first post?

Steve: I presume you made clear the bit about minuses – it’s a hard-wired two digit limit, not two characters. You can have a minus sign (actually a hyphen) as well as two digits (as long as the minus sign comes first…)

I presume you made clear the bit about minuses

I have just put a digits warning in for now. It is not only a minus but also a plus that can be inserted.

It is not only a minus but also a plus that can be inserted.

True, dat. I have to admit to not having a clue why one would want to put a plus or a leading zero in there, whereas a minus is necessary; but nor do I have a clue why anyone would have put a limit on leading zeroes if it actually increased the code size.

why anyone would have put a limit on leading zeroes

It’s a microscopic optimisation: avoid a loop.

I was looking at Basic’s number parsing because I am Making Something™, but I’ve not done anything useful in many days. It’s just too hot. There goes my plan to retire to the Algarve.

It’s just too hot. There goes my plan to retire to the Algarve.

Cognitive dissonance? Or realization of implications?

I’ve retired to where I lived already. If it wasn’t for family connections I’d retire to Iceland I think.

a microscopic optimisation: avoid a loop

But makes the code bigger? And avoids a loop that will almost never be taken more than twice?

OK – I’ve not read the code, and I’m talking out of my posterior orifice.

Although this isn’t the exact code, reading a zero, one or two digit decimal is very titchy:

    LDRB   R0, [R1],#1
    SUB    R0, R0, #"0"
    CMP    R0, #10
    MOVCS  R0, #0
    LDRCCB R2, [R1],#1
    SUBCC  R2, R2, #"0"
    CMPCC  R2, #10
    ADDCC  R0, R0, R0, LSL#2
    ADDCC  R0, R2, R0, LSL#1
    SUBCS  R1, R1, #1

A loop is two fewer instructions, but more cycles because it has to repeat the loop for the terminator too:

    MOV    R0, #0
lp  LDRB   R2, [R1],#1
    SUB    R2, R2, #"0"
    CMP    R2, #10
    ADDCC  R0, R0, R0, LSL#2
    ADDCC  R0, R2, R0, LSL#1
    BCC    lp
    SUB    R1, R1, #1

So I wasn’t talking out of my posterior orifice then. That’s pretty much exactly what I’d imagined.

A microscopic optimization indeed – optimizing speed v. optimizing code size, microscopic either way.