Reorganized documentation regarding normal usage and data model gener…

…ation (#44)

Signed-off-by: Aaron Chong <[email protected]>

README.md

@@ -4,11 +4,6 @@ Collection of json message schemas which bridges the C++ and python components o
 
 # Usage
 
-Download pkg for python model generation
-```bash
-pip3 install datamodel-code-generator
-```
-
 Compile with [colcon](https://colcon.readthedocs.io/en/released/) is recommended
 
 ```bash
@@ -30,10 +25,30 @@ nlohmann::json schema = rmf_api_msgs::schemas::task_state
 
 ```py
 from rmf_api_msgs import schemas
-from rmf_api_msgs.models import task_state
 
 # get schema
 schema = schemas.task_state()
+```
+
+# Python data model generation
+
+To generate the data models based on the schemas, install `datamodel-code-generator`
+
+```bash
+pip3 install datamodel-code-generator
+```
+
+Recompile with [colcon](https://colcon.readthedocs.io/en/released/)
+
+```bash
+colcon build --packages-select rmf_api_msgs
+source install/setup.bash
+```
+
+The generated data models can then be accessed and used for development,
+
+```py
+from rmf_api_msgs.models import task_state
 
 # create task_state model
 booking = task_state.Booking(id = "id0001")