running "bazel build //cc/google/fhir_examples:ParsePatients" on Ubuntu creates errors about fhir-examples HOT 12 CLOSED

Python is version 3.8.6, Bazel is verison 3.7.0, Java 11.0.9, Ubuntu 20.10 64 bit.

from fhir-examples.

Hi @KiwiFPVHarry!

Could you confirm whether there's a pip binary installed for your Python interpreter?

python3 --version
python3 -m pip

The Bazel workspace installs pip requirements during setup, and in doing so will require pip to be discoverable on your PATH. Even if you're trying to build for a target in //cc/... you'll need to have an environment that supports the workspace setup.

Based on the error message you may need to install python3-distutils it seems? I'd be curious to see if:

apt list | grep -i distutils

shows anything?

from fhir-examples.

We'll look at to see if there's any way we can split this more cleanly too -- as for someone just interested in //cc/..., it seems unnecessary to also provision Python, go, ..., etc.

from fhir-examples.

after installing pip it has now completed only to get the next error when tryignt o run bazel-bin

bazel build //cc/google/fhir_examples:ParsePatients
Starting local Bazel server and connecting to it...
INFO: Analyzed target //cc/google/fhir_examples:ParsePatients (60 packages loaded, 2452 targets configured).
INFO: Found 1 target...
Target //cc/google/fhir_examples:ParsePatients up-to-date:
bazel-bin/cc/google/fhir_examples/ParsePatients
INFO: Elapsed time: 19.726s, Critical Path: 2.24s
INFO: 1 process: 1 internal.
INFO: Build completed successfully, 1 total action
h@h-virtual-machine:~/fhir-examples$ bazel-bin/cc/google/fhir_examples:ParsePatient $WORKSPACE
bash: bazel-bin/cc/google/fhir_examples:ParsePatient: No such file or directory

from fhir-examples.

h@h-virtual-machine:~/fhir-examples$ bazel-bin/cc/google/fhir_examples:ParsePatient $WORKSPACE
bash: bazel-bin/cc/google/fhir_examples:ParsePatient: No such file or directory

Ah I see -- this looks to be an error in the README.md which I've since corrected internally. The corrected documentation will be included in the next push to the public repo.

Try running:

bazel-bin/cc/google/fhir_examples/ParsePatients $WORKSPACE

Which is just a relative path to the compiled binary created from the bazel target: //cc/google/fhir_examples:ParsePatients.

Let me know if that ends up fixing things for you!

from fhir-examples.

Thank you - running now..should taht take quite some time? running on a Ubuntu VM inside of WIndows

from fhir-examples.

It shouldn't be considerably long, maybe a minute or so, to parse all 1000 records. It should be dumping to STDOUT as well as it's making progress. What are you seeing?

from fhir-examples.

Reading /home/h/fhirdata and blinking cursor...for about 10 minutes already

from fhir-examples.

did just run it again...cursor stops blinking after a few seconds

from fhir-examples.

Yeah something's not right. I'm guessing that the generate_synthea.sh script didn't execute properly, and that you have no files in $WORKSPACE/ndjson. Could you verify if this is true?

ls $WORKSPACE/ndjson/

from fhir-examples.

all good now..have re-run the synthea script after deleting all but top folder and then all the commands executed successfully..:) thnaks for all the help..:)

from fhir-examples.

Glad to hear it! :)

To add some additional context for what you might have encountered -- unfortunately the generate_synthea.sh script can fail certain steps but continue execution and appear as though it succeeded. I noticed, for example, on one Ubuntu VM, that since I didn't have a bazel alias (the correct invocation was bazel-3.2.0), the bazel build @com_google_fhir//java:SplitBundle command would fail but the rest of the script would continue. This resulted in no files output to $WORKSPACE/ndjson.

Because of this, when we eventually would try to execute the compiled ParsePatients binary, we were actually spinning indefinitely within the while loop as no data was ever read, since no files existed within $WORKSPACE/ndjson to begin with. Hence the hanging cursor you were seeing after Reading <workspace> was printed to STDOUT.

I've added additional safety to parse_patients.cc so that it will error and exit clearly if these problems occur. I'll also look into adding some additional safety to generate_synthea.sh.

Glad things are working, thanks for your help and patience! :)

from fhir-examples.