Will web page elements support XPath in the future

[guoliweb]

Will web page elements support XPath in the future

[Descolada]

I do not have plans to add XPath support, but I am not opposed to PRs to implement such support. There are a few reasons why I haven't done it:

1. I don't have the time
2. It would mostly duplicate pre-existing features. For example suppose XPath used element types as nodes, in which case /Pane[2]/Button[1] would select the second pane and then the first button, which is currently implemented as rootEl[{Type: "Pane", i: 2}, {Type: "Button"}]. The XPath one is more terse, but the second one can easily be more flexible if needed (eg also using callbacks, extra modifiers such as requiring a specific element name etc).
3. It'd require writing a separate parser for XPath. For example string parsing can get tricky because escape characters need to be introduced, and it gets even more complex if expressions were to be supported.