Need mechanism for specifying article content filters for Mediawiki dumps

[aarddict]

Need mechanism for specifying article content filters for Mediawiki dumps

[itkach]

Filters can be XPath expressions evaluated against XML element tree or regular expressions applied to resulting article xhtml text, optionally combined with condition (e.g. remove element that matches given XPath expression but only if all it's children's text is blank). Problem with using XPath to identify content to exclude is that ElementTree has inadequate XPath support, so mwaardhtmlwriter probably needs to be rewritten with lxml, or content needs to be serialized and reparsed with lxml.

[itkach]

Implemented in f4dc271 and further improvements

See http://aarddict.org/aardtools/doc/aardtools.html#content-filters