bug: decoding error with constants

[AntoineFONDEUR]

bug: decoding error with constants

Time spent on this PR: 45 mins

Pull request type

Please check the type of change your PR introduces:

• Bugfix
• Feature
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no api changes)
• Build related changes
• Documentation content changes
• Other (please describe):

What is the current behavior?

Currently an instruction is interpreted as a felt constant as long as it is a NOp and has a regular PCUpdate.

Resolves #54

What is the new behavior?

There is no way to make a difference between a felt constant and an instruction.
So the new behavior is to try to decode all instructions and no matter if it fails or not, it will convert the instruction to a decimal number in the right column.

[ClementWalter]

I suggest that we instead try to use the size of an instruction. If a given instruction is size = 2, then the following encoded instruction is a constant. Otherwise, we try to decode as an instruction, and if an error is raised, again, consider it's a constant

[ClementWalter]

when possible, I prefer not to use "stateful" for loops, ie loops where an iteration depends on the previous ones.

Suggested change

	for (var i = 0; i < bytecode.length; i++) {
	programSheet
	.getRange(`A${i + 2}:B${i + 2}`)
	.setValues([
	[
	bytecode[i],
	isConstant
	? `=TO_SIGNED_INTEGER(A${i + 2})`
	: `=DECODE_INSTRUCTION(A${i + 2})`,
	],
	]);
	if (!isConstant) {
	isConstant = size(decodeInstruction(BigInt(bytecode[i]))) == 2;
	} else {
	isConstant = false;
	}
	}
	programSheet
	.getRange(`A2:B${bytecode.length + 1}`)
	.setValues(
	bytecode.map((instruction, i) => [
	instruction,
	i > 0 && size(decodeInstruction(BigInt(bytecode[i - 1]))) == 2 ? `=TO_SIGNED_INTEGER(A${i + 2})` : `=DECODE_INSTRUCTION(A${i + 2})`,
	]),
	);

[ClementWalter]

of course we compute two times decodeInstruction here so it's probably less efficient.

The merge-base changed after approval.