在机器人中找不到名称为“ Get Child WebElement”的关键字

Question

I want to get all child elements from Current page, for that I am using below code

${childElements} =  Get Child WebElement    xpath=//*

The above code throws an error No keyword with name 'Get Child WebElement'

Is there any other library for this to work?

Answer 1

As you have figured out that these keywords are not yet in the published SeleniumLibrary versions.

In the event that you are processing an element or simply have a string locator then you can use the proposed code in SelenniumLibrary Pull Request #702 mentioned in the answer from @yash in a custom SeleniumLibrary extension.

In the below example there are two custom Robot Framework keywords to give you the keywords using the approach mentioned in #702:

1. Get Child Webelements
2. Get Parent Webelement

In the below example it shows that you can look for the parent and child using a regular locator id=main or a WebElement ${children[0]} . This is handled by the python method find_element(s) of the WebElement python object.

Note: changed the keyword from Evaluate to Call Method for security reasons and readability.

example.robot

*** Settings ***
Library    SeleniumLibrary      

Suite Teardown    Close All Browsers

*** Test Cases ***
Get Children Then Parent
    Open Browser    http://www.google.com    headlesschrome

    ${children}  Get Child Webelements   id=main
    ${parent}    Get Parent Webelement   ${children[0]}

    ${id}        Get Element Attribute   ${parent}    id
    Should Be Equal    ${id}    main

*** Keywords ***
Get Child Webelements
    [Arguments]    ${locator}

    ${element}    Get WebElement    ${locator}    
    ${children}     Call Method       
    ...                ${element}    
    ...                find_elements   
    ...                  by=xpath    value=child::*    

    [Return]      ${children}

Get Parent Webelement
    [Arguments]    ${locator}

    ${element}    Get WebElement    ${locator}      
    ${parent}     Call Method       
    ...                ${element}    
    ...                find_element    
    ...                  by=xpath    value=parent::*

    [Return]    ${parent}  

Answer 2

There is PR #702 raised for the same. There are some ongoing discussion going on around that. As there is no any ETA/Conclusion on this yet. Meanwhile you can control all things from your identifier.

like if you want child element your identifier will be:-
In case of css:- <<parent identifier > child >>
In case of xpath:- <<parent identifier/child>>

There is limitation of using this approach is you cant have 2 different selector type as parent/child.