[BUG] Unusual error when running direct recursion

[MTakahashi-KWH]

Environment

• PyQASM version: 0.3.2
• Python version: 3.12.8
• Operating system: Windows

What happened?

Issue

The OpenQASM standard allows direct recursion by making gate/subroutine identifiers available within their own scope.

However, pyqasm fails when handling recursive calls, producing an unexpected parameter type error.

Example

OPENQASM 3;
include "stdgates.inc";
qubit[5] qb;
bit[5] cb;
const float my_angle = .1;
def test(qubit[1] a,int b){
  p(my_angle) a[0];
  if(b > 0){
    test(a[0],b-1);
  }
}
test(qb[2],3);
'''

Observed Behavior

The recursive call throws the following error:

ERROR :pyqasm: Error at line 9, column 2 in QASM file

 >>>>>> test(a[0], b - 1)
ValidationError: Expecting qubit argument for 'a'. Qubit register 'a' not found for function 'test'

• note that both test(a,b-1); and test(a[0],b-1); threw the same error when I was testing for semantic errors in my code

Expected Behavior

pyqasm should accept recursive calls in accordance with the OpenQASM standard, correctly resolving subroutine identifiers within their own scope.

Suggestions (Optional)

Id assume this is an error regarding scoping/aliasing within subroutine calls rather than a recursion issue, since error originates from: Qasm3SubroutineProcessor.process_quantum_arg(

[TheGupta2012]

Hi @MTakahashi-KWH , could you update your pyqasm version to 0.5.0 and test again? The above example seems to be working for me -

In [1]: import pyqasm

In [2]: pyqasm.__version__
Out[2]: '0.5.0'

In [3]: qasm = '''
   ...: OPENQASM 3;
   ...: include "stdgates.inc";
   ...: qubit[5] qb;
   ...: bit[5] cb;
   ...: const float my_angle = .1;
   ...: def test(qubit[1] a,int b){
   ...:   p(my_angle) a[0];
   ...:   if(b > 0){
   ...:     test(a[0],b-1);
   ...:   }
   ...: }
   ...: test(qb[2],3);
   ...: '''

In [4]: mod = pyqasm.loads(qasm)

In [5]: mod.validate()

In [6]: mod.unroll()

In [7]: print(mod)
OPENQASM 3.0;
include "stdgates.inc";
qubit[5] qb;
bit[5] cb;
h qb[2];
rx(0.1) qb[2];
h qb[2];
h qb[2];
rx(0.1) qb[2];
h qb[2];
h qb[2];
rx(0.1) qb[2];
h qb[2];
h qb[2];
rx(0.1) qb[2];
h qb[2];