Bash script: Get XML elements into array using xmllint

Question

This is a follow up question to this post.

I want to end up with an array, containing all the <description> elements of the xml.

array[0] = "<![CDATA[A title for .... <br />]]>"
array[1] = "<![CDATA[A title for .... <br />]]>"

...

file.xml:

<item>
    <description><![CDATA[A title for the URLs<br /><br />

    http://www.foobar.com/foo/bar
    <br />http://bar.com/foo
    <br />http://myurl.com/foo
    <br />http://desiredURL.com/files/ddd
    <br />http://asdasd.com/onefile/g.html
    <br />http://second.com/link
    <br />]]></description> 


</item>
    </item>
<description> ...</description>
    <item>

Solution

A Bash solution could be

let itemsCount=$(xmllint --xpath 'count(//item/description)' /tmp/so.xml)
declare -a description=( )

for (( i=1; i <= $itemsCount; i++ )); do 
    description[$i]="$(xmllint --xpath '//item['$i']/description' /tmp/so.xml)"
done

echo ${description[@]}

Disclaimer

Consider that bash may not be the right tool. XSLT/XPath could give you direct access to the content of the description element as describe in previous answer. For instance:

xmllint --xpath '//item/description/text()' /tmp/so.xml

Return every <description> content

Comments

[Édouard Lopez]

why do you want to store them in an array ?

[tzippy]

I want to extract values from each array object. What better solution thatn an array would you suggest? What would be an Xpath expression for that?

[tzippy]

Thank you! The thing is that, what you described in your disclaimer, returns all the content of the description elements in one string. That's not too useful for further processing.

[Édouard Lopez]

code before disclaimer should do what you ask. Disclaimer is here to only if you don't care/need the element description that serve as source

[JulianM]

I think in the bash solution it should be "smallerThanOrEqualTo" <= for (( i=1; i <= $itemsCount; i++ )); do