How to parse an xml file in a shell script

Hi Akshay,

#!/bin/bash

value=`sed "s:.*<$2>\([^<]*\)<.*:\1:" $1`
echo "$2 value is $value"

Cheers!
sunnycoder

> I want to write a shell script which would be able to get a number within a particular tag.
you'll have a lot of difficulties doing that with shell only, I guess impossible
If you use other programs like sed, grep, awk, perl (as Tintin suggested), you have a little chance to do what you want if you're very experianced in using these tools.
XML is a strucutred format, use an XML parser, anything else is a pain even for such a simple example as you described.

perl with XML:Simple would be the simplest and quickest way to do it.

akshay,
I already build a simple script for the same goal to help me investigate issues

$ cat a.xml
<?xml version="1.0" encoding="US-ASCII"?>
<trade>
<start_date>2006-09-01T00:00:00.000</start_date>
<end_date>2006-09-08T00:00:00.000</end_date>
<type>PR</type>
</trade>

$ search_xml.ksh a.xml type
searching a.xml for tagname type
PR

$ cat search_xml.ksh
#!/bin/ksh
echo searching $1 for tagname $2
myvalue=`grep "<$2>.*<.$2>" $1 | cut -f2 -d">"| cut -f1 -d"<"`
echo $myvalue


please let me know with any question about that script

> .. which means tags are not seperated line by lines.
HamdyHassan, did you read that?

Hi HamdyHassan,

Script will not work in the following case!

<Order>
       <Some>sdf</Some>
       <Tags>val</Tags>
       <ProductProfile>
            <Service>
                <Attribute Name="no_of_lines">1<\Attribute>
                <Attribute Name="no_of_lines">1<\Attribute>
           </Service>
       <ProductProfile>
       .
       .
       .
Script is not able to search --> Attribute Name="no_of_lines"
Output should be 1
   

Thank you very much for ur help!