Front page | perl.perl5.porters |
Postings from May 2013
On 05/01/2013 07:04 PM, Peter Martini wrote:
> On Wed, May 1, 2013 at 12:53 PM, Steffen Mueller
> <mail@steffen-mueller.net> wrote:
>> [side-track]
>>
>> On 05/01/2013 06:44 PM, demerphq wrote:
>>>
>>> My understanding of pack()s original intention was to allow one to
>>> construct C structures in Perl. As such introduces a change which
>>> basically means one cannot safely put a string into a C structure
>>> without the possibility of the entire structure being corrupted by
>>> doing so doesn't make sense.
>>
>>
>> In all fairness, if you use pack to do that, you'll be writing VERY
>> unportable code due to alignment, padding and friends. But don't listen to
>> me: Have a look at the docs for Convert::Binary::C. Marcus knows a lot
>> better than I do.
> For what it's worth, I often use pack/unpack when I'm working with
> network protocols. The endianness and encoding are well defined and
> may differ from the architecture of the system I'm working on. The
> main example I can think of was talking from a SPARC/Solaris or
> Intel/linux host to a Windows service, which meant not only did it it
> have to pay attention to endianness for the encoded length fields, it
> had to convert encoded text data from UTF-16 to UTF-8 as well.
That's entirely not the same as C structs. :)
This being said, I'm surprised you haven't run into the "passing a
scalar with UTF8bit set to pack("v/a")" issue. Are you running old perls?
--Steffen
Thread Previous
|
Thread Next