Shell Programming and Scripting

Hi,

I have this text in a file where I need to move part of the text....

<Relation1 OriginatingObjectID="Holding_1" RelatedObjectID="Party_1" id="Relation_1">
<OriginatingObjectType tc="4">Holding</OriginatingObjectType>
<RelatedObjectType tc="6">Party</RelatedObjectType>
<RelationRoleCode tc="32">Insured</RelationRoleCode>
</Relation1>
<Relation2 OriginatingObjectID="Holding_2" RelatedObjectID="Party_2" id="Relation_2">
<OriginatingObjectType tc="4">Holding</OriginatingObjectType>
<RelatedObjectType tc="6">Party</RelatedObjectType>
<RelationRoleCode tc="32">Insured</RelationRoleCode>
</Relation2>

I want only the first line of each block in this text to be modified to

<Relation1 id="Relation_1" OriginatingObjectID="Holding_1" RelatedObjectID="Party_1">
similarly for second block ..
<Relation1 id="Relation_2" OriginatingObjectID="Holding_2" RelatedObjectID="Party_2">
and so on....
I tried it this way but could not get the complete solution...
grep "Relation1 " New_Text.txt | awk '{ print "<Relation " $4" "$2" "$3">"}'
but The result is ..
<Relation id="Relation_1"> OriginatingObjectID="Holding_1" RelatedObjectID="Party_1">
I do not want "id="Relation_1"> " instead I want id="Relation_1" This ">"should not appear after "Relation_1"...and how do I implement it in a text file where I have many blocks to be modified...and also have other blocks which does not need any change..?

>cat ty.awk

Code:

BEGIN {
FS=" "
}
{
if( $4 ~ /Relation_/ )
{
	print $1" "$4" "$2" "$3">"
}
else
{
	print $0">"
}
}

Code:

sed 's/>$//' file | awk -f ty.awk

That works perfectly Thanks a lot......

Why the sed and the awk ?

Just the sed would do as well.

Code:

sed -e 's_^\(<Relation[0-9]*\) \(.*\) \(.*\) \(.*\)>$_\1 \4 \2 \3>_g' input.txt

This works with slight modification...
sed -e 's_\(<Relation[0-9]*\) \(.*\) \(.*\) \(.*\)>$_\1 \4 \2 \3>_g' input.txt

Thanks.