blog/data/2012-08-10-docbook-nb.txt

   1 Title: Notes on language codes for Norwegian docbook processing...
   2 Tags: english, docbook, freeculture
   3 Date: 2012-08-10 21:00
   4
   5 <p>In <a href="http://www.docbook.org/">docbook</a> one can specify
   6 the language used at the top, and the processing pipeline will use
   7 this information to pick the correct translations for 'chapter', 'see
   8 also', 'index' etc.  And for most languages used with docbook, I guess
   9 this work just fine.  For example a German user can start the document
  10 with &lt;book lang="de"&gt;, and the document will show up with the
  11 correct content with any of the docbook processors.  This is not the
  12 case for the language
  13 <a href="http://www.hungry.com/~pere/blog/Free_Culture_in_Norwegian___5_chapters_done__74_percent_left_to_do.html">I
  14 am working with at the moment</a>, Norwegian Bokmål.</p>
  15
  16 <p>For a while, I was confused about which language code to use,
  17 because I was unable to find any language code that would work across
  18 all tools.  I am currently testing dblatex, xmlto, docbook-xsl, and
  19 dbtoepub, and they do not handle Norwegian Bokmål the same way.  Some
  20 of them do not handle it at all.</p>
  21
  22 <p>A bit of background information is probably needed to understand
  23 this mess.  Norwegian is not one, but two written variants.  The
  24 variants are Norwegian Nynorsk and Norwegian Bokmål.  There are three
  25 two letter language codes associated with these languages, Norwegian
  26 is 'no', Norwegian Nynorsk is 'nn' and Norwegian Bokmål is 'nb'.
  27 Historically the 'no' language code was used for Norwegian Bokmål, but
  28 many years ago this was found to be å bad idea, and the recommendation
  29 is to use the most specific language code instead, to avoid confusion.
  30 In the transition period it is a good idea to make sure 'no' was an
  31 alias for 'nb'.</p>
  32
  33 <p>Back to docbook processing tools in Debian.  The dblatex tool only
  34 understand 'nn'.  There are translations for 'no', but not 'nb' (BTS
  35 <a href="http://bugs.debian.org/684391">#684391</a>), but due to a bug
  36 (BTS <a href="http://bugs.debian.org/682936">#682936</a>) the 'no'
  37 language code is not recognised.  The docbook-xsl tool chain only
  38 recognise 'nn' and 'nb', but not 'no'.  The xmlto tool only recognise
  39 'nn' and 'nb', but not 'no'.  The end result that there is no language
  40 code I can use to get the docbook file working with all of these tools
  41 at the same time. :(</p>
  42
  43 <p>The correct solution is to use &lt;book lang="nb"&gt;, but it will
  44 take time before that will work with all the free software docbook
  45 processors. :(</p>
  46
  47 <p>Oh, the joy of well integrated tools. :/</p>