[TYPO3-core] RFC #9194: Bug: wrong handling of 'is_in' list in TCA element type 'input'

[Vladimir Podkovanov]

Vladimir Podkovanov wrote:
> Martin Kutschker wrote:
>> Vladimir Podkovanov schrieb:
>>> Hi!
>>>
>>> This is an SVN patch request.
>>>
>>> Type: Bugfix
>>>
>>> Bugtracker references: http://bugs.typo3.org/view.php?id=9194
>>>
>>> Branches: TYPO3_4-1, TYPO3_4-2, trunk
>>>
>>> Problem:
>>> when 'is_in' string in TCA element type 'input' ends with symbol '0'
>>> (say '1234567890') then '0' is not taken into account so 0's stripped
>>> off from field value
>>>
>>> Solution:
>>> changed logical condition to avoid evaluation of string '0' as false
>>
>> Now that you bring that up, I see that this code has another problem:
>> It's not multi-byte safe. Both the loop and the substr() assume that
>> each character is only one byte long. What should be done is to use the
>> character set defined by $LANG->charset and use it with t3lib_cs.
>>
>> Masi
> 
> Well,
> I've checked now whole class.t3lib_tcemain.php and have found two more 
> instances where strstr() or substr() using with possible multibyte 
> values. So maybe we commit currect fix and then fixing 
> class.t3lib_tcemain.php in new thread.
> 
> Vladimir

+ found some more places where using strtoupper(), strtolower(), 
strlen(), str_replace(), ereg_replace() with values from input field 
that could be in unicode and maybe some more functions should be checked

Vladimir