Review Request: Enable a much superior html export filter

[Inge Wallin]

> On Oct. 14, 2012, 7:41 a.m., Boudewijn Rempt wrote:
> > I don't think the character set selection is important at all -- utf8 should be fine. Does the new filter handle tables correctly? That's the most important thing for me, I think.

Yes, and it even handles tables in tables correctly.  It uses a recursive descent parsing strategy for the ODF contents.


> On Oct. 14, 2012, 7:41 a.m., Boudewijn Rempt wrote:
> > filters/words/CMakeLists.txt, line 60
> > <http://git.reviewboard.kde.org/r/106840/diff/1/?file=89465#file89465line60>
> >
> >     There's no need to keep obsolete code around, so if we move to this filter, the old one should be removed.

Agreed, but I wanted to make the patch for review as simple as possible.


- Inge


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/106840/#review20302
-----------------------------------------------------------


On Oct. 13, 2012, 11:45 p.m., Inge Wallin wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/106840/
> -----------------------------------------------------------
> 
> (Updated Oct. 13, 2012, 11:45 p.m.)
> 
> 
> Review request for Calligra.
> 
> 
> Description
> -------
> 
> This patch enables a much more powerful html filter than the current xslt-based one in filters/words/html-odf. It is developed using the same conversion technology that is used in the epub filter. You can read about the full feature set here: http://ingwa2.blogspot.se/2012/09/more-progress-in-calligra-author.html. The features are listed under "Export to epub2", but the html filter uses the same conversion code.
> 
> The most important difference to the current filter is that all the character and paragraph styles are translated into CSS code. Vector images are also supported with automatic conversion from SVM/EMF/WMF to SVG where applicable.
> 
> The one drawback that I can see is that there is an export option in the html-odf filter where the user can choose encoding. The new filter uses UTF-8 for everything. If that's a really important issue the the new filter can be converted to provide that too.
> 
> 
> Diffs
> -----
> 
>   filters/words/CMakeLists.txt 7dfed28 
>   filters/words/epub/CMakeLists.txt 7744e7e 
> 
> Diff: http://git.reviewboard.kde.org/r/106840/diff/
> 
> 
> Testing
> -------
> 
> Tested with a lot of different odt files. While testing the epub filter, the resulting files were run through the epub2 validator which also tests all aspects of the html.
> 
> 
> Thanks,
> 
> Inge Wallin
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/calligra-devel/attachments/20121014/c2f2b729/attachment.htm>