This is a Java library for accessing a Diamond-flavour ISPyB database. It's designed for server-side-only usage, utilising stored procedures to access the database tables. A key benefit of this approach is the ability to modify the behaviour of the library methods without downtime and without building and deploying a new version of the library itself. A simple change to a stored procedure is very quick and runs in a transaction without disrupting access to the database.
The stored procedures are maintained in a separate repository: ispyb-database.
- MariaDB 10.3 or later
- JDK 1.8 (also appears to work with 11) JRE and development packages (e.g. on Fedora: java-1.8.0-openjdk java-1.8.0-openjdk-devel)
- Maven
The javadocs are currently hosted at https://alfred.diamond.ac.uk/documentation/javadocs/gda-ispyb-api/.
For example usage, see the integration tests, e.g. PlateIntegrationTest.
To create the database user, run e.g. these SQL commands:
CREATE USER maven@localhost IDENTIFIED BY 'password_here';
GRANT SELECT, INSERT, UPDATE, DELETE, CREATE, DROP, REFERENCES, INDEX, ALTER, CREATE TEMPORARY TABLES, LOCK TABLES, EXECUTE, CREATE VIEW, SHOW VIEW, CREATE ROUTINE, ALTER ROUTINE, EVENT, TRIGGER ON `maven\_%`.* TO 'maven'@'localhost';
To build and test the JARs:
mvn -Dispyb.url={jdbc_url} -Dispyb.user={user} -Dispyb.pw={password} -Dispyb.host={host} -Dispyb.port={port} package
Example:
mvn -Dispyb.url='jdbc:mariadb://localhost:3306' -Dispyb.user=maven -Dispyb.pw='password_here' -Dispyb.host=localhost -Dispyb.port=3306 package
To run a particular test class:
mvn -Dispyb.url={jdbc_url} -Dispyb.user={user} -Dispyb.pw={password} -Dispyb.host={host} -Dispyb.port={port} -Dtest={TestClassName} test
To generate javadocs:
mvn -Dispyb.url={jdbc_url} -Dispyb.user={user} -Dispyb.pw={password} -Dispyb.host={host} -Dispyb.port={port} install
The javadocs will then appear in a folder called target/apidocs/
.
To release, use the script release.sh
with the argument major
, minor
, or bugfix
. This script will handle updating the version. The result will be added to GitHub as a release by Azure Pipelines.