zopyx.convert2

The zopyx.convert2 package helps you to convert HTML to PDF, RTF, ODT, DOCX and WML using XSL-FO technology or using PrinceXML. This package is used as the low-level API for zopyx.smartprintng.core.

Requirements

  • Java 1.5.0 or higher (FOP 0.94 requires Java 1.6 or higher)
  • csstoxslfo (included)
  • XFC-4.0 (XMLMind) for ODT, RTF, DOCX and WML support (if needed)
  • XINC 2.0 (Lunasil) for PDF support (commercial)
  • or FOP 0.94 (Apache project) for PDF support (free)

Installation

  • install zopyx.convert2 either using easy_install or by downloading the sources from the Python Cheeseshop. This will install automatically the Beautifulsoup and Elementree modules if necessary.
  • the environment variable $XFC_DIR must be set and point to the root of your XFC installation directory
  • the environment variable $XINC_HOME must be set and to point to the root of your XINC installation directory
  • the environment variable $FOP_HOME must be set and point to the root of your FOP installation directory
  • the ‘prince’ binary must be in the $PATH if you are using PrinceXML

Supported platforms

Windows, Unix

Usage

Some examples from the Python command-line:

from zopyx.convert2 import Converter
C = Converter('/path/to/some/file.html')
pdf_filename = C('pdf-xinc')['output_filename']       # using XINC
pdf2_filename = C('pdf-pisa')['output_filename']      # using PISA
pdf3_filename = C('pdf-fop')['output_filename']       # using FOP
pdf4_filename = C('pdf-prince')['output_filename']    # using FOP
rtf_filename = C('rtf-xfc')['output_filename']
pdt_filename = C('odt-xfc')['output_filename']
wml_filename = C('wml-xfc')['output_filename']
docx_filename = C('docx-xfc')['output_filename']

A very simple command-line converter is also available:

html-convert --format rtf --output foo.rtf sample.html

html-convert has a –test option that will convert some sample HTML. If everything is ok then you should see something like that:

>html-convert --test
Entering testmode
pdf: /tmp/tmpuOb37m.html -> /tmp/tmpuOb37m.pdf
rtf: /tmp/tmpuOb37m.html -> /tmp/tmpuOb37m.rtf
docx: /tmp/tmpuOb37m.html -> /tmp/tmpuOb37m.docx
odt: /tmp/tmpuOb37m.html -> /tmp/tmpuOb37m.odt
wml: /tmp/tmpuOb37m.html -> /tmp/tmpuOb37m.wml
pdf: /tmp/tmpZ6PGo9.html -> /tmp/tmpZ6PGo9.pdf
rtf: /tmp/tmpZ6PGo9.html -> /tmp/tmpZ6PGo9.rtf
docx: /tmp/tmpZ6PGo9.html -> /tmp/tmpZ6PGo9.docx
odt: /tmp/tmpZ6PGo9.html -> /tmp/tmpZ6PGo9.odt
wml: /tmp/tmpZ6PGo9.html -> /tmp/tmpZ6PGo9.wml

How zopyx.convert2 works internally

  • The source HTML file is converted to XHTML using mxTidy
  • the XHTML file is converted to FO using the great “csstoxslfo” converter written by Werner Donne.
  • the FO file is passed either to the external XINC or XFC converter to generated the desired output format
  • all converters are based on Java technology make the conversion solution highly portable across operating system (including Windows)

Environment variables

The following environment variables can be used to resolve OS or distribution specific problems:

ZOPYX_CONVERT_SHELL - defaults to sh and is used to as shell command to execute external converters

ZOPYX_CONVERT_EXECUTION_MODE - default to process and refers to the method of Python executing external command (by default using the process module. Other value: system, commands

Known issues

  • If you are using zopyx.convert2 together with FOP: use the latest FOP 0.94 only. Don’t use any packaged FOP version like the one from MacPorts which is known to be broken.
  • Ensure that you have read the csstoxslfo documentation. csstoxslfo has several requirements about the HTML markup. Don’t expect that it is the ultimate HTML converter. Any questions regarding the necessary markup are documented in the csstoxslfo documentation and will not be answered.

Author

zopyx.convert2 was written by Andreas Jung for ZOPYX Ltd., Tuebingen, Germany.

License

zopyx.convert2 is published under the Zope Public License (ZPL 2.1). See LICENSE.txt.

Contact

ZOPYX Ltd.
Charlottenstr. 37/1
D-72070 Tuebingen, Germany
www.zopyx.com