[reportlab-users] ReportLab 2.7 accelerators
Andy Robinson
andy at reportlab.com
Thu Apr 4 18:17:39 EDT 2013
This is now fixed. I chose 'option 2' and uploaded the tested tar.gz file,
and have just completed an installation with 'easy_install reportlab' in a
virtualenv.
I am guessing my colleague followed 'option 1' in the PyPI instructions and
ran 'setup.py upload'. Now we know that this doesn't work, for us at
least. Clearly, 'setup.py sdist' and 'setup.py upload' have somehow
attempted to 'improve' on easily understood deployment concepts like
'tar/gzip' and 'scp'.
However, we'll be doing a lot more rearranging before the next version, and
hopefully simplifying our setup.py as we do it, so hopefully we will end up
compliant with distutils eventually.
Thanks for bringing this to our attention, and my apologies to anyone else
who bumped into this.
- Andy
On 4 April 2013 23:01, Andy Robinson <andy at reportlab.com> wrote:
> Lele, you are quite right, the file on PyPI is different to the one we
> created and tested, and completely missing the subdirectory with all the C
> code.
>
> I will have to check with my colleague who uploaded it in the morning.
> We created and tested all the packages, moved them from a hidden to a
> visible directory on reportlab.com, then attempted to follow the
> instructions on pypi. Right now I will see if I can manually put the
> right file in place...
>
> - Andy
>
>
> On 4 April 2013 22:17, Lele Gaifax <lele at metapensiero.it> wrote:
>
>> Andy Robinson <andy at reportlab.com> writes:
>>
>> > Lele, what platform are you on?
>>
>> Debian Wheezy
>>
>> > We have been testing this every way we could think of on Ubuntu, Mac OS
>> and
>> > Windows for quite a few days. If I un-tar-gz that file, there is
>> > definitely a src/rl_addons/rl_accel subdirectory containing the code,
>> and
>> > the top level 'setup.py install' worked for all of us on the above
>> > platforms. I can only presume that in your environment the setup
>> script is
>> > somehow not finding that subdirectory.
>>
>> I'd find that rather strange, but we must be talking about a different
>> file::
>>
>> $ wget
>> https://pypi.python.org/packages/source/r/reportlab/reportlab-2.7.tar.gz#md5=07f39af5aa569a16ce9a979b9a9e63a9
>> $ tar tvzf reportlab-2.7.tar.gz |grep accel
>> $
>> $ tar xOzf reportlab-2.7.tar.gz reportlab-2.7/setup.py | grep accel
>> def _find_rl_ccode(dn='rl_accel',cn='_rl_accel.c'):
>> '''locate where the accelerator code lives'''
>> RL_ACCEL = _find_rl_ccode('rl_accel','_rl_accel.c')
>> infoline( '*No rl_accel code found, you can obtain it at *')
>> infoline( '*http://www.reportlab.org/downloads.html#_rl_accel*
>> ')
>> infoline( '#Attempting install of _rl_accel, sgmlop & pyHnj')
>> Extension( '_rl_accel',
>> [pjoin(RL_ACCEL,'_rl_accel.c')],
>> infoline( '!No rl_accel code found, you can obtain it at !')
>>
>> So, effectively the setup.py tries to find the C version, but it cannot
>> find it and thus emits the message about obtaining it from the net.
>>
>> > The reduction in file size is possibly because we have been deliberately
>> > removing obsolete documentation and examples.
>>
>> I bet that there's something that should be put back :-) At least on the
>> archive uploaded on PyPI. Later I'll try to download it from rl.org
>> site.
>>
>> Thank you,
>> ciao, lele.
>> --
>> nickname: Lele Gaifax | Quando vivrò di quello che ho pensato ieri
>> real: Emanuele Gaifas | comincerò ad aver paura di chi mi copia.
>> lele at metapensiero.it | -- Fortunato Depero, 1929.
>>
>> _______________________________________________
>> reportlab-users mailing list
>> reportlab-users at lists2.reportlab.com
>> http://two.pairlist.net/mailman/listinfo/reportlab-users
>>
>
>
>
> --
> Andy Robinson
> Managing Director
> ReportLab Europe Ltd.
> Thornton House, Thornton Road, Wimbledon, London SW19 4NG, UK
> Tel +44-20-8405-6420
>
--
Andy Robinson
Managing Director
ReportLab Europe Ltd.
Thornton House, Thornton Road, Wimbledon, London SW19 4NG, UK
Tel +44-20-8405-6420
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://two.pairlist.net/pipermail/reportlab-users/attachments/20130404/98d4f713/attachment.htm>
More information about the reportlab-users
mailing list