artodeto's blog about coding, politics and the world

Assuming you have a large xml file (say 400 megabytes) and you want to grep the content inside one tag, which tool would solve this better then sed?

sed -n -e 's/.*\(.*\)<\/my_magicTag>.*/\1/p' myInputFile.xml > myInputFileFilteredByMyMagicTag.xml

So what we are doing? We are telling sed to search for none or a lot of text before "", store none or a lot of text before "". With "\1", we are using the first remembered pattern (since we only use one "()", we only have one in this command). With "\p", we are telling sed to print this out. After that, as usual, we are using ">" to redirect the standard output into a file.