I think this is actually a desired behavior, I do not want to change
how the existing code works in patch level release. Otherwise the
patch is fine and I've already applied it to the CVS.
On 22-Dec-06, at 10:20 AM, Matt Wilmas wrote:
> Hi Ilia,
>
> Oh I know that. :-) I'm just talking about if there's a backslash
> as the
> last character. For example, date('U\\') (U\ passed after script
> parsing),
> the backslash is gone, which doesn't seem correct... (Actually, it
> "appends" the '\0'.)
>
>
> Matt
>
>
> ----- Original Message -----
> From: "Ilia Alshanetsky"
> Sent: Friday, December 22, 2006
>
>> Matt,
>>
>> The backslash removal is actually intentional. It is designed to
>> handle instances such as "\d", in which case the format value is
>> printed as a literal "d".
>>
>>
>> On 22-Dec-06, at 12:06 AM, Matt Wilmas wrote:
>>
>>> Hi,
>>>
>>> Just a simple patch for 5.2 to save smart_str_appends()'s strlen()
>>> calls in
>>> date_format(). It's already done in HEAD. Seems pretty safe for
>>> 5.2.1. :-)
>>>
>>> I also noticed that if the format string contains a backslash as
>>> the last
>>> character, it is lost (should be printed right?). So I changed
>>> that too and
>>> attached the patch for HEAD.
>>>
>>> 5.2's patch:
>>> http://realplain.com/php/date_optimization_5_2.diff
>>>
>>>
>>> Matt
>>
>> Ilia Alshanetsky
>
>
Ilia Alshanetsky
