XPath parent node's class should not contain specific string

Question

I'm trying to find all div s whose class name is 'phrase' and parent node's class name is not 'extras'.

So in Python I'm using

for phrase in entry.iterfind(".//div[@class='phrase'] and ./parent::div[@class!='extras']]"):

to do that.

But it gives me the error:

SyntaxError: prefix 'parent' not found in prefix map

And I changed the above code to

for phrase in entry.iterfind(".//div[@class='phrase'] and ./..[@class!='extras']]"):

This time the error was

Traceback (most recent call last):File "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/xml/etree/ElementPath.py", line 272, in iterfind
selector = _cache[cache_key] KeyError: (".//div[@class='phrase'] and ./..[@class!='extras']]", None)

Part of the XML structures are as follows:

<div class="phrases">
    <div class="label">Phrases</div>
    <div class="phrase">
    ……

<div class="phrasal verbs">
    <div class="label">Phrases</div>
    <div class="phrase">
    ……

<div class="extras">
    <h2>test test</h2>
    <div class="phrase">
    ……

I'm using Python 3.7 and xml.etree library on Mac OS 10.14.

Answer 1

Problem might be in your current tool as it might not support some XPath syntax.

You can try lxml.html to parse the same HTML-doc:

from lxml import html

source = """<div class="phrases">
                <div class="label">Phrases</div>
                <div class="phrase">this</div>
            </div>

            <div class="phrasal verbs">
                <div class="label">Phrases</div>
                <div class="phrase">this</div>
            </div>

            <div class="extras">
                <h2>test test</h2>
                <div class="phrase">not this</div>
            </div>"""

dom = html.fromstring(source)
dom.xpath(".//div[@class='phrase' and ./parent::div[@class!='extras']]")

Output:

[<Element div at 0x7fb5218d5db8>, <Element div at 0x7fb521018728>] #  Two elements found

or

dom.xpath(".//div[@class='phrase' and ./parent::div[@class!='extras']]/text()")

Output:

['this', 'this']

Answer 2

你可以使用类似"//div[@class!='extras']/div[@class='phrase']"东西，它应该找到所有类'phrase'的div，其中父类不是'extras'