Skip to content


Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Shell OAI Harvester

The OAI-PMH Shell Harvester is able to harvest OAI-PMH targets. It supports multiple configurable targets which can be updated individually. Furthermore, it is able to execute a preset command for each record it updates or deletes. View the CHANGELOG for important changes.


Use sudo/root where needed.


  • bash
  • curl
  • coreutils (cat, cut, head, date, printf, test)
  • grep
  • sed
  • xmllint (optional for validating responses)
  • xsltproc (libxslt)


  • git clone
  • cd shell-oaiharvester
  • cp oaiharvester /usr/bin/oaiharvester
  • mkdir /usr/share/shell-oaiharvester
  • cp libs/* /usr/share/shell-oaiharvester/.
  • mkdir /etc/shell-oaiharvester
  • cp config.example.xml /etc/shell-oaiharvester/config.xml



Package available in the flavour overlay.



The harvester has several configuration options that apply to the general behaviour while other options are repository specific. Different config files can be used with different levels of precedence (without fallback).

  1. commandline option provided location: oaiharvester -c <config file>
  2. $HOME/.config/shell-oaiharvester/config.xml
  3. /etc/shell-oaiharvester/config.xml

Some notes on the options:

  • deletecmd: executed before a record is deleted, optional
  • updatecmd: executed after a record is updated, optional
    • ${identifier} holds the identifier
    • ${filename) is the identifier with optional xz extension
    • ${path} holds the absolute path of the record
    • ${responsedatetime} the full oai page responsedate
    • ${responsedate} the YYYY-MM-DD format of responsedatetime
  • repository update and delete cmd's are executed before the generic ones if available
  • set, from, until, conditional and recordpath are optional


View all available options: oaiharvester -h

To harvest records from a specified repository, run: oaiharvester -r <repository_id>

To only retrieve identifiers: oaiharvester -r <repository_id> -n

Not harvest, but validate the repository oaiharvester -r <repository_id> -t

To perform strict validation use: oaiharvester -r <repository_id> --test-strict

To list configured repositories: oaiharvester -l


Unless customized, the log file of the harvest process is stored at /tmp/oaiharvester-log.csv. The logger uses a simple csv format with a line for each downloaded page: YYYY-MM-DD HH:MM:SS,PID,repository,record count,download time,process time.

A recordlog can also be set, by default at /dev/null because depending on repositories this file can increase in size a lot. This has the following format: YYYY-MM-DD HH:MM:SS,repository,record datestamp,record identifier.


Once harvested, each repository directory will hold a status file. In version 1, this is lasttimestamp.txt containing just the timestamp where the harvester should continue. Starting from version 2, a hidden .oaiharvester file is used to contain not only the timestamp, but also harvester version information which should help with future migrations.