Hello there,
I'm trying to build a conversion chain from docx- to PDF-documents. I
decided to try an OpenOffice-service which is used by some converter
scripts. By the way: has anybody tried OOo 3.3 RC4-7 with tables
inside a docx? The support seems to be broken (row heights are wrong).
That's why I'm using 3.2.1 right now.

What I do: I convert the docx-doc to an odt-document, then I fix some
things such as background picture settings by an external script. In
the next step, I want to convert the odt-document which really looks
good when openend by swriter to a PDF. I did some tests on my Windows
7 machine which ran fine but when switching to a production system,
things started to fail. On 5 different settings, I get 4 different
PDFs. OpenOffice-service ist started with '-accept="socket,port=8100;
urp;" -headless', for conversion under Windows 7 and Ubuntu I use a
python script (http://www.artofsolving.com/opensource/pyodconverter),
for Mac I have to use the Java program from the same page. The virtual
machines were launched on VirtualBox 3.2:

This is where you can download the odt file and all the results:
http://aplh.de/marscel/3dc/pdfs.zip

system 1) Windows 7 64bit, python script: PDF looks good, size is
172KB (script_py_win.pdf)
system 2) Ubuntu SE 64 bit as VM on 1), python script: PDF looks good,
size is 156KB (script_py_ubuntu_home.pdf)
system 3) Mac OSX 10.5.8, Java converter:
- exported from swriter via export-button: PDF looks good, size is
172KB (direct_mac.pdf)
- converted by the Java program: the image on the last page is
missing!, size is 136KB (script_java_mac.pdf)
system 4) Ubuntu SE 64 bit as VM on 3), python script: the text on the
first page floats into a second page, image on the last page is
missing, BUT size is 332KB (script_py_ubuntu.pdf)

What's up here? I have no clue why there's such a difference in
exporting a PDF directly or by script or why one of two fresh Ubuntu
machines is producing a much bigger and broken PDF but the other one
doesn't (any dependencies?). Does anybody know this problem? For the
time OpenOffice.org 3.3 is not usable for this task (the mentioned
table problem; but these issues seem to be gone), I'd really love to
have a reliable converter chain. Or any workaround to be applied on
the underlying OpenDocument-markup for this to work in any case.

Thanks in advance,
Marcel

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to