Extract string from XML node using Bash script

Question

I can't seem to find exactly how to do this. I have this XML file

<session>
  <translations>
   <translation>
      <inside>198.18.133.1</inside>
      <name>adfs.domain1.com</name>
    </translation>
    <translation>
      <inside>198.18.135.60</inside>
      <name>hds.domain2.com</name>
    </translation>
   </translations>
 </session>

and I want to extract the domain from a particular name node based on the string found in the inside node. As you can see I have multiple name and inside nodes. With the following Bash file I can extract the first instance of name

#!/bin/bash
domain="$(echo "cat /session/translations/translation/name/text()" | xmllint --nocdata --shell session.xml | sed '1d;$d')"
domain="${domain:5}"
printf '%s\n' "Domain is: $domain"

This will give me domain1.com.

Sometimes I could have more translations or less and they aren't always in the same order. So I need a way to pull the name IF the inside node matches 198.18.133.1 or pull the name IF the inside node matches 198.18.135.60, etc.

Have you considered XSLT? You can do it with Java SE and Saxonica HE. — Jeff Holt, Sep 10 '18 at 20:31
@jeff6times7, there's no need for a JVM -- every modern Linux distro (and MacOS) ships with `xsltproc`. Maybe if you needed something newer than 1.0, but this query doesn't need a modern version of the standard. — Charles Duffy, Sep 10 '18 at 20:31
Even if it's not Linux, it's something that ships with bash, and that narrows it down enough. xsltproc is pretty much ubiquitous these days -- it's widely used for compiling documentation to HTML at build time. If you have a recent (last-15-years) UNIX-family system with a compiler/development toolchain, it's more more common to not have a JVM than to not have xsltproc. — Charles Duffy, Sep 10 '18 at 20:33
@CharlesDuffy Then it's hopefully an easy decision based upon the long-term requirements. If the user's *not* headed towards a complicated transformation, then XSLT is overkill. — Jeff Holt, Sep 10 '18 at 20:38
Incidentally, XMLStarlet compiles command-line queries down to XSLT -- one can (in the relevant usage modes) ask it to output an XSLT template which will perform the same operation given on the command line. — Charles Duffy, Sep 10 '18 at 20:39
BTW, I'd suggest `domain=${domain#*.}` to strip off everything up to and including the first period -- that way it'll work the same way if that prefix's length isn't exactly four characters. — Charles Duffy, Sep 10 '18 at 20:51

Charles Duffy · Accepted Answer · 2018-09-10T20:50:15.090

Your current XPath expression, of:

/session/translations/translation/name/text()

...can easily be changed to:

/session/translations/translation[inside="198.18.133.1"]/name/text()

...to perform the desired filtering.

Doing this with XMLStarlet, rather than XMLLint, might look more like:

xmlstarlet sel -t -m '/session/translations/translation[inside="198.18.133.1"]/name' -v . -n

If adding the -C argument to the sel subcommand, it will emit the XSLT template which it's evaluating under-the-hood, which you could run anywhere with xsltproc installed, including systems without XMLStarlet:

<?xml version="1.0"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" xmlns:exslt="http://exslt.org/common" version="1.0" extension-element-prefixes="exslt">
  <xsl:output omit-xml-declaration="yes" indent="no"/>
  <xsl:template match="/">
    <xsl:for-each select="/session/translation/translation[inside=&quot;198.18.133.1&quot;]/name">
      <xsl:call-template name="value-of-template">
        <xsl:with-param name="select" select="."/>
      </xsl:call-template>
      <xsl:value-of select="'&#10;'"/>
    </xsl:for-each>
  </xsl:template>
  <xsl:template name="value-of-template">
    <xsl:param name="select"/>
    <xsl:value-of select="$select"/>
    <xsl:for-each select="exslt:node-set($select)[position()&gt;1]">
      <xsl:value-of select="'&#10;'"/>
      <xsl:value-of select="."/>
    </xsl:for-each>
  </xsl:template>
</xsl:stylesheet>

Thanks, I tried to change the XPath expression and it gave me this error. "XPath error : Invalid predicate /session/translations/translation[inside=198.18.133.1]/name/text() ^ xmlXPathEval: evaluation failed /session/translations/translation[inside=198.18.133.1]/name/text(): no such node — Jason Murray, Sep 10 '18 at 20:44
See how it's just `inside=198.18.133.1`, not `inside="198.18.133.1"`, in the error? You don't want to let the shell eat your quotes. — Charles Duffy, Sep 10 '18 at 20:45
...there's a reason I used single quotes on the outside in this answer -- so the double-quotes inside them would be preserved. — Charles Duffy, Sep 10 '18 at 20:47
Sweet yes that was it! Thanks so much for the quick response! Works like a champ! — Jason Murray, Sep 10 '18 at 20:52

Extract string from XML node using Bash script

1 Answers1