// Create 1 dimensional tensor with values from `values`
TF_Tensor* Int32Tensor(const std::vector<int32_t>& values);

TF_Tensor* Int32Tensor(int32_t v);

TF_Tensor* DoubleTensor(double v);

TF_Tensor* FloatTensor(float v);

TF_Operation* Placeholder(TF_Graph* graph, TF_Status* s,
                          const char* name = "feed",
                          TF_DataType dtype = TF_INT32,

    )

    class FakeNumpyArray:
        def __init__(self):
            self.data = [1.0, 2.0, 3.0]

    FakeNumpyArrayPydantic = Annotated[
        FakeNumpyArray,
        WithJsonSchema(TypeAdapter(list[float]).json_schema()),
        PlainSerializer(lambda v: v.data),
    ]

    class MyModel(BaseModel):
        model_config = ConfigDict(arbitrary_types_allowed=True)
        custom_field: FakeNumpyArrayPydantic

	Timestamp1 *PrimaryTerm `parser:" @@ \",\" "`
	Timestamp2 *PrimaryTerm `parser:" @@ \")\" "`
}

// LitValue represents a literal value parsed from the sql
type LitValue struct {
	Float   *float64       `parser:"(  @Float"`
	Int     *float64       `parser:" | @Int"` // To avoid value out of range, use float64 instead
	String  *LiteralString `parser:" | @LitString"`
	Boolean *Boolean       `parser:" | @(\"TRUE\" | \"FALSE\")"`

* Datenkonvertierung
* Datenvalidierung
* Automatische Dokumentation

## Spezielle Typen und Validierungen { #special-types-and-validation }

Abgesehen von normalen einfachen Typen wie `str`, `int`, `float`, usw. können Sie komplexere einfache Typen verwenden, die von `str` erben.

   */
  ObjectCountHashMap(int capacity) {
    this(capacity, DEFAULT_LOAD_FACTOR);
  }

  ObjectCountHashMap(int expectedSize, float loadFactor) {
    init(expectedSize, loadFactor);
  }

  void init(int expectedSize, float loadFactor) {
    Preconditions.checkArgument(expectedSize >= 0, "Initial capacity must be non-negative");
    Preconditions.checkArgument(loadFactor > 0, "Illegal load factor");

from fastapi.testclient import TestClient
from pydantic import BaseModel

app = FastAPI()


class ItemGroup(BaseModel):
    items: list[tuple[str, str]]


class Coordinate(BaseModel):
    x: float
    y: float


@app.post("/model-with-tuple/")
def post_model_with_tuple(item_group: ItemGroup):
    return item_group


@app.post("/tuple-of-models/")

  @CanIgnoreReturnValue
  PrimitiveSink putInt(int i);

  /** Puts a long into this sink. */
  @CanIgnoreReturnValue
  PrimitiveSink putLong(long l);

  /** Puts a float into this sink. */
  @CanIgnoreReturnValue
  PrimitiveSink putFloat(float f);

  /** Puts a double into this sink. */
  @CanIgnoreReturnValue
  PrimitiveSink putDouble(double d);

  /** Puts a boolean into this sink. */
  @CanIgnoreReturnValue

def get_items(item_a: str, item_b: int, item_c: float, item_d: bool, item_e: bytes):

                            || actual == Short.class
                            || actual == Byte.class)) {
                return true;
            }
            if (formal == Float.TYPE
                    && (actual == Float.class
                            || actual == Long.class
                            || actual == Integer.class
                            || actual == Short.class

        "short" to "Short",
        "java.lang.Integer" to "Int",
        "int" to "Int",
        "java.lang.Long" to "Long",
        "long" to "Long",
        "java.lang.Float" to "Float",
        "float" to "Float",
        "java.lang.Double" to "Double",
        "double" to "Double"
    )


// See `org.gradle.kotlin.dsl.internal.sharedruntime.codegen.ApiTypeProviderKt.mappedTypeStrings`
private