PyOpenOffice 0.31

Process OpenOffice.org Writer Files and transform them to PDF without installing OpenOffice.org

	Advent and Christmas Season in the Ore Mountains ("Erzgebirge"), Saxony High Tech in Saxony

Test it "live" on this server.

FAQ:

What is PyOpenOffice?

• It is a class library, written in the Python Language.
• It is a platform-independent command-line utility (many abilities of the commandline interface are still missing)

What can PyOpenOffice do for you?

• Make serial letters and reports from templates in the SXW-Format and store them anew as SXW-Files
• Replace pictures - generate catalogues, worksheets, photo galleries etc. from an SXW-Template.
  
• Extract the pure text information from the SXW-Document and store it in a plain text file or a PDF-File
  
• Perform simple transformations from SXW-Files to PDF-Files without installing OO - for example on a webserver

At what stage of development is PyOpenOffice now?

• It is an Beta Release. Nevertheless many parts of it are a "spin-off" product of my software "Bezirksreiter" for church administration - which is considered to be stable now.
• Many features are still missing: better transformation of tables to PDF, making use of the stylesheet information in "styles.xml" and so on.

What license has PyOpenOffice?

• It is distributed under the GNU Public License

What software is required?

• Any operating system that supports the following software:
  
• Python 2.3 (maybe it works form 2.1 upwards, but this is not recommended) for the SXW-Processing
• The ReportlabToolkit and the Python Imaging Library for the transformation from SXW to PDF
• OpenOffice.org for viewing and editing the templates
• The PyOpenOffice-Zip-Package
• To get the latest (unstable) version browse the CVS tree on SourceForge.
  

What are the principles behind PyOpenOffice?

• Unpacking and Re-Packing of the SXW-Archive-File is done by the Python zipfile Module
  
• Serial letters and reports are processed by using Regular Expressions and string extract and replace operations
• Generating PDF is performed using the XML-SAX-Parser from the Python Standard Library, the ReportlabToolkit and the Python Imaging Library
• For some purposes an XML-DOM-API is used (Minidom from the Python Standard Library)
  
• Plain text extract uses  the SAX-Parser
• When transforming to PDF four trials are made: with all known attributes, without underlining, without images, plain text only.
  
• Read the sources and use pydoc from the Standard Library! But: Too many functions and comments are still in German - sorry

Usage

• Unzip PyOpenOffice.zip into your path
• Make sure that pyopenoffice.py and textutilities_pyopenoffice.py have write access to this path
  
• Write your skript and import pyopenoffice. Mind: there are still global variables in use, so you always have to use the whole module
  
• make an instance of the interface class: tool = pyopenoffice.PyOpenOffice()
• provide a replace_listofdicts, for example from an SQL-Query: [{"Surname": "Simon", "FirstName": "Martin"}, {"Surname": "Miller", "FirstName": "Steve"}]
• provide a template file with data fields: "TFA-Surname-TFE" etc.
• provide a list with picture files
  
• use a table with a row in it, containing "Multi:TFA-Surname-TFE" - the table row will be repeated, using all the records from your replace_listofdicts
• have a look at the example template "test.sxw" - run "python pyopenoffice" from the command line and check the result
• have a look at "test_pic.sxw" (it explains the replacePictures() function) - run "python pyopenoffice  -pictest" from the command line and check the result
  
• or: provide any SXW-File for transforming it to PDF, with or without template fields
  
• work with it:

Restrictions and Known Bugs

• the serial letter function will not work with frames that are anchored at the page. Use the "anchor at paragraph" option instead.

• only the "automatic-styles" of the current document are parsed - inherited styles from master documents etc. will be dropped
• all fonts are mapped to the three standard postscript fonts of PDF: Times, Helvetica and Courier. Only "pt" is accepted for fontsizes
• no page-numbering function
• headers of long tables will not be repeated on every page
  
• the "underline"-function of the ReportlabToolkit has several bugs. When difficulties occur, PyOpenOffice will change all underlines to italics. If text alignment "justified" is used in the document, this will be done by default.
• the "underline"-function should not be used with coloured background (else you will see a little white point after the underlined word)
  
• do not use frames and columns - the text within is printed as normal paragraphs (and sometimes it is just dropped...)
  
• If at least one cell has borders, the whole table is printed with a grid
• not all image formats are processed (due to the use of the Python Imaging Library)
• image sizes may only be given either in cm or inch
• only headings form heading 1 to heading 3 can be mapped - they will have always a certain standard appearence
• ........... try it out: I think nevertheless the PDFs do not look too bad.

Release History

• 0.31: new class persistentDict for the history and cleanup function (the shelve module was not always reliable). Use data with carriage returns in it (for example from a database) and make them printable in SXW or PDF - read the docstring of makeSerialLetters. Several bugfixes.
  
• 0.3 : Generate PDF with arbitrary page sizes (Letter, DIN A5, portrait or landscape etc.) and arbitrary text margins: page sizes and text margins are now taken exactly from the SXW-Template by using an XML-DOM-API (more functions using this API will follow in the next versions). PyOpenOffice now comes with two modules: pyopenoffice.py (main module) and textutilities_pyopenoffice (helper classes and functions).
  
• 0.23 Tables with different column widths now can be mapped to PDF (have a look at test.sxw). History and cleanup functions for the temporary files generated by PyOpenOffice.
  
• 0.22 Bugfixes for makeSerialLetters(): Multi:TFA-...-TFE rows do not have to be the last row of this table anymore. You may now provide sourcefiles that are not in the directory of pyopenoffice.py (in earlier releases PyOpenOffice stopped with an error). New feature for makeSerialLetters(): provide a preprocessed content.xml to the variable "content" (read the docstrings).
  
• 0.21 Coloured background for paragraphs and table cells in PDF. User-defined constants possible for makeSerialLetters() - not only "today" or "heute" allowed. Better explanations in test.sxw.
  
• 0.2 New function to replace pictures. Patch for makeSerialLetters() to make it usable with Asian fonts.  Mind: no PDF support for such fonts yet.
  
• 0.13 Bugfix for transforming tables to PDF: tables with headers are allowed now in the SXW-Template. Support for special signs like the Euro-Symbol by changing from Latin-1 to cp1252-transformation.
  
• 0.12 Bugfix for the makeSerialLetters()-function. Better test.sxw to explain this function.
  
• 0.11 Bugfix for Windows XP, test.sxw adapted for English language users