short presentation material (php-i18n-20050227.ppt)
>> > which I made it for the meeting with Zeev.
>>
>> > Please look page 6, there are short descriptiion about the current
>> > implementation.
>>
>> > Some multibyte encodings like ShiftJIS (Japanese) and Big5 (Chinese)
>> > are including 0x5c as secodnig byte of some multibyte characters.
>> > In the current implementation, pre-scanner and after-scanner encoding
>> > conversion is applied (er. Shift-JIS -> UCS-2 -> Shift-JIS)
>> > because flex based scanner cannot handle these characters.
>>
>> > Rui
>>
>> > On Mon, 03 Mar 2008 20:43:23 +0100
>> > Johannes Schlueter <[EMAIL PROTECTED]> wrote:
>>
>> >> Rui,
>> >>
>> >> as you maybe have read in [1] Marcus and a few other guys are currently
>> >> working on replacing the flex based parser by one based on re2c. In the
>> >> current development version --enable-zend-multibyte is ignored and we
>> >> (see internals list) have no clue what exactly it is meant to do and
>> >> what won't be possible without that switch. Since you're listed as one
>> >> of the authors of zend_multibyte.c and are still working on mbstring I
>> >> was wondering if you might have some examples of stuff which won't work
>> >> without --enble-zend-multibyte, best would be .phpt tests, so we can
>> >> make sure we don't break anything if we go and replace the scanner.
>> >>
>> >> [1] http://news.php.net/php.internals/35891
>> >>
>> >> Thanks,
>> >> johannes
>> >>
>> >>
--
Best regards,
Marcus mailto:[EMAIL PROTECTED]
--
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php
