Update config model docs

docs/developer/meta/config-models.md

@@ -2,9 +2,9 @@
 
 -----
 
-All integrations use [pydantic](https://github.com/samuelcolvin/pydantic) models as the primary way to validate and interface with configuration.
+All integrations use [pydantic](https://github.com/pydantic/pydantic) models as the primary way to validate and interface with configuration.
 
-As config spec data types are based on OpenAPI 3, we [automatically generate](https://github.com/koxudaxi/datamodel-code-generator) the necessary code.
+As [config spec](config-specs.md) data types are based on OpenAPI 3, we [automatically generate](https://github.com/koxudaxi/datamodel-code-generator) the necessary code.
 
 The models reside in a package named `config_models` located at the root of a check's namespaced package. For example, a new integration named `foo`:
 
@@ -26,25 +26,22 @@ foo
 
 There are 2 possible models:
 
-- `SharedConfig` (ID: `shared`) that corresponds to the `init_config` section
 - `InstanceConfig` (ID: `instance`) that corresponds to a check's entry in the `instances` section
+- `SharedConfig` (ID: `shared`) that corresponds to the `init_config` section that is shared by all instances
 
 All models are defined in `<ID>.py` and are available for import directly under `config_models`.
 
 ## Default values
 
 The default values for optional settings are populated in `defaults.py` and are derived from the
-[value](../meta/config-specs.md#values) property of config spec options.
-
-The precedence is:
-
-1. the `default` key
-2. the `example` key, if it appears to represent a real value rather than an illustrative example and the `type` is a primitive
-3. the default value of the `type` e.g. `string` -> `str()`, `object` -> `dict()`, etc.
+[value](config-specs.md#values) property of config spec options. The precedence is the `default` key
+followed by the `example` key (if it appears to represent a real value rather than an illustrative example
+and the `type` is a primitive). In all other cases the default will be `None` and there will be no default
+getter function.
 
 ## Validation
 
-The validation of fields for every model occurs in 6 stages.
+The validation of fields for every model occurs in 3 high-level stages.
 
 ### Initial
 
@@ -56,46 +53,55 @@ def initialize_<ID>(values: dict[str, Any], **kwargs) -> dict[str, Any]:
 If such a validator exists in `validators.py`, then it is called once with the raw config that was supplied by the user.
 The returned mapping is used as the input config for the subsequent stages.
 
-### Default value population
+### Field
 
-If a field was not supplied by the user nor during the initialization stage, then its default value is
-taken from `defaults.py`. This stage is skipped for required fields.
+The value of each field goes through the following steps.
 
-### Default field validators
+#### Default value population
 
-At this point `pydantic` will parse the values and perform validation of types, etc.
+If a field was not supplied by the user nor during the [initialization stage](#initial), then its default value is
+taken from `defaults.py`. This stage is skipped for required fields.
 
-### Custom field validators
+#### Custom field validators
 
 The contents of `validators.py` are entirely custom and contain functions to perform extra validation if necessary.
 
 ```python
-def <ID>_<OPTION_NAME>(value: Any, *, field: pydantic.fields.ModelField, **kwargs) -> Any:
+def <ID>_<OPTION_NAME>(value: Any, *, field: pydantic.fields.FieldInfo, **kwargs) -> Any:
     ...
 ```
 
-Such validators are called for the appropriate field of the proper model if the option was supplied by the user.
+Such validators are called for the appropriate field of the proper model. The returned value is used as the
+new value of the option for the subsequent stages.
 
-The returned value is used as the new value of the option for the subsequent stages.
+!!! note
+    This only occurs if the option was supplied by the user.
 
-### Pre-defined field validators
+#### Pre-defined field validators
 
-A new `validators` key under the [value](https://datadoghq.dev/integrations-core/meta/config-specs/#values) property of config
+A `validators` key under the [value](https://datadoghq.dev/integrations-core/meta/config-specs/#values) property of config
 spec options is considered. Every entry will refer to a relative import path to a [field validator](#custom-field-validators)
 under `datadog_checks.base.utils.models.validation` and is executed in the defined order.
 
-The last returned value is used as the new value of the option for the [final](#final) stage.
+!!! note
+    This only occurs if the option was supplied by the user.
+
+#### Conversion to immutable types
+
+Every `list` is converted to `tuple` and every `dict` is converted to [types.MappingProxyType](https://docs.python.org/3/library/types.html#types.MappingProxyType).
+
+!!! note
+    A field or nested field would only be a `dict` when it is defined as a mapping with arbitrary keys. Otherwise, it would be a model with its own properties as usual.
 
 ### Final
 
 ```python
-def finalize_<ID>(values: dict[str, Any], **kwargs) -> dict[str, Any]:
+def check_<ID>(model: pydantic.BaseModel) -> pydantic.BaseModel:
     ...
 ```
 
-If such a validator exists in `validators.py`, then it is called with the cumulative result of all fields.
-
-The returned mapping is used to instantiate the model.
+If such a validator exists in `validators.py`, then it is called with the final constructed model. At this point it cannot
+be mutated so you may only raise errors.
 
 ## Loading
 
@@ -116,17 +122,16 @@ class Check(AgentCheck, ConfigMixin):
     ...
 ```
 
-It exposes the instantiated `InstanceConfig` model at `self.config` and `SharedConfig` model at `self.shared_config`.
+It exposes the instantiated `InstanceConfig` model as `self.config` and `SharedConfig` model as `self.shared_config`.
 
 ## Immutability
 
-All generated models are [configured as immutable](https://pydantic-docs.helpmanual.io/usage/models/#faux-immutability).
-Additionally, every `list` is converted to `tuple` and every `dict` is converted to [immutables.Map](https://github.com/MagicStack/immutables).
+In addition to each field being [converted to an immutable type](#conversion-to-immutable-types), all generated models are [configured as immutable](https://docs.pydantic.dev/2.0/usage/models/#faux-immutability).
 
 ## Deprecation
 
 Every option marked as deprecated in the config spec will log a warning with information about when it will be removed and what to do.
 
 ## Enforcement
 
-A validation command `ddev validate models` runs in our CI. To locally generate the proper files, run `ddev validate models [CHECK] --sync`.
+A validation command [`validate models`](../ddev/cli.md#ddev-validate-models) runs in our CI. To locally generate the proper files, run `ddev validate models [INTEGRATION] --sync`.