python 3.8 or higher
pip install pyyaml requests tx-functional
To use the cli tools, if your mapper runs on the local host, you can put your spec under config and put your custom python functions under a sub dir in that dir and specify configDir in the command line. The cli will automatically copy your spec and custom python functions to that dir.
If your mapper runs on a remote host, you need to manually copy them to the config dir.
Start your pds server.
There is a cli.py. You can run it as
python cli.py <file containing your spec> <a yaml file containing resource types> <a yaml file contains pids> <timestamp> --libraryPath <sub dir containing custom python functions> --nthreads <number of threads> --level <level> --pdsHost <pds plugin host> --pdsPort <pds plugin port> --configDir <mapper config dir>
For example,
python cli.py spec.py ../resourceTypes.yaml ../patientIds.yaml "2000-01-01T00:00:00Z" --libraryPath module --nthreads 4 --level 1 --pdsHost pds --pdsPort 1234 --configDir ../pdspi-mapper-parallex-example/config
In this example, you would put your spec in config/spec.py. Any python module under the config/modules directory can be imported in your spec. For example, if you have config/modules/clivar.py, you can reference functions in that module in various ways, for example from clivar import *. ../resourceTypes.yaml contains a list of resource types. ../patientIds.yaml contains a list of patient ids. Your spec should output the format that the api specifies. See config/spec4.py's return statement for example. <level> specifies the number of nested for loops to parallelize. It automatically copies the config files in the config dir to ../pdspi-mapper-parallex-example/config.