How do you extract text matching a pattern in XPATH?

Question

I have data that looks like this:

<value>v13772   @FBst0451145:w&lt;up&gt;1118&lt;/up&gt;; P{GD3649}v13772@
v13773  @FBst0451146:w&lt;up&gt;1118&lt;/up&gt;; P{GD3649}v13773@</value>

How can I process this string in XPATH to extract any and all @FBst####### numbers?

I know of the xpath matches() function... but that only returns true or false. No good if I want the matching string. I've searched around but cannot find a satisfactory answer to this problem, which is probably really common.

Thanks!

Answer 1

In addition to the good answer by Michael Kay, if you want to use only the replace() function, then use:

replace(.,'.*?(@FBst\d+).*','$1')

The result is:

@FBst0451145
@FBst0451146

And if you only want the numbers from the above result, use:

replace(replace(.,'.*?(@FBst\d+).*','$1'),
          '[^0-9]+', ' ')

This produces:

 0451145 0451146

Answer 2

I Assume you can also use XQuery. The get_matches() function from the FunctX module should work for you. Download the file which supports your version of XQuery. Then import the module whenever you need its functionality.

import module namespace functx = "http://www.functx.com" at "functx-1.0-doc-2007-01.xq";

functx:get-matches(string-join(//text()),'xyz')

Answer 3

Try

tokenize(value, '[^0-9]+')

which should return the sequence of tokens separated by sequences of non-digits.