convert docx to pdf creates a document which display all []

fdong fuchun.dong at
Fri Mar 13 14:01:33 PDT 2015

I need help.

After spent many days of struggling, finally I built libreoffcie- on
Arch linux. I built it without X enabled so I can't use the UI to see is
anything wrong apparently.

I run the command line on new built software: soffice --headless
--convert-to pdf my.docx 
(here my.docx is a regular English document with size 19K) 

my.pdf has been created successfully. but the display in Adobe Reader looks
strange: The structure of pdf resembles original docx format. but all
characters in pdf are display as [], in Adobe Reader. 

I used an online tool which claims to extract text from pdf file. I got back
space plus FFFFFFFFFF...

I did the compare with the result from running the same command against
libreoffice installed on Mac. Here is the summary: 

                                                    Original file size
(.docx)        size of pdf after convert (.pdf) 
using software built on linux box:       19K                                         
using software installed on Mac:         19K                                         

I did further testing:
1. convert docx to txt or html, there is no issue
2. create a txt document which contains one character, after convert to pdf
file. The display is a single []

I used pacman to install required libraries first. and build libreoffice.

Following is the options in command line:
                --with-build-version="${_LOver} Arch Linux build-${pkgrel}"
--with-vendor="Arch Linux"

Any idea how this happened or how to debug the issue (which module ? Is
vcl/source/gdi/pdfwriter_impl.cxx the correct place to look at it?).

View this message in context:
Sent from the Dev mailing list archive at

More information about the LibreOffice mailing list