How to get parent tag while using XPaths in xml.etree.elementree

[PBLN RAO]

I am trying to retrieve data from xml.etree.cElementTree. i have the following code Code Snippet import xml.etree.cElementTree as ET xmldata =""" <pipeline> <ep_150> <stage name="lay" longname="layout" department="layout" process="production"> <review name="R1" reviewer="sridhar reddy" role="supervisor" id="p1234"> </review> </stage> <stage name="lip" longname="lipsync" department="lipsync" process="production"> <review name="R2" reviewer="someone" role="supervisor" id="p2345"> </review> </stage> <stage name="blk" longname="blocking" department="animation" process="production"> <review name="R3" reviewer="sudeepth" role="supervisor" id="p4645" dependson='R1'> </review> <review name="R4" reviewer="chandu" role="director" id="p5678"> </review> </stage> <stage name="pri" longname="primary" department="animation" process="production"> <review name="R5" reviewer="sudeepth" role="supervisor" id="p4645" style="dep" > </review> <review name="R6" reviewer="sudeepth" role="bld_supervisor" id="p2556" style="dep"> </review> </stage> <stage name="sec" longname="secondary" department="animation" process="production"> <review name="R7" reviewer="sha" role="supervisor" id="p1234" style="dep"> </review> <review name="R8" reviewer="chandu" role="director" id="p5678"> </review> </stage> </ep_150> </pipeline> """ root = ET.fromstring(xmldata) review = root.findall("./ep_150/stage/review") print '\n\nreviews for id=p4645\n' for rev in review: if rev.attrib['id']=='p4645': print (rev.attrib['name']) with the above code i am getting the result as below reviews for id=p4645 R3 R5 But i need the output as reviewes for id=p4645 blk - R3 pri - R5 i.e, i need the parent tag along with the element using Xpaths. I know its possible if i iterate through entire tree. But need to use more loops.

[damon shelton]

perhaps loop through each stage and check its children


root = ET.fromstring(xmldata)
stages = root.findall("./ep_150/stage")

print '\n\nreviews for id=p4645\n'

for stage in stages:
	children = stage.getchildren()
	for child in children:
		if child.attrib['id']=='p4645':
			print('%s - %s' % (stage.attrib['name'], child.attrib['name']))

--
view archives: http://groups.google.com/group/python_inside_maya
change your subscription settings: http://groups.google.com/group/python_inside_maya/subscribe

[Justin Israel]

I'm not very experienced with xpath, but this link suggests a way to build a parent-child map once that you can use:

http://effbot.org/zone/element.htm#accessing-parents

root = ET.fromstring(xmldata)

parent_map = dict((c, p) for p in root.getiterator() for c in p)

...

for rev in review:

    parent = parent_map[rev]

    ...

If you are using the newest cElementTree from pypi then you can also modify your xpath to directly filter the id attribute and save the loop:

review = root.findall('./ep_150/stage/review[@id="p4645"]')

[" - ".join((parent_map[r].get('name'), r.get('name'))) for r in review]

# ['blk - R3', 'pri - R5']

And if you have lxml installed, it seems the Element objects let you directly access the parent:

review = root.xpath('./ep_150/stage/review[@id="p4645"]')

[" - ".join((r.getparent().get('name'), r.get('name'))) for r in review]

# ['blk - R3', 'pri - R5']

-- justin

[PBLN RAO]

thx justin,

parent map worked for me.

but i think your 2nd option below dont work in python 2.6.
bcoz the link which gave is cElementree 1.0.2 and in the site it says cElementree 1.0.5 comes with python 2.5 onwards. i am using python 2.6.4.
even in python 2.6 xpath did not work directly to filter the attributes.

i have posted a question on this on stackoverflow

If you are using the newest cElementTree from pypi then you can also modify your xpath to directly filter the id attribute and save the loop:

review = root.findall('./ep_150/stage/review[@id="p4645"]')

[" - ".join((parent_map[r].get('name'), r.get('name'))) for r in review]

# ['blk - R3', 'pri - R5']

[Justin Israel]

Ya, I think I wasn't clear about that part. The cElementTree that is shipped with python isn't the one that has the full xpath support I mentioned. You just need to install the newer one (or lxml if you want that).

You can use easy_install or pip to get them from pypi

easy_install cElementTree

pip install cElementTree

pip install lxml

If you don't have easy_install you can get it by downloading this:

http://peak.telecommunity.com/dist/ez_setup.py

And doing:

python ez_setup.py

Then you will have easy_install. You could further get pip and just use that if you want:

easy_install pip