Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

remove element with ElementTree

23 views
Skip to first unread message

tdan

unread,
Mar 8, 2010, 10:35:50 PM3/8/10
to
I have been using ElementTree to write an app, and would like to
simply remove an element.
But in ElementTree, you must know both the parent and the child
element to do this.
There is no getparent() function, so I am stuck if I only have an
element.

I am iterating over a table and getting all <td> tags, checking their
text, and conditionally deleting them:

def RemoveElementWithText( topEl, subEl, text ):
for el in topEl.getiterator( subEl ):
if el.text = text:
break
else:
el = None
return el

RemoveElementWithText( table, 'td', 'sometext' )

My table is like so:

<table>
<thead>...
<tbody><tr><td>...

Is there any way to do this in ElementTree? I see lxml.etree does
this nicely, but I want to use python's standard library.

Justin Ezequiel

unread,
Mar 9, 2010, 4:34:52 AM3/9/10
to
On Mar 9, 11:35 am, tdan <df.tr...@gmail.com> wrote:
> I have been using ElementTree to write an app, and would like to
> simply remove an element.
> But in ElementTree, you must know both the parent and the child
> element to do this.
> There is no getparent() function, so I am stuck if I only have an
> element.
>

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

Stefan Behnel

unread,
Mar 9, 2010, 10:07:54 AM3/9/10
to pytho...@python.org
Justin Ezequiel, 09.03.2010 10:34:

Also note that there is an independent ElementTree implementation called
lxml.etree, which has parent pointers.

Stefan

Stefan Behnel

unread,
Mar 9, 2010, 10:45:48 AM3/9/10
to pytho...@python.org
tdan, 09.03.2010 04:35:

> I have been using ElementTree to write an app, and would like to
> simply remove an element.
> But in ElementTree, you must know both the parent and the child
> element to do this.
> There is no getparent() function, so I am stuck if I only have an
> element.
>
> I am iterating over a table and getting all<td> tags, checking their
> text, and conditionally deleting them:
>
> def RemoveElementWithText( topEl, subEl, text ):

Note that all-camel-case names are rather unusual for function names and
rather used for class names. See PEP 8 for a style guide.


> for el in topEl.getiterator( subEl ):
> if el.text = text:
> break
> else:
> el = None
> return el
>
> RemoveElementWithText( table, 'td', 'sometext' )
>
> My table is like so:
>
> <table>
> <thead>...
> <tbody><tr><td>...
>
> Is there any way to do this in ElementTree?

I think the easiest is to iterate not over the elements themselves, but
over their parents, and then to remove all children of the specified tag in
each step.

Stefan

0 new messages