wiki to xml

[Pascal]

Hi all,

I was wondering how do you get mediawiki content into xml. Is it an home
made script ? Simple ? A sophisticated and a heavy process ?

I didn't found any detailed information about it on the fedoraproject
website.

I am making some latex/pdfs files from mediawiki content of
fedora-fr.org with an ugly php script and I think it's time to think to
something more efficient. So any advices are welcome :)

I already have a look to the wiki2xml extension[1], which can be a
possibility.


Regards,
Pascal

[1] http://toolserver.org/~magnus/wiki2xml/w2x.php

[Chris Curran]

Pascal wrote:
>  Hi all,
>
>  I was wondering how do you get mediawiki content into xml. Is it an home
>  made script ? Simple ? A sophisticated and a heavy process ?
>
>  I didn't found any detailed information about it on the fedoraproject
>  website.
>
>  I am making some latex/pdfs files from mediawiki content of
>  fedora-fr.org with an ugly php script and I think it's time to think to
>  something more efficient. So any advices are welcome :)
>
>  I already have a look to the wiki2xml extension[1], which can be a
>  possibility.
>
>
>  Regards,
>  Pascal
>
>  [1] http://toolserver.org/~magnus/wiki2xml/w2x.php
>
>     
Unfortunately those scripts are mostly snakeoil. I've tried many of them 
and even had my hand at writing one myself but there is no easy way to 
convert them. The main issue is the conflicting views over what 
constitutes good XML. Most of the scripts produce horrendous XML.

Also bear in mind what you are trying to do. Wiki is a simple markup 
language, docbook XML is a rich markup language. Wiki has elements 
related to formatting, docbook usually ignores formatting meta data. 
Wiki sometimes has inline PHP which just kills everything (if you have 
ever tried to write something which parses code as well as data you will 
know it's easier to write a cross compiler). To summarize it is a near 
impossible tasks to do well.

All that said, it's much easier to go the other way, from XML to wiki.

Chris

[Sibir Chakraborty]

Hi all,

There are several ways to do that. Depending on if you have physical access,
there are a few simple and complete steps.

1. Please visit MWDump tool. This is available at:
http://www.mediawiki.org/wiki/Mwdumper
2. The other way is also directly omnipresent at:
http://toolserver.org/~magnus/wiki2xml/w2x.php
3. MediaWiki also allows exporting articles to a simplified XML format. Now
if the job is to export and import docs. The link to do that is at:
http://en.wikibooks.org/wiki/MediaWiki_Administrator's_Handbook/Impor...

I could guide you thru' each of these, if you would like to do so. Let me
know.

Thanks,
Sibir

On Sun, Mar 1, 2009 at 8:19 PM, Pascal <pascal22p(a)parois.net&gt; wrote:

>  Hi all,
>
>  I was wondering how do you get mediawiki content into xml. Is it an home
>  made script ? Simple ? A sophisticated and a heavy process ?
>
>  I didn't found any detailed information about it on the fedoraproject
>  website.
>
>  I am making some latex/pdfs files from mediawiki content of
>  fedora-fr.org with an ugly php script and I think it's time to think to
>  something more efficient. So any advices are welcome :)
>
>  I already have a look to the wiki2xml extension[1], which can be a
>  possibility.
>
>
>  Regards,
>  Pascal
>
>  [1] http://toolserver.org/~magnus/wiki2xml/w2x.php
>
>  --
>  fedora-docs-list mailing list
>  fedora-docs-list(a)redhat.com
>  To unsubscribe:
>  https://www.redhat.com/mailman/listinfo/fedora-docs-list
>

[Paul W. Frields]

On Sun, Mar 01, 2009 at 02:49:01PM +0000, Pascal wrote:
>  Hi all,
>  
>  I was wondering how do you get mediawiki content into xml. Is it an home
>  made script ? Simple ? A sophisticated and a heavy process ?
>  
>  I didn't found any detailed information about it on the fedoraproject
>  website.
>  
>  I am making some latex/pdfs files from mediawiki content of
>  fedora-fr.org with an ugly php script and I think it's time to think to
>  something more efficient. So any advices are welcome :)
>  
>  I already have a look to the wiki2xml extension[1], which can be a
>  possibility. 
There is a python-mwlib to which the MediaWiki developers actually
contribute.  It's in Fedora 10, but I seem to recall there may be some
breakage going on right now because of an underlying library (odfpy?)
that was pushed prematurely.

I think Ian Weller was working on this.  Ian, what's the status?

