mjhjmtu

I have the following XML file:

<?xml version="1.0" encoding="utf-8"?>
<deviceID>IO238470374109730497</devicdeID>
<KeyValuePairs>
 <KeyValuePair>
 <BASE-URL>2</BASE-URL>
 <Key>2</Key>
 <Value>2</Value>
 </KeyValuePair>
 <KeyValuePair>
 <BASE-URL>3</BASE-URL>
 <Key>3</Key>
 <Value>3</Value>
 </KeyValuePair>
</KeyValuePairs>

I want to delete only this section:

<KeyValuePair>
 <BASE-URL>3</BASE-URL>
 <Key>3</Key>
 <Value>3</Value>
</KeyValuePair>

What I want to achieve is to delete what is inside and including <KeyValuePair> only for these specific <BASE-URL>, <Key> and <Value>. The <KeyValuePair> tuple is always unique and won't be repeated, but it can be one or more.

How can I do that using bash?

Your example isn't a valid XML file. I've altered it slightly by adding a <root/> element surrounding the document (and fixing your typo in the closing </deviceID> marker):

<?xml version="1.0" encoding="utf-8"?>
<root>
 <deviceID>IO238470374109730497</deviceID>
 <KeyValuePairs>
 <KeyValuePair>
 <BASE-URL>2</BASE-URL>
 <Key>2</Key>
 <Value>2</Value>
 </KeyValuePair>
 <KeyValuePair>
 <BASE-URL>3</BASE-URL>
 <Key>3</Key>
 <Value>3</Value>
 </KeyValuePair>
 </KeyValuePairs>
</root>

Now it's possible to use an XML editing tool to edit your XML. This XPath expression tells xmlstarlet to delete the XML element <KeyValuePair/> (and its content) where <BASE-URL/> has a text value 3.

xmlstarlet ed -d '//KeyValuePair[BASE-URL/text()="3"]' /tmp/500185.xml

And the output is

<?xml version="1.0" encoding="utf-8"?>
<root>
 <deviceID>IO238470374109730497</deviceID>
 <KeyValuePairs>
 <KeyValuePair>
 <BASE-URL>2</BASE-URL>
 <Key>2</Key>
 <Value>2</Value>
 </KeyValuePair>
 </KeyValuePairs>
</root>