How to remove specific attributes from an ElementTree

Question

I'm trying to remove all lines from an xml file that have one of these two forms:

<attr key="filename"><string>[SOME_FILENAME]</string></attr>
<attr key="line_number"><integer>[SOME_NUMBER]</integer></attr>

Right now my code looks like this:

for parent in tree.iter():
    for child in parent:
           if 'key' in child.attrib:
                   if child.attrib['key'] == 'phc.filename':
                           del child.attrib['key']
                   elif child.attrib['key'] == 'phc.line_number':
                           del child.attrib['key']

But the output isn't what I want, it's changing this:

<attr key="filename"><string>[SOME_FILENAME]</string></attr>
<attr key="line_number"><integer>[SOME_NUMBER]</integer></attr>

into this

<attr><string>[SOME_FILENAME]</string></attr>
<attr><integer>[SOME_NUMBER]</integer></attr>

When I'd rather have both of those lines gone altogether.

I've also tried replacing the "del child.attrib['key']" lines with parent.remove(child) but that doesn't work the way I tried it either.

Answer 1

That is because you are only removing attributes, not the elements themselves

Try:

    dict = {}
    for parent in tree.iter():
        for child in parent:
               if 'key' in child.attrib:
                       if child.attrib['key'] == 'phc.filename':
                               dict[child] = parent
                       elif child.attrib['key'] == 'phc.line_number':
                               dict[child] = parent

    for child in dict:
        parent = dict[child]
        parent.remove(child)