[Development] utf-8 BOM and parsers

[Knoll Lars]

On 22/04/14 16:36, "Thiago Macieira" <thiago.macieira at intel.com> wrote:

>Em ter 22 abr 2014, às 12:35:33, Knoll Lars escreveu:
>> Hi,
>> 
>> Just came back from vacation today.
>> 
>> Unfortunately BOM’s at the beginning of files seem to still be used
>>quite
>> a bit esp. in the Windows world. So I would actually vote for option 1
>>and
>> rather keep compatibility. Reason is that stripping the BOM will not
>>break
>> anything, but leaving it in will.
>> 
>> We could also consider using our builtin utf8 decoder for all utf8
>> locales, so that we don’t use iconv or ICU if the locale is utf-8 (and
>> thus always strip the BOM). That would at least give us consistent cross
>> platform behaviour.
>
>I'll send the update to the release branch in the next few hours.

Thanks!

Lars