Bug with ; as concatenation operator in conditional If...The

[Brandon Parker]

I am not certain if this has been reported previously, but here is a bug with the ; as the concatenation operator in the conditional of an If...Then statement.

var$ = "[-]"

'This works...
'Notice the () around the right side of the (a = b)
'(i.e. the b term is in parentheses)
If (var$ = ("[";Trim$("- ");"]")) Then var$ = "-"

var$ = "[-]"
'This also works using the + concatenation operator
'with or without the spaces around the + operator
If (var$ = "[" + Trim$("- ") + "]") Then var$ = "-"

var$ = "[-]"
'But, this dose not
'No parentheses around the b term
If (var$ = "[";Trim$("- ");"]") Then var$ = "-"


{:0)

Brandon Parker

[tsh73]

I would say not a bug but priority difference (well, may be issue)
print "12"="1";"2"
produces 02 - "=" executes first, gives False (0), then it get concatenated by (semicolon)
print "12"="1"+"2"
produces 1 - first (+) concatenates, then (=) evaluates to True (1)

[Carl Gundel]

Oct 12, 2021 3:08:10 GMT -5 tsh73 said:

I would say not a bug but priority difference (well, may be issue)
print "12"="1";"2"
produces 02 - "=" executes first, gives False (0), then it get concatenated by (semicolon)
print "12"="1"+"2"
produces 1 - first (+) concatenates, then (=) evaluates to True (1)

I would say that it's a bug in the way that it doesn't present with a proper error or warning.  LB5 does produce a type mismatch error in the compile.  Liberty BASIC doesn't compile expressions the same way in PRINT statements as it does in other situations.

[tsh73]

produce a type mismatch error in the compile

we used to (semicolon) automatically convert stuff to text then necessary - so no type mismatch expected?

[Carl Gundel]

Oct 12, 2021 7:50:07 GMT -5 tsh73 said:

produce a type mismatch error in the compile

we used to (semicolon) automatically convert stuff to text then necessary - so no type mismatch expected?

It works properly in the PRINT statement where it was intended to work properly.    It works better in LB5 because it doesn't blow up with a meaningless error.

[tsh73]

(shame on me - I thought I understand the problem but I didn't actually RUN original code, so I didn't saw error message)