Fix YAML scientific notation parsing as float (#2913)

Author: koxudaxi

src/datamodel_code_generator/util.py

@@ -59,6 +59,9 @@ def is_pydantic_v2() -> bool:
 
 _YAML_1_2_BOOL_PATTERN = re.compile(r"^(?:true|false|True|False|TRUE|FALSE)$")
 _YAML_DEPRECATED_BOOL_VALUES = {"True", "False", "TRUE", "FALSE"}
+# Pattern for scientific notation without decimal point (e.g., 1e-5, 1E+10)
+# Standard YAML only matches floats with decimal points, missing patterns like "1e-5"
+_YAML_SCIENTIFIC_NOTATION_PATTERN = re.compile(r"^[-+]?[0-9][0-9_]*[eE][-+]?[0-9]+$")
 
 
 def _construct_yaml_bool_with_warning(loader: Any, node: Any) -> bool:
@@ -104,6 +107,13 @@ class CustomSafeLoader(_SafeLoader):  # type: ignore[valid-type,misc]
         ))
     CustomSafeLoader.yaml_constructors["tag:yaml.org,2002:bool"] = _construct_yaml_bool_with_warning
 
+    # Add scientific notation without decimal point (e.g., 1e-5) as float
+    for key in ["-", "+", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9"]:
+        CustomSafeLoader.yaml_implicit_resolvers.setdefault(key, []).append((
+            "tag:yaml.org,2002:float",
+            _YAML_SCIENTIFIC_NOTATION_PATTERN,
+        ))
+
     return CustomSafeLoader
 
 

tests/data/expected/main/yaml/scientific_notation.py

@@ -0,0 +1,14 @@
+# generated by datamodel-codegen:
+#   filename:  scientific_notation.yaml
+#   timestamp: 2019-07-26T00:00:00+00:00
+
+from __future__ import annotations
+
+from pydantic import BaseModel
+
+
+class Model(BaseModel):
+    exponential_default: float | None = 1e-05
+    positive_exp: float | None = 20000000000.0
+    negative_prefix: float | None = -30000.0
+    with_decimal: float | None = 1.5e-05

tests/data/yaml/scientific_notation.yaml

@@ -0,0 +1,15 @@
+$schema: http://json-schema.org/draft-07/schema#
+type: object
+properties:
+  exponential_default:
+    type: number
+    default: 1e-5
+  positive_exp:
+    type: number
+    default: 2E+10
+  negative_prefix:
+    type: number
+    default: -3e4
+  with_decimal:
+    type: number
+    default: 1.5e-5

tests/main/test_main_yaml.py

@@ -73,3 +73,19 @@ def test_main_yaml_deprecated_bool(output_file: Path) -> None:
             output_path=output_file,
             input_file_type="openapi",
         )
+
+
+def test_main_yaml_scientific_notation(output_file: Path) -> None:
+    """Test YAML file with scientific notation default values (issue #1955).
+
+    Scientific notation without decimal point (e.g., 1e-5) should be parsed as float,
+    not as string.
+    """
+    run_main_and_assert(
+        input_path=YAML_DATA_PATH / "scientific_notation.yaml",
+        output_path=output_file,
+        input_file_type="jsonschema",
+        extra_args=["--output-model-type", "pydantic_v2.BaseModel"],
+        assert_func=assert_file_content,
+        expected_file="yaml/scientific_notation.py",
+    )