Juno for py3o

py3o.renderers.juno is a Java driver for py3o to transform an OpenOffice document into a PDF

It is intended to be used in conjunction with py3o.renderserver But can be used outside it if you wish.

Note for end-users

If you just search for an easy way to render LibreOffice files to PDF or DOCX, with ot without templating capabilities, you should directly look at py3o.fusion

py3o.fusion is a packaged webservice that lets you send a template, a target format and your data and will return the resulting file.


Since this is a Java implementation you will need to install jpype1 and to have a recent Java runtime on the rendering machine. You will also need a running OpenOffice instance. (If you are on windows this can be addressed by using the py3o.renderserver Open Office service.)

We have a dependency on gson from google. Which can be installed on ubuntu like this:

bash $ sudo apt-get install libgoogle-gson-java

This has been tested to build correctly with:

  • Oracle JDK 1.6 and OpenOffice 3.2.1 on Windows 7 and Windows server 2003
  • Oracle JDK 1.6 and LibreOffice 3.4 on Windows 7 64bit
  • OpenJDK 6 and LibreOffice 3.4 on Linux (Ubuntu and RHEL 5)
  • OpenJDK 7 and LibreOffice 4.0.4 on Linux (Ubuntu 13.04)
  • OpenJDK 7 and LibreOffice on Linux (Ubuntu 14.04)

For example if you are on Ubuntu you should run this command:

bash $ sudo apt-get install default-jdk


Start a JVM from python and then initialize you convertor and use it:

```python # -- encoding: utf-8 -- from py3o.renderers.juno import start_jvm, Convertor, formats import datetime

# first arg is the or .dll # second arg is the basedir where we can find the basis3.3/program/classes/unoil.jar # third argument it the ure basedir where we can find ure/share/java/*.jar containing # java_uno.jar, juh.jar, jurt.jar, unoloader.jar # the fourth argument was the openoffice version but is no more used # fifth argument is the max memory you want to give to the JVM start_jvm( "/usr/lib/jvm/java-7-openjdk-amd64/jre/lib/amd64/server/", "/usr/lib/libreoffice", "/usr/lib", "", 140) c = Convertor("", "8997")

t1 = c.convert("py3o_example.odt", "py3o_example.pdf", formats['PDF']) t2 = ```

For more information please read the example provided in the examples dir and read the API documentation.


bash $ pip install --upgrade py3o.renderers.juno


We just made a change of requirement from jpype to jpype1 in version 0.6 which should be pip installable by anyone with the correct toolchain. This means all requirements should now install automatically on a development machine.

Driver compilation and installation

NOTE: this is optionnal and reserved for developpers who want to compile the jar file by themselves

If you want to install from source you'll need to clone our repository:

bash $ hg clone $ cd py3o.renderers.juno/java/py3oconvertor $ ./ $ cd ../../ $ python develop

Please note how you must first compile the jar file with our script (some more example scripts are available for windows and OpenOffice). If something fails, first try to edit the script and find if all referenced jar files are present on your system.


By order of appearance:

  • Florent Aide
  • Thomas Desvenain
  • Eric Bréhault
  • Alexis de Lattre
  • Fabien George

Release Notes

2018-07-04 0.8.1

  • Add support for PDF options
  • Fix ods to pdf conversion

2016-11-28 0.8

  • Update java classpath depending on plateform (Thomas Desvenain & Eric Bréhault)