Shell Programming and Scripting

How can I parse file containing xml ?
I am sure that its best to use perl - but my perl is not very good - can someone help?

Example below contents of file containing the xml - I basically want to parse the file and have each field contained in a variable..

ie. I want to store the account number in a variable, name in a variable, add in a variable

So I could just echo $accountnumb $name $add etc and get the following
65004 Bob Daly Ireland



XML Sample File

<?xml version="1.0"?>




<po:Message
xmlnso="http://192.168.50.167/cust/api"
xmlns="http://192.168.50.167/cust"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://192.168.50.167/cust/api http://database1.com:1008/cust/api/jsp/xsd/api_v3_xsd.jsp"
>
<po:Response>

<po:Header>
<po:RequestUrl><![CDATA[/cust/api/ACCOUNTDetail]]></po:RequestUrl>
<po:RequestCommand>ACCOUNT_DETAILS</po:RequestCommand>
<po:Version>6</po:Version>
<po:Id>000001</po:Id>
<po:Opid><![CDATA[XXX]]></po:Opid>
<po:Status>
<po:Response>SUCCESS</po:Response>
<po:Text>
<![CDATA[Service successfully completed.]]>
</po:Text>
<po:StatusKey><![CDATA[SERVICE_SUCCESS]]></po:StatusKey>
<po:StatusParams cACCOUNT="0">
</po:StatusParams>
</po:Status>
</po:Header>

<po:Body>
<po:RequestKey>
<po:RequestCust>65004</po:RequestCust>
</po:RequestKey>

<ACCOUNT>
<ACCOUNT>65004</ACCOUNT>
<MasterACCOUNT>65004</MasterACCOUNT>
<SubscriberType>SUBS_TYPE_STANDALONE</SubscriberType>
<Name>Bob Daly</Name>
<Add>Ireland</Add>

<ClassDescs>
<ClassDesc><![CDATA[Social Account]]></ClassDesc>
<TempClassDesc><![CDATA[]]></TempClassDesc>
</ClassDescs>
<ClassChangedDateTime></ClassChangedDateTime>

<PreferredLanguage><![CDATA[EN]]></PreferredLanguage>
<PreferredCurrency><![CDATA[EUR]]></PreferredCurrency>
<CurrentPromotionPlan><![CDATA[]]></CurrentPromotionPlan>
<SubscriptionStatus>ACTIVE</SubscriptionStatus>

<TempStatus>NOT_BLOCKED</TempStatus>
<EocnSelStructId>255</EocnSelStructId>

<Agent><![CDATA[00000000]]></Agent>
<SubAgent><![CDATA[]]></SubAgent>
<DisconnectReason><![CDATA[]]></DisconnectReason>

<DisconReasonText><![CDATA[]]></DisconReasonText>

<BeginDate>16-Jul-2008</BeginDate>

<StartDate>16-Jul-2008</StartDate>

<ServiceRemovalDate>12-Sep-2009</ServiceRemovalDate>

<LastModification>12-Sep-2008 13:24:33</LastModification>




</ACCOUNT>
</po:Body>

</po:Response>
</po:Message>

An example to read those variables in a shell script with awk, adjust it if you want more fields:

Code:

#/bin/sh

awk -F"<|>" '
$2=="ACCOUNT" && NF > 3{s=S3}
$2=="Name"{s=s" "S3}
$2=="Add"{s=s" "$3}
$2=="/ACCOUNT"{print s}
' file | 
while read accountnumb name add; do
  echo "$accountnumb" "$name" "$add"
done
# do something with "$accountnumb" "$name" "$add"
# more commands..

Regards

Thanks for reply - I entered the xml in to a file called file..
Then ran the script you provided but it just returned to command prompt with no output.... Do you know why?

Use nawk, gawk or /usr/xpg4/bin/awk on Solaris.
Do you get any output with this?

Code:

#/bin/sh

awk -F"<|>" '
$2=="ACCOUNT" && NF > 3{s=S3}
$2=="Name"{s=s" "S3}
$2=="Add"{s=s" "$3}
$2=="/ACCOUNT"{print s}
' file

Okay - I now specified to use
/usr/xpg4/bin/awk


I get the following output only when I run the latest script..


Ireland

hmmm - I made a small change to the script and now I get

Bob Daly Ireland

I changed from :
$2=="ACCOUNT" && NF > 3{s=S3}
$2=="Name"{s=s" "S3}
$2=="Add"{s=s" "$3}

to:
$2=="ACCOUNT" && NF > 3{s=S3}
$2=="Name"{s=s" "$3}
$2=="Add"{s=s" "$3}


I dont know much about awk to know why this worked and what to do to get the ACCOUNT to work as well... Can you explain what each row of code in the script is actually doing?

Sorry for the typos, S3 should be $3.
Try this, I've change the internal field separator to split the fields properly if you have spaces in the variables:

Code:

#/bin/sh

OFS=$IFS
IFS=:

/usr/xpg4/bin/awk -F"<|>" '
$2=="ACCOUNT" && NF > 3{s=$3}
$2=="Name"{s=s":"$3}
$2=="Add"{s=s":"$3}
$2=="/ACCOUNT"{print s}
' file |
while read accountnumb name add; do
  echo "$accountnumb" "$name" "$add"
done

IFS=$OFS

# Here you can use the variables $accountnumb" "$name" and "$add"
.
.
exit 0