Experimenting with DMCG flags to have more readable pydantic models

.gitignore

@@ -170,4 +170,10 @@ cython_debug/
 # PyPI configuration file
 .pypirc
 
-.envrc
+.envrc
+
+private
+
+# build dir contains the cloned specs
+# from which we generate the pydantic classes
+build

README.md

@@ -1,27 +1,33 @@
 # Pydantic EDA
 
-**WIP!**
-
 Pydantic models for EDA OpenAPI spec. Models are generated for the EDA Core API as well as for the apps shipped with EDA Playground at the time of the generation.
 
 ## Usage
 
 ## Generation
 
-Store Github auth token in a `GH_AUTH_TOKEN` environment variable. For example, with `gh` cli:
+Install dev dependencies:
 
 ```
-export GH_AUTH_TOKEN=$(gh auth token)
+uv sync --all-groups
 ```
 
-Install dev dependencies:
+Generate models for a specific version of the openapi repo (git ref):
 
 ```
-uv add --dev 'datamodel-code-generator[http]' rich ruff httpx
+python gen_models.py --version v25.4.1
 ```
 
-Generate models for a specific version of openapi repo:
+The generation script clones the eda-labs/openapi repo at the provided ref and generated models from there.
 
-```
-python gen_models.py --version v24.12.1
-```
+## Modifications
+
+The generation script transforms all schema objects in the source openapi files by removing `com.nokia.eda.<name>.<version>`, as DMCG project has issues with treating schema nodes with dots in their names as module-based schemas. Therefore, the original schema nodes undergo that mutation by the script.
+
+## Versions
+
+The following table matches the project version with the version of the EDA delivery from which the models were generated.
+
+| pydantic_eda | EDA release |
+| ------------ | ----------- |
+| 0.3.0        | 25.4.1      |