Read files as named graphs and set default graph as union of named graphs.

Add -m max time for query execution. Attempt to fix FROMs not being sent to remote SPARQL endpoint.

Use logger rather than out/err

Move the uncurry to later to eagerly parse the config and load the data from files only once.

Remove debugging.

Move error reason to message, not stack-trace.

-i -> ignore errors (exceptions) rather than failures & errors.

Override default test directory.

Some refactoring to add tests to figure out the reason for exceptions.

Allow multiple test directories to be specified.

Bump Jena -> 3.17.0, redo build.sbt and start adding GraalVM build.

Use Using properly. Add stacktrace to exception/error.

Clean & use Using()

Bump versions: scala -> 2.13.4; scopt -> .0.0; scala-xml -> 1.3.0; sbt -> 1.4.7.

Don't use prettyprinter so to avoid adding whitespace in error/failure reasons.

Put output message in body of <failure>, using CDATA to preserve line endings.

Found the JUnit/Jenkins reverse engineered docs; updated to distinguish errors & failures.

Add -l option to add/set limits on SELECT queries

Add --from to be able to specify one or more graphs to restrict queries to (currently over HTTP)

Attempt to add limited oAuth support via supplied token.

Remove ENTRYPOINT as it messes things up.

Create Docker image

Bump scala to 2.12.8 and sbt to 1.2.8. Create sparql-test-runner jar and make it run the test runner directly.

Add -p option to bind query vars directly by replacing given vars in query text.

Catch exceptions running queries.

Add APLv2

Bump version to 1.3

Allow basic auth against remote endpoint.

Fix comparing absolute file path with relative one.

Add option to alternatively run against a remote SPARQL endpoint.