Function apply 

pub fn apply(
    op: Arc<dyn ExtensionOp>,
    inputs: &[&TracedTensor],
) -> Result<Vec<TracedTensor>>

Apply an extension op in the traced graph.

The op value is cloned into a StdTensorOp::Extension(Arc<dyn ExtensionOp>) carrier. The returned vector contains one TracedTensor per declared output slot of the extension. Output shapes are inferred via ExtensionOp::infer_output_meta using the input shape hints.

inputs.len() must equal op.input_count(), and each input’s shape_hint must be present (i.e. the extension must be used on tensors whose rank is known at graph-build time). For symbolic-shape composition, pass concrete tensors to crate::Runtime::run_compiled at evaluation time.

Examples

use std::sync::Arc;
use tenferro_runtime::extension::{apply, ExtensionOp};
use tenferro_runtime::{DType, SymDim, TracedTensor};
use tenferro_ops::ExtensionShapeContext;

let op: Arc<dyn ExtensionOp> = Arc::new(IdentityExt);
let a = TracedTensor::from_vec_col_major(vec![2], vec![1.0_f64, 2.0]).unwrap();
let outputs = apply(op, &[&a])?;
assert_eq!(outputs.len(), 1);

Errors

Returns Error::Validation with ValidationError::InvalidArgument when the extension receives the wrong number of traced inputs or produces an unknown output shape. Canonical metadata inference failures, including a returned metadata count that differs from ExtensionOp::output_count, are returned as Error::TensorRuntime containing the typed tensor validation source, while poisoned metadata state is retained as Error::RuntimeStateSource.