Sometimes you have a BIG XML, and you want, like, a lot of little XMLs?
Which is funny, because the little XMLs each describe something different, and yet someone has herded them into one enormous tree like an infinite number of monkeys.
And SOMETIMES, that XML tree full of monkeys is SO BIG that it's bigger than your RAM. Someone should do something about your XML getting so BIG all the time!
Oh well, until that happens, you can use this command line tool. Just install sbt, check out this project, go to the xmll directory and do:
sbt "runMain dpla.xmll.Main <name of record element> <infile> <outfile>"
The outfile will end up containing one row for every that is found at a sibling level as the first one found. Each line will contain the XML corresponding to that element and it's descendants. All newlines in the xml will be replaced with spaces.
If you'd like, you can package the project up into a portable JAR file using the command sbt assembly. The JAR will be saved at the path target/cala-2.13/xmll-assembly-0.1.jar, and then you can copy it to wherever a java install is handy, and run it with:
java -jar xmll-assembly-0.1.jar <name of record element> <infile> <outfile>
This software is able to process over a gigabyte of input xml per minute on this laptop that I'm using to write this. YMMV.
Happy trees!