XPath to pick an element based on the text in parent or any of its child

Is there any better XPath to pick the div based on the text inside the parent or any of the child?

Use ., not text().

//*[contains(., 'Tag 1')]

text() does not give you the element's "text".

It gives you a list (!) of text nodes that are direct children of the current context node. When the context node is <div> in example #2, that list would be three text nodes containing only whitespace. I've highlighted them with brackets:

<div title='Title2'>[
  ]<input type='checkbox' />[
  ]<span>Tag 1<span>[
]</div>

'Tag 1' is a child of <span>, not of <div>.

Now, contains() does not accept node lists. If you give it a node list, it will only consider the string value of the very first node in that list. The string value of a node is the concatenation of all text nodes it contains, not just direct children.

. refers to the context node. In example #2 that's the <div> itself. contains() again converts it to string, but this time, that string actually contains Tag 1. Another way to write this is:

//*[contains(string(.), 'Tag 1')]

That's what you thought text() would do.

Now //* is recursive, this means the <div> will be selected, the <span> and all the <div> ancestors, too, because they all contain Tag 1 at some point.

Use something more specific than //* to fix this.