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
  irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug

[Ian Weller]

On Mon, Mar 02, 2009 at 09:01:44AM -0500, Paul W. Frields wrote:
>  On Sun, Mar 01, 2009 at 02:49:01PM +0000, Pascal wrote:
>  > Hi all,
>  > 
>  > I was wondering how do you get mediawiki content into xml. Is it an home
>  > made script ? Simple ? A sophisticated and a heavy process ?
>  > 
>  > I didn't found any detailed information about it on the fedoraproject
>  > website.
>  > 
>  > I am making some latex/pdfs files from mediawiki content of
>  > fedora-fr.org with an ugly php script and I think it's time to think to
>  > something more efficient. So any advices are welcome :)
>  > 
>  > I already have a look to the wiki2xml extension[1], which can be a
>  > possibility.
>  
>  There is a python-mwlib to which the MediaWiki developers actually
>  contribute.  It's in Fedora 10, but I seem to recall there may be some
>  breakage going on right now because of an underlying library (odfpy?)
>  that was pushed prematurely.
>  
>  I think Ian Weller was working on this.  Ian, what's the status?
>   
I've been highly distracted and if anyone else can work on it that would
be wonderful.
  https://bugzilla.redhat.com/show_bug.cgi?id=486678
I think Jef has released a new dateutil but I can't remember.

-- 
Ian Weller <ianweller(a)gmail.com&gt;                  http://ianweller.org
GnuPG fingerprint:  E51E 0517 7A92 70A2 4226  B050 87ED 7C97 EFA8 4A36
"Technology is a word that describes something that doesn't work yet."
  ~ Douglas Adams

[Pascal]

Le Sun, 1 Mar 2009 14:49:01 +0000,
Pascal <pascal22p(a)parois.net&gt; a écrit :

>  Hi all,
>  
>  I was wondering how do you get mediawiki content into xml. Is it an
>  home made script ? Simple ? A sophisticated and a heavy process ?
>   
Thanks for all the answers, I need to have a closer to all of them now.

Pascal

[Karsten Wade]

On Tue, Mar 03, 2009 at 12:25:37AM +0000, Pascal wrote:
>  Le Sun, 1 Mar 2009 14:49:01 +0000,
>  Pascal <pascal22p(a)parois.net&gt; a écrit :
>  
>  > Hi all,
>  > 
>  > I was wondering how do you get mediawiki content into xml. Is it an
>  > home made script ? Simple ? A sophisticated and a heavy process ?
>  > 
>  
>  Thanks for all the answers, I need to have a closer to all of them now. 
I used mw-render form python-mwlibs for all the F10 work, here are the
notes I took:

http://fedoraproject.org/wiki/User:Kwade/Post-processing_wiki2xml_results

As was pointed out, the output is accurate for formatting but there is
no contextual meaning.  This is why we use a strict wiki markup
pattern that allows us to manually clean-up.

http://fedoraproject.org/wiki/WikiEditing#Marking_Technical_Terms

What I mean is, if you use '''Application Name''' consistently,
then
that shows in the XML as <emphasis role="bold">Application
Name</emphasis>.  You can then do a search for each instance of
<emphasis role="bold"></emphasis> and change it to the contextually
meaningful <application></application>.

It is a bit tedious, but in the end ... it takes the same energy to
convert from wiki to XML as XML to wiki.  One way you compress out
meaning, the other way you add in meaning.

- Karsten
-- 
Karsten 'quaid' Wade, Community Gardener
http://quaid.fedorapeople.org
AD0E0C41

[Pascal]

Le Wed, 4 Mar 2009 08:20:55 -0800,
Karsten Wade <kwade(a)redhat.com&gt; a écrit :

>  
>  I used mw-render form python-mwlibs for all the F10 work, here are the
>  notes I took:
>  
>  http://fedoraproject.org/wiki/User:Kwade/Post-processing_wiki2xml_results
>  
>  As was pointed out, the output is accurate for formatting but there is
>  no contextual meaning.  This is why we use a strict wiki markup
>  pattern that allows us to manually clean-up. 
I didn't look at it yet :)

We already have some rules for the wiki and we try to keep it simple.
No condition case, php, complicated/nested wiki markup...
The contextual meaning is a problem.

>  
>  http://fedoraproject.org/wiki/WikiEditing#Marking_Technical_Terms
>  
>  What I mean is, if you use '''Application Name''' consistently,
> then
>  that shows in the XML as <emphasis role="bold">Application
>  Name</emphasis>.  You can then do a search for each instance of
>  <emphasis role="bold"></emphasis> and change it to the
> contextually
>  meaningful <application></application>. 
We've got specific wiki markup for these that need to be kept in the
xml:

<app>Application name</app>
<path>/chemin/vers</path>
<cmd>command</cmd>
<touche>Key</touche> and so on

The other thing I want to keep is the templates. Templates shouldn't be
expanded in the xml. 

Actually, these specific markup and templates are kept into latex
commands or environnements:

\newcommand{\cmd}[1]{\textit{{\large\guilsinglleft}\,#1\,{\large\guilsinglright}}\xspace}

\newenvironment{Warning}{%
\definecolor{shadecolor}{rgb}{1,0.9,1}%
\begin{center}\begin{shaded}\begin{footnotesize}%
\raisebox{\dimexpr-\height+0.5\baselineskip}%
{\includegraphics[width=1cm]{InProgress_tip}}
\hfill\begin{minipage}[t]{\dimexpr\textwidth-1.2cm}%
}
{%
\end{minipage}\end{footnotesize}\end{shaded}\end{center}}

Pascal