On Wed, Apr 23, 2014 at 12:37:26AM +0200, drago01 wrote:
On Tue, Apr 22, 2014 at 11:49 PM, Michal Jaegermann
<michal(a)harddata.com> wrote:
> On Tue, Apr 22, 2014 at 07:30:35PM +0000, Jack Peirce wrote:
>>
>> No, that's just the way pdfseparate works. Separates into individual files
>>
>> May not be ideal, but you can use pdfunite with it to achieve what
>> you're trying to do.
>
> What if you are trying to do those other things that pdftk is capable
> of doing and they are not just splitting and joining pages?
Port pdftk to openjdk? Not sure why it depends on gcj didn't look at
it in detail.
I am not sure if there is really anything to port. At
http://www.pdflabs.com/docs/install-pdftk-on-redhat-or-centos/
you can find pdftk-2.02-1.el6.src.rpm. I do not now about rawhide
but out of curiosity I tried on a Fedora 20 installation:
rpmbuild --rebuild pdftk-2.02-1.el6.src.rpm
That whines profusely, as to be expected from a thingy using Java,
but after a not so long while it produces
pdftk-2.02-1.fc20.x86_64.rpm
pdftk-debuginfo-2.02-1.fc20.x86_64.rpm
A spec for all this is somewhat broken as every single file, with an
exception of .../man1/pdftk.1.gz, gets installed with -rwxr-xr-x
permissions. This is trivially fixable. OTOH after light testing
pdftk binary seems to work just fine and it appears to be somwewhat
faster then the one from pdftk-1.44-11.fc19.x86_64 (the last one
available on Fedora 64-bits mirrors) although I did not try to do any
measurements.
Oh, and even in a limited application pdfseparate/pdfunite quickly
becomes a major PITA if your source file has more than nine pages.
Extracting some pieces from a few hundreds pages long text could become
"interesting".
Michal
p.s. URL above has also links to ready binaries for RHEL5, RHEL6,
CentOS5 and CentOS6. I would be a bit leery to install those "as is"
but in any case they are there.