Molecule transparently runs a unified Scala interface with only minor differences against the full matrix of Datomic systems/apis/languages which from a Molecule perspective boils down to 3 systems:
| System | API | Protocol | Language | Download | License |
|---|---|---|---|---|---|
| peer | datomic.api | free | Clojure/Java | On-Prem free | Free |
| peer-server | datomic.client.api | dev | Clojure | On-Prem starter/pro | Perpetual use, 1-year upgrades |
| dev-local (cloud) | datomic.client.api | dev | Clojure | Dev-tools | Email reg |
Molecule is published with only a dependency on Datomic Free to be freely distributed. To run tests only against the free version, please follow the instructions in README_free.md
To run tests against the two proprietary systems too, please follow the steps below (some steps are only required once during setup).
In order for Intellij to be happy, set VM option -Xmx2G in preferences:
Preferences > Build, Execution, Deployment > Compiler > Scala Compiler > Scala Compiler Server / VM options
Ensure that the java version (like 1.8) here is the same as in your terminal that you get from java -version
Register and install the two last proprietary Datomic systems from the links shown above into some folder locally.
Install libraries to your local maven repository (.m2):
- Run
bin/maven-installin datomic-pro-[version] - Run
./installin cognitect-dev-tools-[version]
Open project.SettingsDatomic and set datomicDownloadsDir to the path of the directory where you downloaded the Datomic distributions.
In project.Settings, add "-SNAPSHOT" to the molecule version version in ThisBuild := "<molecule-version>-SNAPSHOT" and run sbt publishLocal from the molecule project directory. Let your projects use this proprietary-aware molecule version.
Molecule will automatically detect and default to using the latest available Datomic pro version if this is downloaded.
To enforce using the free version, you can compile molecule with this flag:
sbt compile -Dfree=true
To enforce using a specific version, you can compile molecule with one or both of these flags:
sbt compile -Ddatomic.pro=1.0.6362 -Ddatomic.dev-local=1.0.242
Start Datomic transactor (update path/version number):
cd <path-to-datomic-downloads>/datomic-pro-1.0.6362
bin/transactor config/samples/dev-transactor-template.properties
While the transactor is running, create a new tab/process in the terminal to run the next steps:
Run commands below to perform these 3 operations:
-
Install
datomic-samplesfor dev-local -
Download and unpack mbrainz-1968-1973 special sample for peer and peer-server
-
Restore mbrainz sample
wget https://datomic-samples.s3.amazonaws.com/datomic-samples-2020-07-07.zip unzip datomic-samples-2020-07-07.zip rm datomic-samples-2020-07-07.zip wget https://s3.amazonaws.com/mbrainz/datomic-mbrainz-1968-1973-backup-2017-07-20.tar -O mbrainz.tar tar -xvf mbrainz.tar -C datomic-samples rm mbrainz.tar bin/datomic restore-db file://
pwd/datomic-samples/mbrainz-1968-1973 datomic:dev://localhost:4334/mbrainz-1968-1973
In the IDE, run moleculeTests.restore.RecreateTestDbs in jvm moleculeTests (main) to create sample databases for the Peer Server.
Copy all lines below and paste into terminal to start the Peer Server and having it serve all the sample databases used by the tests:
bin/run -m datomic.peer-server -h localhost -p 8998 -a k,s \
-d m_coretests,datomic:mem://m_coretests \
-d m_bidirectional,datomic:mem://m_bidirectional \
-d m_partitions,datomic:mem://m_partitions \
-d m_nested,datomic:mem://m_nested \
-d m_selfjoin,datomic:mem://m_selfjoin \
-d m_aggregates,datomic:mem://m_aggregates \
-d m_socialNews,datomic:mem://m_socialNews \
-d m_graph,datomic:mem://m_graph \
-d m_graph2,datomic:mem://m_graph2 \
-d m_modernGraph1,datomic:mem://m_modernGraph1 \
-d m_modernGraph2,datomic:mem://m_modernGraph2 \
-d m_productsOrder,datomic:mem://m_productsOrder \
-d m_seattle,datomic:mem://m_seattle \
-d mbrainz-1968-1973,datomic:dev://localhost:4334/mbrainz-1968-1973
Now you can run molecule tests or projects against peer, peer-server and dev-local (cloud).
Run jvm tests in sbt:
sbt
project moleculeTestsJVM
// Run all tests (against scala 2.13 as default)
sbt:moleculeTests> test
// Run selection of tests (against scala 2.13 as default)
sbt:moleculeTests> testOnly moleculeTests.tests.*
sbt:moleculeTests> testOnly moleculeTests.tests.core.ref.*
sbt:moleculeTests> testOnly moleculeTests.tests.core.ref.TwoStepQueries
// Test against scala 2.12
sbt:moleculeTests> ++2.12.15; test
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.*
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.core.ref.*
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.core.ref.TwoStepQueries
Remember to ctrl-c the sbt process to avoid process locks if testing in IDE.
To run molecule tests from the client/js, a jvm server has to be running to receive ajax calls from the client. It can be of any stack, as long as it can respond to ajax calls. In our test setup we use MoleculeRpcServer which is just a simple Akka-Http server with this only purpose. But it can be your server running just serving molecule ajax requests too.
In one process in the terminal we start the MoleculeRpcServer with
sbt
moleculeTestsJVM/run
And in another we run the tests, this time just choosing the JS project. From here on, all tests are run as with the jvm.
sbt
project moleculeTestsJS
// Run all tests (against scala 2.13 as default)
sbt:moleculeTests> test
// Run selection of tests (against scala 2.13 as default)
sbt:moleculeTests> testOnly moleculeTests.tests.*
sbt:moleculeTests> testOnly moleculeTests.tests.core.ref.*
sbt:moleculeTests> testOnly moleculeTests.tests.core.ref.TwoStepQueries
// Test against scala 2.12
sbt:moleculeTests> ++2.12.15; test
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.*
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.core.ref.*
sbt:moleculeTests> ++2.12.15; testOnly moleculeTests.tests.core.ref.TwoStepQueries
Info about the mbrainz sample data: