type Cell struct {
	Value      string
	Attributes []color.Attribute
}

type Row struct {
	Cells []Cell
}

func NewCell(value string, attributes ...color.Attribute) Cell {
	attrs := append([]color.Attribute{}, attributes...)
	return Cell{value, attrs}
}

func (cell Cell) String() string {
	if len(cell.Attributes) == 0 {
		return cell.Value
	}
	s := color.New(cell.Attributes...)
	s.EnableColor()

                        const char* operation_name,
                        const TFE_OpAttrs* attributes, int expected_max_outputs,
                        TF_Status* status) const {
  std::vector<PartialTensorShape> expected_output_shapes(expected_max_outputs);
  StartExecute(context, inputs, operation_name, attributes,
               expected_max_outputs, *default_cancellation_manager_);

/// warning

Notice that `Field` is imported directly from `pydantic`, not from `fastapi` as are all the rest (`Query`, `Path`, `Body`, etc).

///

## Declare model attributes

You can then use `Field` with model attributes:

{* ../../docs_src/body_fields/tutorial001_an_py310.py hl[11:14] *}

`Field` works the same way as `Query`, `Path` and `Body`, it has all the same parameters, etc.

private
val zipEntryBaseTimestamp = LocalDateTime.of(1980, 2, 1, 0, 0, 0)
    .atZone(ZoneId.systemDefault()).toInstant().toEpochMilli()


object Attributes {
    val artifactType = Attribute.of("artifactType", String::class.java)
    val minified = Attribute.of("minified", Boolean::class.javaObjectType)
}


class JarAnalyzer(
    private val shadowPackage: String,
    private val keepPackages: Set<String>,

## Create your data model

Then you declare your data model as a class that inherits from `BaseModel`.

Use standard Python types for all the attributes:

{* ../../docs_src/body/tutorial001_py310.py hl[5:9] *}


The same as when declaring query parameters, when a model attribute has a default value, it is not required. Otherwise, it is required. Use `None` to make it just optional.

                }
                final NamedNodeMap attributes = imgNode.getAttributes();
                final String thumbnailUrl = getThumbnailSrc(responseData.getUrl(), attributes);
                final Integer height = getAttributeAsInteger(attributes, "height");
                final Integer width = getAttributeAsInteger(attributes, "width");
                if (!fessConfig.isThumbnailHtmlImageUrl(thumbnailUrl)) {

  // its corresponding inputs from the input ParallelTensors. Wraps the
  // resulting per-device and per-output TFE_TensorHandles into one
  // ParallelTensor per output of the original operation.
  //
  // Attributes are forwarded to executed operations unmodified.
  //
  // The returned optional has a value if and only if `status` evaluates to
  // TF_OK. Bad statuses are forwarded from underlying `TFE_Execute` calls, or

    const ParallelDevice& parallel_device,
    const std::string& parallel_device_name, TFE_Context* context,
    std::vector<MaybeParallelTensorUnowned> inputs, const char* operation_name,
    const TFE_OpAttrs* attributes, int expected_max_outputs,
    TF_Status* status) {
  absl::optional<std::vector<MaybeParallelTensorOwned>> result;
  // TODO(allenl): We should remove "TPU" from these op names at the very least,

'right' ? -1 : 1;\n    styles[sideA] = top * invertTop;\n    styles[sideB] = left * invertLeft;\n    styles.willChange = `${sideA}, ${sideB}`;\n  }\n\n  // Attributes\n  const attributes = {\n    'x-placement': data.placement,\n  };\n\n  // Update `data` attributes, styles and arrowStyles\n  data.attributes = { ...attributes, ...data.attributes };\n  data.styles = { ...styles, ...data.styles };\n  data.arrowStyles = { ...data.offsets.arrow, ...data.arrowStyles };\n\n  return data;\n}\n","import getOppositePlacement...

The same way, this database wouldn't receive a Pydantic model (an object with attributes), only a `dict`.

You can use `jsonable_encoder` for that.

It receives an object, like a Pydantic model, and returns a JSON compatible version:

{* ../../docs_src/encoder/tutorial001_py310.py hl[4,21] *}