Xpath for finding element that does not have sibling

Question

So I'm trying to extract the a href locations from the links below but only for the a elements that do NOT have a span sibling. I've managed to figure out the Xpath for finding the ones that DO, but how do I negate this? I've tried using count() and not() but can't seem to get them to be valid or return anything.

<div class="member">
    <a href="/test1" class="ahrefclass">shouldntfindme</a>
    <span class="evilspan">evilspan</span>
</div>
<div class="member">
    <a href="/test2" class="ahrefclass">findme</a>
</div>
<div class="member">
    <a href="/test3" class="ahrefclass">shouldntfindme</a>
    <span class="evilspan">evilspan</span>
</div>
<div class="member">
    <a href="/test4" class="ahrefclass">findme</a>
</div>
<div class="member">
    <a href="/test5" class="ahrefclass">shouldntfindme</a>
    <span class="evilspan">evilspan</span>
</div>

My Xpath so far is

//a[@class="ahrefclass"][../span[@class="evilspan"]]

choroba · Accepted Answer · 2014-11-29T11:01:54.987

10

Just negate the condition:

//a[@class="ahrefclass"][not(../span[@class="evilspan"])]

which is, in this case, equivalent to

//a[@class="ahrefclass" and not(../span[@class="evilspan"])]

edited Nov 29 '14 at 11:01

answered Nov 29 '14 at 10:52

choroba

231,213
25
204
289

There is a difference between `[x][y]` and `[x and y]`. See http://stackoverflow.com/a/1006439/18771 – Tomalak Nov 29 '14 at 11:04
ah thanks! I did try using not, unsure why I had no luck, must of been using it in the wrong position or something :-) – Mr J Nov 29 '14 at 11:05
@Tomalak: In this case, they're equivalent. – choroba Nov 29 '14 at 11:07
In that case yes. It wasn't meant as a correction, but as an addition. – Tomalak Nov 29 '14 at 11:21
A little simpler (to my mind) is `//div[not(span/@class='evilspan')]/a[@class='ahrefclass']` – Michael Kay Nov 29 '14 at 16:52

Xpath for finding element that does not have sibling

1 Answers1