[Rfc-markdown] New xml2rfc release: v2.9.0

[Henrik Levkowetz <[e-mail]>]

Hi,

Re-sending this, to fix broken formatting.

Release notes:

xml2rfc (2.9.0) ietf; urgency=medium

  This release introduces preptool functionality, through a --preptool output
  mode.  With reservation for some points for which issues has been raised,
  this follows the specification in RFC7998.

  The preptool currently takes vocabulary v3 input, and produces prepped
  output.  When work on the text formatter commences, the idea is that the
  input xml source will always be run through v2v3 conversion and preptool
  processing before the output formatting, in order to increase consistency
  and reduce complexity of the output formatter.

  There are also some changes which are not related to the preptool
  functionality: The tox tests have been changed to add testing under Python
  3.6, and removed test runs for Python 3.3.  Although there is no intention
  of breaking compatibility with Python 3.3, it may happen eventually since
  there will not be any release testing with that version of Python.

  The v2v3 converter in some cases could insert <seriesInfo> elements with
  only a name= attribute, because the required seriesNo= attribute on <rfc>
  was missing.  This has been changed.

  In order to work around a debilitating issue with relax-ng validation in
  libxml2 (time to validate increases exponentially with number of attributes
  on the root element: https://bugzilla.gnome.org/show_bug.cgi?id=133736)
  some empty attributes on <rfc> are removed during processing; for instance
  obsoletes="" and updates="".  They don't contribute information, but
  increase validation time with a factor ~20.

  In order to identify the unicode scripts needed to display a document,
  a module to efficiently identify the scripts related to unicode codepoints
  has been written.  The 'uniscripts' module which was originally intended to
  be used for this turned out not to be viable.  The new 'scripts' module can
  be broken out for separate release as a library module, if desired.

  In order to work with vocabulary v3 input, the parser has been slightly
  modified to not do input validation according to rfc2629.dtd if not
  appropriate.


The preferred way to install xml2rfc is by doing 'pip install xml2rfc',
and 'pip install --upgrade xml2rfc' to upgrade.

The new version is also available through SVN checkout, with
  'svn checkout http://svn.tools.ietf.org/svn/tools/xml2rfc/tags/cli/2.9.0'

Regards,

	Henrik
	(via the mkrelease script)