pyXML beginner questions

[Sebastian Fey]

hi,

id like to do the following:

(1) open a .xml, change something and save it.
problem is: how to save/serialize?
i tried xml.dom.ext.Print, but this resolves all entities and serializes
the xml with resolved entities. (see example below)

(2) id also like to load external parsed entities referenced in the xml.
MSXML provides an extension(?) to DOM which returns the uri to an
entityReference-NODE.
any similar in pyXML. actually, is nodetype entityReference implemented
in pyXML. i always get the nodeType of the resolved entity, ie 3
(NODE_TEXT) with a internal unparsed entity.
#############
<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE xbel [
<!ENTITY intTxt 'GIGI'>
<!ENTITY intMarkup '<entIntern>text</entIntern>'>
<!ENTITY extParsed SYSTEM "ent.xml">
]>
<root>
<text>&intTxt;</text>
&intMarkup;
&extParsed;
</root>
############

becomes:

#############
<?xml version="1.0" encoding="iso-8859-1"?>
<root>
<text>some text</text>
<entIntern>text</entIntern>'
<entIntern>text</entIntern>'
</root>
#############

thx,
Sebastian

[Uche Ogbuji]

Sebastian Fey <fe*@parsytec.de> wrote in message news:<c2*************@ID-190842.news.uni-berlin.de>...

hi,

id like to do the following:

(1) open a .xml, change something and save it.
problem is: how to save/serialize?
i tried xml.dom.ext.Print, but this resolves all entities and serializes
the xml with resolved entities. (see example below)
Sounds as if you want a lexical round-trip. Very few XML processing
packages allow for this. I'd check whether pxdom supports this. If
not, I don't expect you'll find it in Python.

http://www.doxdesk.com/software/py/pxdom.html
(2) id also like to load external parsed entities referenced in the xml.
MSXML provides an extension(?) to DOM which returns the uri to an
entityReference-NODE.
any similar in pyXML. actually, is nodetype entityReference implemented
in pyXML. i always get the nodeType of the resolved entity, ie 3
(NODE_TEXT) with a internal unparsed entity.

Again pxdom will get you closest.

--Uche

[Andrew Clover]

Sebastian Fey <fe*@parsytec.de> wrote:

actually, is nodetype entityReference implemented in pyXML.
Yes, but you won't ever see them from a parse operation.

Print() will happily serialise an entity reference as &e; providing you
can get one into the document in the first place. Using
Document.createEntityReference() is the only way I know.
(2) id also like to load external parsed entities referenced in the xml.
MSXML provides an extension(?) to DOM which returns the uri to an
entityReference-NODE. any similar in pyXML.

The standard DOM way of doing it is to use the DocumentType.entities
interface:

doctype= entref.ownerDocument.doctype
entdecl= doctype.entities.getNamedItem(entref.nodeName)
uri= entdecl.systemId # see also baseURI if using DOM Level 3 Core

This isn't any use for 4DOM as you won't get any Entity objects from its
parse stage and you can't create your own.

In DOM Level 3 Load/Save, control of whether Entity and EntityReference
objects are kept in the document is achieved with the DOMConfiguration
parameter 'entities':

parser= implementation.createLSParser(1, None)
parser.domConfig.setParameter('entities', True) # False by default
doc= parser.parseURI('file:///in.xml')
serialiser= implementation.createLSSerializer()
serialiser.domConfig.setParameter('entities', True)
serialiser.writeToURI(doc, 'file:///out.xml')

DOM 3 LS is still at Proposed Recommendation stage and isn't supported
by 4DOM yet. (Insert customary pxdom plug here.)

--
Andrew Clover
mailto:an*@doxdesk.com
http://www.doxdesk.com/