Merge pull request #144 from eWaterCycle/julia

Add Julia support to gprc4bmi

.github/workflows/ci.yml

@@ -22,7 +22,7 @@ jobs:
       run: |
         python -m pip install --upgrade pip wheel
         pip install -r dev-requirements.txt
-        pip install -e .[R]
+        pip install -e .[R,julia]
     - name: Setup Apptainer
       uses: eWaterCycle/setup-apptainer@v2
       with:
@@ -45,13 +45,18 @@ jobs:
       run: |
         Rscript -e "install.packages('remotes')"
         Rscript -e "install.packages('R6')"
+    - name: Install Julia
+      uses: julia-actions/setup-julia@v1
+      with:
+        version: '^1.9'
     - name: Test with pytest
       run: |
         pytest -vv --cov=grpc4bmi --cov-report xml
+      timeout-minutes: 20
     - name: Correct coverage paths
       run: sed -i "s+$PWD/++g" coverage.xml
     - name: SonarCloud analysis
-      uses: sonarsource/sonarcloud-github-action@v1.3
+      uses: sonarsource/sonarcloud-github-action@v2.0.2
       env:
         GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         SONAR_TOKEN: ${{ secrets.SONAR_TOKEN }}

README.md

@@ -24,6 +24,12 @@ on the client (Python) side. If your server model is implemented in Python, do t
 pip install grpc4bmi[R]
 ```
 
+If the model is implemented in Julia, run instead
+
+```bash
+pip install grpc4bmi[julia]
+```
+
 in the server environment. For bleeding edge version from GitHub use
 
 ```bash
@@ -90,6 +96,35 @@ For example with [WALRUS](https://github.com/eWaterCycle/grpc4bmi-examples/tree/
 run-bmi-server --lang R --path ~/git/eWaterCycle/grpc4bmi-examples/walrus/walrus-bmi.r --name WalrusBmi --port 55555
 ```
 
+### Models written in Julia
+
+The grpc4bmi Python package can also run BMI models written in Julia if the model has an implementation of the [BasicModelInterface.jl](https://github.com/Deltares/BasicModelInterface.jl).
+
+Run the Julia model in Python with
+
+```bash
+from grpc4bmi.bmi_julia_model import BmiJulia
+
+mymodel = BmiJulia.from_name('<package>.<model>', 'BasicModelInterface')
+```
+
+For example with [Wflow.jl](https://github.com/Deltares/Wflow.jl/) use
+
+```bash
+# Install Wflow.jl package in the Julia environment managed by the juliacall Python package.
+from juliacall import Main as jl
+jl.Pkg.add("Wflow")
+# Create the model
+from grpc4bmi.bmi_julia_model import BmiJulia
+mymodel = BmiJulia.from_name('Wflow.Model', 'Wflow.bmi.BMI')
+```
+
+A Julia model has to be run locally. It can not be run in the default gRPC client/server Docker container mode because:
+
+1. Julia has no gRPC server implementation
+2. Calling Julia methods from Python gRPC server causes 100% CPU usage and no progress
+3. Calling Julia methods from C++ gRPC server causes segmentation faults
+
 ### The client side
 
 The client side has only a Python implementation. The default BMI client assumes a running server process on a given port.
@@ -154,7 +189,3 @@ pip install -e .[docs]
 
 and install the C++ runtime and `protoc` command as described in <https://github.com/google/protobuf/blob/master/src/README.md>.
 After this, simply executing the `proto_gen.sh` script should do the job.
-
-## Future work
-
-More language bindings are underway.

docs/container/building.rst

@@ -68,6 +68,11 @@ The WALRUS model has a `Dockerfile`_  file which can be used as an example.
 
 .. _Dockerfile: https://github.com/eWaterCycle/grpc4bmi-examples/blob/master/walrus/Dockerfile
 
+Julia
+-----
+
+A Julia model can not be run as a server, see https://github.com/eWaterCycle/grpc4bmi/blob/main/README.md#model-written-in-julia .
+
 C/C++/Fortran
 -------------
 

grpc4bmi/bmi_client_docker.py

@@ -102,7 +102,8 @@ def __init__(self, image: str, work_dir: str, image_port=50051, host=None,
         super(BmiClientDocker, self).__init__(BmiClient.create_grpc_channel(port=port, host=host), timeout=timeout)
 
     def __del__(self):
-        self.container.stop()
+        if hasattr(self, 'container'):
+            self.container.stop()
 
     def logs(self) -> str:
         """Returns complete combined stdout and stderr written by the Docker container.

grpc4bmi/bmi_client_subproc.py

@@ -15,19 +15,19 @@ class BmiClientSubProcess(BmiClient):
     >>> mymodel = BmiClientSubProcess(<PACKAGE>.<MODULE>.<CLASS>)
     """
 
-    def __init__(self, module_name, path=None, timeout=None):
+    def __init__(self, module_name, path=None, timeout=None, delay=1):
         host = "localhost"
         port = BmiClient.get_unique_port(host)
         name_options = ["--name", module_name]
         port_options = ["--port", str(port)]
         path_options = ["--path", path] if path else []
         self.pipe = subprocess.Popen(["run-bmi-server"] + name_options + port_options + path_options, env=dict(os.environ))
-        time.sleep(1)
+        time.sleep(delay)
         super(BmiClientSubProcess, self).__init__(BmiClient.create_grpc_channel(port=port, host=host), timeout=timeout)
 
     def __del__(self):
-        self.pipe.terminate()
-        self.pipe.wait()
+        self.pipe.kill()
+        self.pipe.wait(timeout=0.1)
 
     def get_value_ref(self, var_name):
         raise NotImplementedError("Cannot exchange memory references across process boundary")