I have a large XML file and I want to extract blocks from it.

For example, there is a large block:

<tag>
....
</tag>

I want to check if there is a specific string within the block - if yes, then extract the entire block from <tag> to </tag>.

I tried using nawk but I didnt know how to make the "if" part. It's a Sun machine, thus Sun OS.

I do it this way:

:label
/<tag>/,/<\/tag>/ {
/<\/tag>/! {
N;
b label
}
p
}

put the above in a file, say tag.sed and run:

cat xmlfile.xml | sed -n -f tag.sed > outputfile

Regards

Thanks, just one question: Where is the "wanted string" ?

Also I got this when I ran the above:

Unrecognized command: /<\/tag>/! {
Broken Pipe

I am running gnu sed v3.02, maybe it's a sed version issue or a type mismatch? because it works for me.

Instead of just 'p' on line 7 of the tag.sed file it should read
/wanted-string/p

regards

I don't know it still doesn't work and I checked it etc.

I'm using Sun OS 5.8 which has its own pecularities. Any chance you could do the same with Sun's sed?