Xpath: matching one element based on attribute of neighboring element

Question

I've got markup of the following format I'm attempting to work with using Selenium/Python:

    <tr>
        <td><a href="www.google.com">google</a></td>
        <td>useless text</td>
        <td>useless text2</td>
        <td>useless text3</td>
        <td><a href="needle@email.com">emailaddress</a></td>
    </tr>

The idea being that given a known email address (part of the href in the emailaddress td), I can get to (and click) the a in the first td. It looks like xpath is the best choice to accomplish this with Selenium. I'm trying the following xpath:

//*[@id="page_content"]/table/tbody/tr[2]/td[2]/div/table[1]/tbody/tr/td[4]/a[contains(@href, "mailto:needle@email.com")]/../../td/a[0]

But I'm getting this error:

selenium.common.exceptions.NoSuchElementException: Message: no such element: Unable to locate element: {"method":"xpath","selector":"xpathhere"}

I do know the xpath to get to the "needle@email.com" a is correct, as it's just copied from chrome dev tools, so the error must be with the part of the xpath after reaching the first a element. Can anyone shed some light on the problem with my xpath?

The problem is I'm attempting to retrieve the `a` in the first `td`, so I need to figure out how to get back to that level once the "needle@email.com" `a` is matched. — Marcatectura, May 08 '17 at 17:12
You should avoid using indexes, always use something you know will not change, like part of an attribute for example, using indexes is a bad practice and not very flexible. — lauda, May 08 '17 at 17:28
The best answer is @Andersson's. No index, robust against unexpected issues and Selenium state of the art practices. — keepAlive, May 08 '17 at 18:01

score 2 · Answer 1 · answered May 08 '17 at 17:52

Try to use below code:

from selenium.webdriver.common.by import By
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.support.ui import WebDriverWait as wait

xpath = '//td[a[@href="needle@email.com"]]/preceding-sibling::td/a'
wait(driver, 10).until(EC.element_to_be_clickable((By.XPATH, xpath))).click()

This should allow you to match first link based on href attribute of the last link in table row (tr) and click it once it became clickable

No index, noisily robust against unexpected issues and Selenium state of the art practices. — keepAlive, May 08 '17 at 18:10

score 1 · Answer 2 · edited May 23 '17 at 12:26

First, note that (this may be a meaningless typo) you are looking for "mailto:needle@email.com" while the value of your href attribute is "needle@email.com".

Second, you actually know how to get back [...]. But Xpath indexing starts with 1. Thus why this 'a[0]', is this also a meaningless typo ?

Anyway, this xpath would get your sibling

'//a[contains(@href, "needle@email.com")]/../../td[1]/a[1]'

Or more accurately than using contains (since you may have other email adresses that can be matched, e.g. "otherneedle@email.com")

'//a[@href="needle@email.com"]/../../td[1]/a[1]'

Or even better, i.e. with no index and no parent/child like exploration.

'//td[a[@href="needle@email.com"]]/preceding-sibling::td/a'

All tested.

score 0 · Answer 3 · answered May 08 '17 at 17:25

Try to find the tr that contains that email, and click on the first link from it.

//tr[.//a[contains(@href, 'your_email')]]//a

or

//tr[.//a[contains(@href, 'your_email')]]//a[@href]

or

//tr[.//a[contains(@href, 'your_email')]]//a[contains(@href, 'common_url_part')]

score 0 · Answer 4 · answered May 08 '17 at 17:48

Your HTML should be this.

<tr>
    <td><a href="www.google.com">google</a></td>
    <td>useless text</td>
    <td>useless text2</td>
    <td>useless text3</td>
    <td><a href="mailto:needle@email.com">emailaddress</a></td>
</tr>

Otherwise, your user can click on the link until he or she works himself into a frenzy. :)

Then you can do this in selenium.

>>> from selenium import webdriver
>>> driver = webdriver.Chrome()
>>> driver.get("file://c:/scratch/temp2.htm")
>>> link = driver.find_element_by_xpath('.//a[contains(@href,"needle@email.com")]')
>>> link.click()

I've used contains because an email address in a link can be something like mailto:Jose Greco <needle.email.com>.

PS: And incidentally I've just executed this stuff on my machine.

Xpath: matching one element based on attribute of neighboring element

4 Answers4