TDB code has been removed, but can be reverted through commit fa97f914c69935e07e6521c4b15be4a0e3b7db89.
Minor overhaul of shared test resources to reduce chance of overlapping test resources (now individual modules should only not have a shared directory themselves).
Creation of HDT was done using the vibe-3.1.0 TTL files (so not from the start), so output should be identical (except perhaps the order of diseases with the same score within the results of a single gene).
Readme is updated.
Testing notes
New HDT still gives same results as original TDB (script was adjusted to use the optimized HDT instead of optimized TDB):
$ sh TestOptimizedQueries.sh -or ~/Programming/data/vibe/database_creation/vibe-5.0.0-sources-tdb/ -op ~/Programming/data/vibe/database_creation/vibe-5.0.0-hdt/vibe-5.0.0.hdt
### Running original TDB/query.
Time: 74,405 sec
### Running optimized HDT/query.
### Validating if optimized TDB/query output files are equal to their original counterparts.
test/genes_for_hpo-optimized.tsv: OK
Some new tests are not supported on Jenkins (due to changing file/dir permissions within a unit test), so a tag skipOnJenkins has been created and -DexcludedGroups='skipOnJenkins' was added to the Jenkinsfile on several places. These tests are still run locally by default and mvn clean install did not cause any issues.
md5 of the final .tsv files seem to differ between the TDB & HDT files, though multiple runs for a single format do not differ. It seems however that the cause of this might be that certain genes (or diseases for a single gene) with the same score are outputted in a different order.
Changes
shared
directory themselves).vibe-3.1.0
TTL files (so not from the start), so output should be identical (except perhaps the order of diseases with the same score within the results of a single gene).Testing notes
skipOnJenkins
has been created and-DexcludedGroups='skipOnJenkins'
was added to the Jenkinsfile on several places. These tests are still run locally by default andmvn clean install
did not cause any issues..tsv
files seem to differ between the TDB & HDT files, though multiple runs for a single format do not differ. It seems however that the cause of this might be that certain genes (or diseases for a single gene) with the same score are outputted in a different order.