get specific value from xml file using bash script

Question

I am not familiar with bash and strings, and need to get a value from xml file, after curling the xml I get:

<metadata modelVersion="1.1.0">
  <groupId>com.company.NYC</groupId>
  <artifactId>scripts</artifactId>
  <version>2.3.7-SNAPSHOT</version>
  <versioning>
    <snapshot>
      <timestamp>20201209.102124</timestamp>
      <buildNumber>38</buildNumber>
    </snapshot>
    <lastUpdated>20201209104133</lastUpdated>
    <snapshotVersions>
      <snapshotVersion>
        <classifier>NYC</classifier>
        <extension>zip</extension>
        <value>2.3.7-20201209.102124-38</value>
        <updated>20201209102124</updated>
      </snapshotVersion>
      <snapshotVersion>
        <classifier>tests</classifier>
        <extension>jar</extension>
        <value>2.3.7-20201209.102124-38</value>
        <updated>20201209102124</updated>
      </snapshotVersion>
      <snapshotVersion>
        <extension>pom</extension>
        <value>2.3.7-20201209.102124-38</value>
        <updated>20201209102124</updated>
      </snapshotVersion>
    </snapshotVersions>
  </versioning>
</metadata>

using this command, I get the version tags value:

#wget -q -N $urlversion=$(curl -s https://www.artifactoty.company.net/artifactory/maven-dev-local/com/google/api/framework/maven-metadata.xml --insecure | grep  -oP '(?<=<value>).*?(?=</value>))

And all what i need is value after the zip file:

extension>zip</extension>
<value>2.3.7-20201209.102124-38</value>

The 2.3.7-20201209.102124-38 value.

How can i get it alone.

Thanks friends.

Does this answer your question? [How to parse XML in Bash?](https://stackoverflow.com/questions/893585/how-to-parse-xml-in-bash) — Andrew, Dec 09 '20 at 14:21

score 0 · Answer 1 · answered Dec 09 '20 at 14:38

Using xmlstarlet:

curl ... | xmlstarlet sel -t -v "metadata/versioning/snapshotVersions/snapshotVersion/value" | head -1

Take the output of the curl command and process through xmlstarlet. Use sel to select the path to all values and then print only the first value (corresponding to zip)

markp-fuso · Accepted Answer · 2020-12-09T14:58:34.817

Assuming OP has a reason for not using a XML-aware tool to parse the data ...

NOTE: I've copied the sample xml data into file tag.xml for the purposes of this answer.

One awk solution:

awk -F'[<>]' '             # define multiple field delimiters as "<" and ">"
$3=="zip" {getline         # if field #3 = "zip" then read the next line and ...
           print $3        # print field #3
           exit}           # at this point we have what we want so exit
' tag.xml

Collapsing into a one-liner and storing in a variable:

zipver=$(awk -F'[<>]' '$3=="zip" {getline; print $3; exit}' tag.xml)
echo "${zipver}"

Both of the above generate:

2.3.7-20201209.102124-38

OP could modify the current command as follows:

wget -q -N $urlversion=$(curl -s ... | awk -F'[<>]' '$3=="zip" {getline; print $3; exit}')

score 0 · Answer 3 · answered Dec 12 '20 at 11:59

Please, don't use regex to parse XML!
When the url would still work (it doesn't for me at the moment) you could use xidel with a XPath expression to get what you want:

xidel -s "<url>" -e '//snapshotVersion[extension="zip"]/value'
2.3.7-20201209.102124-38

score 0 · Answer 4 · answered Jul 15 '22 at 15:08

So, however you got to this:

extension>zip</extension>
<value>2.3.7-20201209.102124-38</value>

Given :
$ echo "$var1" extension>zip 2.3.7-20201209.102124-38

This gets you what you want:

$ echo "$var1" |tail -1|cut -d">" -f2 |cut -d"<" -f1
2.3.7-20201209.102124-38

get specific value from xml file using bash script

4 Answers4