Add __hash__ to Pydantic v2 models used in sets (#2918)

* Add __hash__ to Pydantic v2 models used in sets

* Fix lint errors: refactor to classmethods and reduce nesting

* Fix cross-module set item hash detection

* Refactor: use generic class_body_lines instead of set_item_hashable

Author: koxudaxi

src/datamodel_code_generator/model/base.py

@@ -678,6 +678,7 @@ def __init__(  # noqa: PLR0913
 
         self.reference.source = self
 
+        self.extra_template_data: dict[str, Any]
         if extra_template_data is not None:
             # The supplied defaultdict will either create a new entry,
             # or already contain a predefined entry for this type

src/datamodel_code_generator/model/template/pydantic_v2/BaseModel.jinja2

@@ -7,14 +7,17 @@ class {{ class_name }}({{ base_class }}):{% if comment is defined %}  # {{ comme
     {{ description | escape_docstring | indent(4) }}
     """
 {%- endif %}
-{%- if not fields and not description and not config %}
+{%- if not fields and not description and not config and not class_body_lines %}
     pass
 {%- endif %}
 {%- if config %}
 {%- filter indent(4) %}
 {% include 'ConfigDict.jinja2' %}
 {%- endfilter %}
 {%- endif %}
+{%- for line in class_body_lines %}
+    {{ line }}
+{%- endfor %}
 {%- for field in fields %}
     {%- if not field.annotated and field.field %}
     {{ field.name }}: {{ field.type_hint }} = {{ field.field }}

src/datamodel_code_generator/model/template/pydantic_v2/RootModel.jinja2

@@ -27,7 +27,10 @@ class {{ class_name }}({{ base_class }}{%- if fields -%}[{{get_type_hint(fields,
 {% include 'ConfigDict.jinja2' %}
 {%- endfilter %}
 {%- endif %}
-{%- if not fields and not description %}
+{%- for line in class_body_lines %}
+    {{ line }}
+{%- endfor %}
+{%- if not fields and not description and not config and not class_body_lines %}
     pass
 {%- else %}
     {%- set field = fields[0] %}

src/datamodel_code_generator/parser/base.py

@@ -1564,6 +1564,32 @@ def __replace_unique_list_to_set(self, models: list[DataModel]) -> None:
                         model_field.default = converted_default
                     model_field.replace_data_type(set_data_type)
 
+    @classmethod
+    def __collect_set_item_references(cls, models: list[DataModel]) -> set[str]:
+        """Collect reference paths of all types used as set/frozenset items."""
+        references: set[str] = set()
+        for model in models:
+            for field in model.fields:
+                for data_type in field.data_type.all_data_types:
+                    if data_type.is_set or data_type.is_frozen_set:
+                        for item_type in data_type.data_types:
+                            references.update(
+                                nested.reference.path for nested in item_type.all_data_types if nested.reference
+                            )
+        return references
+
+    @classmethod
+    def __mark_set_item_models_hashable(cls, models: list[DataModel]) -> None:
+        """Mark models used as set/frozenset items with hash flag for __hash__ generation."""
+        set_item_references = cls.__collect_set_item_references(models)
+
+        for model in models:
+            if model.reference.path in set_item_references:
+                if isinstance(model, Enum):
+                    continue
+                class_body_lines = model.extra_template_data.setdefault("class_body_lines", [])
+                class_body_lines.append("__hash__ = object.__hash__")
+
     @classmethod
     def __set_reference_default_value_to_field(cls, models: list[DataModel]) -> None:
         for model in models:
@@ -2981,6 +3007,8 @@ def _finalize_modules(
         module_to_import: dict[ModulePath, Imports],
     ) -> None:
         """Finalize module processing: apply generic base class and remove unused imports."""
+        all_models = [model for ctx in contexts for model in ctx.models]
+        self.__mark_set_item_models_hashable(all_models)
         self.__apply_generic_base_class(contexts)
 
         for ctx in contexts:

tests/data/expected/main/jsonschema/unique_items_enum_set.py

@@ -0,0 +1,25 @@
+# generated by datamodel-codegen:
+#   filename:  unique_items_enum_set.json
+#   timestamp: 2019-07-26T00:00:00+00:00
+
+from __future__ import annotations
+
+from enum import Enum
+
+from pydantic import BaseModel
+
+
+class Status(Enum):
+    active = 'active'
+    inactive = 'inactive'
+    pending = 'pending'
+
+
+class Item(BaseModel):
+    __hash__ = object.__hash__
+    name: str | None = None
+
+
+class Container(BaseModel):
+    statuses: set[Status] | None = None
+    items: set[Item] | None = None

tests/data/expected/main/openapi/with_field_constraints_pydantic_v2_use_generic_container_types_set.py

@@ -10,6 +10,7 @@
 
 
 class Pet(BaseModel):
+    __hash__ = object.__hash__
     id: int = Field(..., ge=0, le=9223372036854775807)
     name: str = Field(..., max_length=256)
     tag: str | None = Field(None, max_length=64)

tests/data/expected/main/openapi/with_field_constraints_pydantic_v2_use_standard_collections_set.py

@@ -8,6 +8,7 @@
 
 
 class Pet(BaseModel):
+    __hash__ = object.__hash__
     id: int = Field(..., ge=0, le=9223372036854775807)
     name: str = Field(..., max_length=256)
     tag: str | None = Field(None, max_length=64)

tests/data/jsonschema/unique_items_enum_set.json

@@ -0,0 +1,29 @@
+{
+    "$schema": "http://json-schema.org/draft-07/schema#",
+    "title": "Container",
+    "type": "object",
+    "definitions": {
+        "Status": {
+            "type": "string",
+            "enum": ["active", "inactive", "pending"]
+        },
+        "Item": {
+            "type": "object",
+            "properties": {
+                "name": {"type": "string"}
+            }
+        }
+    },
+    "properties": {
+        "statuses": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {"$ref": "#/definitions/Status"}
+        },
+        "items": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {"$ref": "#/definitions/Item"}
+        }
+    }
+}

tests/main/jsonschema/test_main_jsonschema.py

@@ -7957,3 +7957,21 @@ def test_validators_requires_pydantic_v2(output_file: Path, tmp_path: Path, caps
         capsys=capsys,
         expected_stderr_contains="--validators option requires Pydantic v2",
     )
+
+
+@PYDANTIC_V2_SKIP
+def test_unique_items_enum_set(output_file: Path) -> None:
+    """Test set with enum items does not add __hash__ to enum (already hashable)."""
+    run_main_and_assert(
+        input_path=JSON_SCHEMA_DATA_PATH / "unique_items_enum_set.json",
+        output_path=output_file,
+        input_file_type="jsonschema",
+        assert_func=assert_file_content,
+        expected_file="unique_items_enum_set.py",
+        extra_args=[
+            "--output-model-type",
+            "pydantic_v2.BaseModel",
+            "--use-unique-items-as-set",
+            "--use-standard-collections",
+        ],
+    )