Question

Okay, so I’ve got some example xml that looks like this:

<Node name="details">
   <Node name="1130482269">
      <Val name="type"><u>2</u></Val>
      <Val name="owner"><n/></Val>
      <Val name="account_number"><u>99-71-39</u></Val>
   </Node>
   <Node name="570722270">
      <Val name="type"><u>2</u></Val>
      <Val name="owner"><n/></Val>
      <Val name="account_number"><u>99 71 40</u></Val>
   </Node>
   <Node name="1845243341">
      <Val name="type"><u>2</u></Val>
      <Val name="owner"><n/></Val>
      <Val name="account_number"><u>9971-41</u></Val>
   </Node>
</Node>

I’m trying to do a search based on account_number. But there are so many different formats that it could be in. Some have spaces, some have dashes, some have a mixture etc. If I was able to do a search removing spaces and dashes, I should be able to get any account number.

With accountnumber being the search query, the ultimate goal is to get the second level node name number by doing a search in XPATH doing this:

doc = libxml2.parseDoc(xml)
ctxt = doc.xpathNewContext()
res = ctxt.xpathEval("/Node/Node[Val[@name='account_number']/*='" + str(accountnumber) + "']/@name")

Is there a way to use XPATH search, to remove all dashes and spaces, and leave only numbers?

Thanks

Answer 1

Yes, it is possible using translate() function. Example:

"translate(Node/Node[@name='1130482269']/Val[@name='account_number'],'- ','')"

will output 997139.

You can use then (code line broken for readability):

"translate(
  /Node/Node[@name='" + 
  str(accountnumber) + 
  "']/Val[@name='account_number'],'- ','')"

Note the white space included in the second argument of translate '- ' (or even '- ') necessary to remove white spaces.