Hello, I'm doing a spider to capture with XPath some web data.
But the creation of xpath is a bit of work. Does anyone know of any way to train XPath? Example; I click 5 times on a link and some tool generates the xpath.
Any tips are welcome.
Thank you!
I compiled the answers found here:
When using Chrome Dev Tools (you can access by right-clicking on an element and asking to "inspect element") you can right-click the element, then "Copy XPath".
You can also use $ x in the Chrome javascript console.
ex: $ x ("// img")
I did not get to test, but this tool apparently seems to be what you want.
Try the bookmarklet SelectorGadget .
It works like this: right after you hit, you click on something you want to capture. It generates a very generic selector, and highlights the captured sections in yellow.
From here on, you can refine the selection by: 1) clicking on a highlighted area to remove from the selection; or 2) by clicking on an undefined area to add the selection.
By default it generates a CSS selector, but has an XPath button to get the equivalent XPath expression.
It has a add-on for Firefox that queries XPath on a page, called XPath Checker .
Install xmllint (tiny tool, usually already installed on linux) and play xpath:
xmllint --xpath '//symbol' file.xml
or, in the case of html:
xmllint --html --xpath '//table' file.html
or even
xmllint --html --xpath '//a' 'http://a.b.c/index.html' > links.txt