modified the library, tests and applications to support both Python3 …

…and Python2

README

@@ -0,0 +1,81 @@
+This is a collection of utility programs written in Python to manipulate
+OpenDocument files.
+
+How to proceed: Each application has its own directory. In there, look
+at the manual pages. The Python-based tools need the odf library. Just
+make a symbolic link like this: ln -s ../odf odf
+... or type: make
+
+For your own use of the odf library, see api-for-odfpy.odt
+
+This version has been reworked by Georges Khaznadar <georgesk@debian.org>,
+to add Python3 support.
+                             -o- INSTALLATION -o-
+
+First you get the package.
+
+$ git clone https://github.com/georgesk/odfpy.git
+
+Then you can build and install the library for Python2 and Python3:
+
+$ python setup.py build
+$ python3 setup.py build
+$ su
+# python setup.py install
+# python3 setup.py install
+
+The library is incompatible with PyXML.
+
+                            -o- TODO / IDEAS -o-
+
+* tests:
+  Some part of the files tests/test*.py are still unsuccessful. This
+  announces that there is still something to fix (2014-10-21).
+  .
+  The file tests/testxhtml.py is particularly strange: when the test
+  is run twice, it does not yeld the same result, part of the procedures
+  have unpredictable results. The same ODF source document can be
+  translated to different XHTML targets: this is visible when one watches
+  CSS output like margin-left and margin-right attributes, which have
+  different values for successive calls to the same test.
+  .
+  This unpredictable behavior occurs both with Python2 and Python3.
+
+* html2odf
+  Alex Hudson has been contracted to produce a command-line html2odf 
+  converter. It should include support for images, tables, CSS, etc.
+  He will provide a C# version first, and later a C version.
+
+* odf2pdf
+  A valuable tool, but one that is hard to do. PDF is an immensely
+  popular format, but it's tricky to make PDFs. With an odf2pdf tool
+  available, many developers would use ODF purely for the purpose of
+  generating a PDF later. The latest idea is to hire KOffice 
+  developers and get them to trim down KOffice into a converter.
+
+* pdf2odf
+  This conversion is less likely to produce good results, but it 
+  might be worth a shot. Poppler is a pdf library that can convert 
+  PDF into XML. Maybe we can convert that XML to ODF.
+  http://webcvs.freedesktop.org/poppler/poppler/
+
+* odfclean
+  A command-line program that removes unused automatic styles, 
+  metadata and track-changes. Some companies might like to send all
+  out-going files through odfclean to remove any information they
+  don't want others to see.
+
+* odf2xliff
+  Create XLIFF extraction and merge application. XLIFF is a OASIS file
+  for translations. You extract the text strings, send them to the translator
+  and then import them. It allows you to work on the document in the
+  meantime and only retranslate the changed parts.
+
+* odfdiff
+  A program that can generate a diff between two ODF files. Useful for 
+  SVN commit messages. This is very difficult to do. But see:
+  http://www.manageability.org/blog/stuff/open-source-xml-diff-in-java/view
+  http://freshmeat.net/projects/xmldiff/
+
+* odfsign
+  Sign and verify the signature(s) of an ODF document.

README.md

@@ -1,64 +1,43 @@
-# ODFPY
+odfpy
+=====
 
-This is a collection of utility programs written in Python to manipulate
-OpenDocument files.
+a new branch of odfpy to support Python3
 
-How to proceed: Each application has its own directory. In there, look
-at the manual pages. The Python-based tools need the odf library. Just
-make a symbolic link like this: ln -s ../odf odf
-... or type: make
+This version has been reworked by Georges Khaznadar <georgesk@debian.org>,
+to add Python3 support.
 
-For your own use of the odf library, see api-for-odfpy.odt
-
-## INSTALLATION
+INSTALLATION
+============
 
 First you get the package.
 
-    $ git clone https://github.com/eea/odfpy.git
-
-    $ python setup.py build
-    $ su
-    # python setup.py install
-
-The library is incompatible with PyXML.
-
-## TODO / IDEAS
+$ git clone https://github.com/georgesk/odfpy.git
 
-* html2odf
-  Alex Hudson has been contracted to produce a command-line html2odf 
-  converter. It should include support for images, tables, CSS, etc.
-  He will provide a C# version first, and later a C version.
+Then you can build and install the library for Python2 and Python3:
 
-* odf2pdf
-  A valuable tool, but one that is hard to do. PDF is an immensely
-  popular format, but it's tricky to make PDFs. With an odf2pdf tool
-  available, many developers would use ODF purely for the purpose of
-  generating a PDF later. The latest idea is to hire KOffice 
-  developers and get them to trim down KOffice into a converter.
+```
+$ python setup.py build
+$ python3 setup.py build
+$ su
+# python setup.py install
+# python3 setup.py install
+```
 
-* pdf2odf
-  This conversion is less likely to produce good results, but it 
-  might be worth a shot. Poppler is a pdf library that can convert 
-  PDF into XML. Maybe we can convert that XML to ODF.
-  http://webcvs.freedesktop.org/poppler/poppler/
-
-* odfclean
-  A command-line program that removes unused automatic styles, 
-  metadata and track-changes. Some companies might like to send all
-  out-going files through odfclean to remove any information they
-  don't want others to see.
-
-* odf2xliff
-  Create XLIFF extraction and merge application. XLIFF is a OASIS file
-  for translations. You extract the text strings, send them to the translator
-  and then import them. It allows you to work on the document in the
-  meantime and only retranslate the changed parts.
-
-* odfdiff
-  A program that can generate a diff between two ODF files. Useful for 
-  SVN commit messages. This is very difficult to do. But see:
-  http://www.manageability.org/blog/stuff/open-source-xml-diff-in-java/view
-  http://freshmeat.net/projects/xmldiff/
+The library is incompatible with PyXML.
 
-* odfsign
-  Sign and verify the signature(s) of an ODF document.
+                            -o- TODO / IDEAS -o-
+
+* tests:
+  Some part of the files tests/test*.py are still unsuccessful. This
+  announces that there is still something to fix (2014-10-21).
+  .
+  The file tests/testxhtml.py is particularly strange: when the test
+  is run twice, it does not yeld the same result, part of the procedures
+  have unpredictable results. The same ODF source document can be
+  translated to different XHTML targets: this is visible when one watches
+  CSS output like margin-left and margin-right attributes, which have
+  different values for successive calls to the same test.
+  .
+  This unpredictable behavior occurs both with Python2 and Python3.
+
+... see the file README (without extension) for more information.