Revert "Allow CircuitOperations in cirq.optimized_for_sycamore" (quan…

…tumlib#4356)

Reverts quantumlib#4336.

Enabling CircuitOp serialization is causing errors in QCS.

cirq-google/cirq_google/optimizers/convert_to_sycamore_gates.py

@@ -97,10 +97,6 @@ def _is_native_sycamore_op(self, op: cirq.Operation) -> bool:
         ):
             return True
 
-        if gate is None and isinstance(op.untagged, cirq.CircuitOperation):
-            subcircuit = op.untagged.circuit
-            return all(self._is_native_sycamore_op(op) for op in subcircuit.all_operations())
-
         return False
 
     # TODO(#3388) Add summary line to docstring.
@@ -135,13 +131,12 @@ def on_stuck_raise(bad):
             keep=self._is_native_sycamore_op,
             intercepting_decomposer=self._convert_one,
             on_stuck_raise=None if self.ignore_failures else on_stuck_raise,
-            preserve_structure=True,  # keep CircuitOps but decompose their contents
         )
 
     def optimization_at(
         self, circuit: cirq.Circuit, index: int, op: cirq.Operation
     ) -> Optional[cirq.PointOptimizationSummary]:
-        if not isinstance(op, (cirq.GateOperation, cirq.CircuitOperation)):
+        if not isinstance(op, cirq.GateOperation):
             return None
 
         gate = op.gate

cirq-google/cirq_google/optimizers/convert_to_sycamore_gates_test.py

@@ -78,52 +78,13 @@ def test_single_qubit_gate_phased_xz():
     assert ops[0].gate == gate
 
 
-def test_circuit_operation_inspection():
-    q0, q1 = cirq.LineQubit.range(2)
-    gate = cirq.PhasedXZGate(axis_phase_exponent=0.2, x_exponent=0.3, z_exponent=0.4)
-    cop = cirq.CircuitOperation(cirq.FrozenCircuit(gate(q0)))
-    assert cgoc.ConvertToSycamoreGates()._is_native_sycamore_op(cop)
-
-    cop2 = cirq.CircuitOperation(cirq.FrozenCircuit(cirq.SWAP(q0, q1)))
-    assert not cgoc.ConvertToSycamoreGates()._is_native_sycamore_op(cop2)
-
-
-def test_circuit_operation_conversion():
-    q0, q1 = cirq.LineQubit.range(2)
-    subcircuit = cirq.FrozenCircuit(cirq.X(q0), cirq.SWAP(q0, q1))
-    circuit = cirq.Circuit(cirq.CircuitOperation(subcircuit))
-    converted_circuit = circuit.copy()
-    cgoc.ConvertToSycamoreGates().optimize_circuit(converted_circuit)
-    # Verify that the CircuitOperation was preserved.
-    ops = list(converted_circuit.all_operations())
-    assert isinstance(ops[0], cirq.CircuitOperation)
-    # Verify that the contents of the CircuitOperation were optimized.
-    reconverted_subcircuit = ops[0].circuit.unfreeze().copy()
-    cgoc.ConvertToSycamoreGates().optimize_circuit(reconverted_subcircuit)
-    assert ops[0].circuit == reconverted_subcircuit
-    cirq.testing.assert_circuits_with_terminal_measurements_are_equivalent(
-        circuit, converted_circuit, atol=1e-8
-    )
-
-
 def test_unsupported_gate():
     class UnknownGate(cirq.TwoQubitGate):
         pass
 
-    q0, q1 = cirq.LineQubit.range(2)
-    circuit = cirq.Circuit(UnknownGate()(q0, q1))
-    with pytest.raises(ValueError, match='Unrecognized gate: '):
-        cgoc.ConvertToSycamoreGates().optimize_circuit(circuit)
-
-
-def test_nested_unsupported_gate():
-    class UnknownGate(cirq.TwoQubitGate):
-        pass
-
     q0 = cirq.LineQubit(0)
     q1 = cirq.LineQubit(1)
-    subcircuit = cirq.FrozenCircuit(UnknownGate()(q0, q1))
-    circuit = cirq.Circuit(cirq.CircuitOperation(subcircuit))
+    circuit = cirq.Circuit(UnknownGate()(q0, q1))
     with pytest.raises(ValueError, match='Unrecognized gate: '):
         cgoc.ConvertToSycamoreGates().optimize_circuit(circuit)
 

docs/google/best_practices.md

@@ -49,39 +49,6 @@ my_circuit = cirq.Circuit()
 sycamore_circuit = cg.optimized_for_sycamore(my_circuit, new_device=cg.Sycamore, optimizer_type='sqrt_iswap')
 ```
 
-### Using CircuitOperation to reduce circuit size
-
-Particularly large batches (or sweeps) of circuits may encounter errors when
-sent to Quantum Engine due to an upper limit on request size. If the circuits
-in question have a repetitive structure, `cirq.CircuitOperation`s can be used
-to reduce the request size and avoid this limit.
-
-`optimized_for_sycamore` will preserve `CircuitOperation`s while optimizing
-their contents.
-
-```python
-import cirq
-import cirq_google as cg
-
-# Repeatedly apply Hadamard and measurement to 10 qubits.
-my_circuit = cirq.Circuit()
-qubits = cirq.GridQubit.rect(2, 5)
-for i in range(100):
-    my_circuit.append(cirq.H.on_each(*qubits))
-    for q in qubits:
-        my_circuit.append(cirq.measure(q, key=f'm{q}'))
-
-# The same circuit, but defined using CircuitOperations.
-# This is ~1000x smaller when serialized!
-sub_circuit = cirq.Circuit(cirq.H(qubits[0]), cirq.measure(qubits[0], key='m'))
-circuit_op = cirq.CircuitOperation(sub_circuit.freeze())
-circuit_op = circuit_op.with_qubits([q])
-circuit_op = circuit_op.with_measurement_key_mapping({'m': f'm{q}'})
-circuit_op = circuit_op.repeat(100)
-short_circuit = cirq.Circuit(circuit_op for q in qubits)
-```
-
-
 ## Running circuits faster
 
 The following sections give tips and tricks that allow you to improve your

docs/google/devices.md

@@ -223,15 +223,6 @@ For decay experiments and other applications, a WaitGate is provided
 that causes the device to idle for a specified amount of time.
 This can be accomplished by specifying a `cirq.WaitGate`.
 
-
-### Subcircuits
-
-Circuits with a repetitive structure can benefit from using
-`cirq.CircuitOperation` to specify "subcircuits" within the overall circuit.
-Using this type condenses the serialized representation of the circuit, which
-may help for circuits that would otherwise run into size limitations.
-
-
 ## Specific Device Layouts
 
 The following devices are provided as part of cirq and can help you get your