Add documentation for building, running and testing benchcab

docs/dev_docs.md

@@ -0,0 +1,69 @@
+# Developer Guide
+## Building
+### Dependencies
+- conda env
+```sh
+module load conda
+conda env create --name benchcab-dev --file .conda/benchcab-dev.yaml
+conda activate benchcab-dev
+pip install --user -e .
+```
+
+### Dev-Dependencies
+- `black`, `ruff`, `autoformatter`
+
+!!! tip
+    Add a docstring using numpy style. (autodocstring extension in VS Code can pre-populate for you, make sure to select numpy style on settings first).
+
+### Documentation
+- mkdocs
+
+```sh
+pip install -r mkdocs-requirements.txt
+```
+
+!!! tip 
+    Using `mkdocs` on Gadi 
+    Use ssh port forwarding
+    ```sh
+    ssh -L 8000:localhost:8000 <username>@gadi.nci.org.au
+    ```
+
+```sh
+mkdocs serve
+```
+
+## Running
+
+Follow the user guide
+
+## Testing
+- Static Type Checking  `mypy`
+- Unit tests `pytest`
+- Integration tests
+```sh
+/bin/bash benchcab/data/test/integration.sh
+```
+
+### Comparing Models
+
+```sh
+module load <name>
+```
+
+Outputs are stored in `$SRC_DIR/runs/outputs/`
+
+- `netcdf`
+
+```sh
+ncdump -h
+```
+
+- `nccmp`/`cdo` - Bitwise comparison
+
+```sh
+cdo </path/to/output1> </path/to/output2> 
+```
+
+!!! tip
+    Generally, try to keep a separate copy of `output2` from the `main` branch and `output1` from the feature branch to compare the ouputs, thus checking the validity of integration testing.

mkdocs.yml

@@ -75,7 +75,9 @@ nav:
   - About: index.md
   - User Guide:
       - user_guide/index.md
+      - user_guide/dev_docs.md
       - user_guide/config_options.md
       - user_guide/default_science_configurations.md
       - user_guide/running_CABLE_v2.md
       - user_guide/expected_output.md
+  - Developer Guide: dev_docs.md