Looking for source preservation features in XML libs

[and-google at doxdesk.com]

Grzegorz Adam Hankiewicz <gra... at titanium.sabren.com> wrote:

> I have looked at xml.minidom, elementtree and gnosis and haven't
> found any such features. Are there libs providing these?

pxdom (http://www.doxdesk.com/software/py/pxdom.html) has some of this,
but I think it's still way off what you're envisaging.

> One is to be able to tell which source file line a tag starts
> and ends.

You can get the file and line/column where a node begins in pxdom using
the non-standard property Node.pxdomLocation, which returns a DOM Level
3 DOMLocator object, eg.:

uri= node.pxdomLocation.uri
line= node.pxdomLocation.lineNumber
col= node.pxdomLocation.columnNumber

There is no way to get the location of an Element's end-tag, however.
Except guessing by looking at the positions of adjacent nodes, which is
kind of cheating and probably not reliable.

SAX processors can in theory use Locator information too, but AFAIK (?)
this isn't currently implemented.

> Another feature is to be able to save the processed XML code in a way
> that unmodified tags preserve the original identation.

Do you mean whitespace *inside* the start-tag? I don't know of any XML
processor that will do anything but ignore whitespace here; in XML
terms it is utterly insignificant and there is no place to store the
information in the infoset or DOM properties.

pxdom will preserve the *order* of the attributes, but even that is not
required by any XML standard.

> Or in the worst case, all identation is lost, but I can control to
> some degree the outlook of the final XML output.

The DOM Level 3 LS feature format-pretty-print (and PyXML's
PrettyPrint) influence whitespace in content. However if you do want
control of whitespace inside the tags themselves I don't know of any
XML tools that will do it. You might have to write your own serializer,
or hack it into a DOM implementation of your choice.
-- 
Andrew Clover
mailto:and at doxdesk.com
http://www.doxdesk.com/