使用XPath，如何在没有前置文本的情况下获得标记的完全匹配？

Question

The doubt is with xpath in the given scenario:
I have this two dropdown item inside a customized dropdown component, this component has a textbox to search items that contains the informed text. Using XPath expression, I want to get the dropdown-item that exact match the informed text which I put at the textbox and not returns the dropdown-item that contains a text before the span tag. This span surround the item that match the informed criteria.

To be clearer, let's look at the html code itself where we have this issue:

This first dropdown-item, contains the item that I want to return. The span that contains text Steak and doesn't contains any above text, inside the div.

<div id="dropdown-item" role="option">
<span class="select2-match">Steak</span>
 - Delicious Salmon Steak
</div>

This second item, contains the item that the xpath expression should ignore, because it contains a text above the span node, inside the dropdown-item div.

<div id="dropdown-item" role="option">
Double
<span class="select2-match">Steak</span>
 - Monster Fillet Steak
</div>

I tried a lot of XPath expressions and I'm starting to think that only with XPath I won't be able to select the item I want. I know that if I concatenate two or more XPaths in programming (C# or Java, for example) and give a substring by the hyphen and in the end find the index that I need, I'll be able to find the required node, but with the amount of information I'll be working, it can not be performatic. The idea would be to use only xpath, perhaps with regex... Any suggestion?

Answer 1

This is one possible XPath (formatted for readability) :

//div[@id='dropdown-item']
/span[.='Steak']
     [not(
            preceding-sibling::text()[normalize-space()]
       )
     ]

brief explanation :

• //div[@id='dropdown-item'] : find all div element, anywhere in the HTML document, having id attribute value equals "dropdown-item"
• /span[.='Steak'] : from each of such div , find child element span , having inner text exactly equals "Steak" ....
• [not(preceding-sibling::text()[normalize-space()])] : ... and not having a non-empty preceding-sibling text node. This will match span having white-space only preceding sibling text, or not having any preceding sibling text at all.

Answer 2

我会让你调整XPath以满足你的需求，但这是一个开始：

//div[span[normalize-space(preceding-sibling::text())='']]