1  Quick start guide

1.1 Installation

See INSTALL.md for dependencies.

  1. wget https://github.com/pkiraly/metadata-qa-marc/releases/download/v0.6.0/metadata-qa-marc-0.6.0-release.zip
  2. unzip metadata-qa-marc-0.6.0-release.zip
  3. cd metadata-qa-marc-0.6.0/

1.2 Configuration

Either use the script qa-catalogue or create configuration files:

  1. cp setdir.sh.template setdir.sh

Change the input and output base directories in setdir.sh. Local directories input/ and output/ will be used by default. Files of each catalogue are in a subdirectory of theses base directories:

  1. Create configuration based on some existing config files:
  • cp catalogues/loc.sh catalogues/[abbreviation-of-your-library].sh
  • edit catalogues/[abbreviation-of-your-library].sh according to configuration guide

1.3 With Docker

A more detailed instruction how to use qa-catalogue with Docker can be found in the wiki

A Docker image bundling qa-catalogue with all of its dependencies and the web interface qa-catalogue-web is made available:

To download, configure and start an image in a new container the file docker-compose.yml is needed in the current directory. It can be configured with the following environment variables:

  • IMAGE: which Docker image to download and run. By default the latest image from Docker Hub is used (pkiraly/metadata-qa-marc). Alternatives include

    • IMAGE=ghcr.io/pkiraly/qa-catalogue:main for most recent image from GitHub packages
    • IMAGE=metadata-qa-marc if you have locally build the Docker image

  • CONTAINER: the name of the docker container. Default: metadata-qa-marc.

  • INPUT: Base directory to put your bibliographic record files in subdirectories of. Set to ./input by default, so record files are expected to be in input/$NAME.

  • OUTPUT: Base directory to put result of qa-catalogue in subdirectory of. Set to ./output by default, so files are put in output/$NAME.

  • WEBCONFIG: directory to expose configuration of qa-catalogue-web. Set to ./web-config by default. If using non-default configuration for data analysis (for instance PICA instead of MARC) then you likely need to adjust configuration of the web interface as well. This directory should contain a configuration file configuration.cnf.

  • WEBPORT: port to expose the web interface. For instance WEBPORT=9000 will make it available at http://localhost:9000/ instead of http://localhost/.

  • SOLRPORT: port to expose Solr to. Default: 8983.

Environment variables can be set on command line or be put in local file .env, e.g.:

WEBPORT=9000 docker compose up -d

or

docker compose --env-file config.env up -d

When the application has been started this way, run analyses with script ./docker/qa-catalogue the same ways as script ./qa-catalogue is called when not using Docker (see usage for details). The following example uses parameters for Gent university library catalogue:

./docker/qa-catalogue \
  --params "--marcVersion GENT --alephseq" \
  --mask "rug01.export" \
  --catalogue gent \
  all

Now you can reach the web interface (qa-catalogue-web) at http://localhost:80/ (or at another port as configured with environment variable WEBPORT). To further modify appearance of the interface, create templates in your WEBCONFIG directory and/or create a file configuration.cnf in this directory to extend UI configuration without having to restart the Docker container.

This example works under Linux. Windows users should consult the Docker on Windows wiki page. Other useful Docker commands at QA catalogue’s wiki.

Everything else should work the same way as in other environments, so follow the next sections.

1.4 Use

catalogues/[abbreviation-of-your-library].sh all-analyses
catalogues/[abbreviation-of-your-library].sh all-solr

For a catalogue with around 1 million record the first command will take 5-10 minutes, the later 1-2 hours.