Unix shell script for harvesting metadata records from OAI-PMH repositories. Runs on macOS and Unix / Linux.

Records are aggregated in a single file, serialized to a single record per line. Optionally, individual records can be compressed to save space.

Use git to clone the data

$ git clone https://github.com/renevoorburg/oai2linerec.git

or simply download the oai2linerec.sh source and make it executable using $ chmod +x oai2linerec.sh.

Invoking it is as easy as

$ oai2linerec.sh -s set -p dc -f 2012-02-03 -o results.txt -b http://base.nl/oai

Enable logging like this

$ oai2linerec.sh -d -s set -p dc -f 2012-02-03 -o results.txt -b http://base.nl/oai 2>harvest.log

For an example of how to process such a single linerec output file, see companion script parselinerec.sh. Gzipped output files are fast and easy to use when indexed using Matt Godbolt's zindex. For parallel processing of linecrec file with shell scripts see A beginners guide to processing 'lots of' data.