Shell Programming and Scripting

Hi,

I have an XML file with following structure. Between following tags I have pipedelimited records with newline characters (Data1|1|2|3)

<![CDATA[
and
]]>

I need to read the data between above tags so that my output is a flat file with pipedelimited records.

<BOS>
<Header>
<TTC>ABC</TTC>
</Header>
<Payload>
<Data>
<![CDATA[ Data1|1|2|3 \n
Data2|4|5|6 \n
--
--
Data3|3|5|6|7 \n
]]>
<Data>
</Payload>
</BOS>

Thanks
DSR

Code:

 
awk '/<\!/,/]]/{print}' infile | sed -e 's/\<\!\[CDATA\[ //' -e 's/\]\]\>//'

or....

Code:

awk '/\!\[CDATA\[/ , /]]/ {if (!/\!\[CDATA\[/&&!/]]/)print}' file

That was good one but it is ignoring the whole line.
But after, <![CDATA[, there is ,Data1|1|2|3, which need to be displayed.

The command line awk displayed this.....

Data2|4|5|6 \n
--
--
Data3|3|5|6|7 \n

is it not the aim ????

Code:

>cat file
<BOS>
<Header>
<TTC>ABC</TTC> 
</Header>
<Payload>
<Data>
<![CDATA[ Data1|1|2|3 \n
Data2|4|5|6 \n
--
--
Data3|3|5|6|7 \n
]]> 
<Data>
</Payload>
</BOS>

If you see the OP infile, you are missing the one which is on the red when you get your result.

Yes, i haven't seen, it's true...