The int8 operation converts exactly one input value to an 8-bit signed
integer. The operand may be a numeric literal, a numeric value already held in
a variable, or a string literal that represents a decimal or hexadecimal
integer. If the operand is a floating-point value, it must have no fractional
part. The result is stored as a true 8-bit signed integer ranging from -128 to
int8 raises an error. No other types are accepted. The operation never
changes its input; it produces a new value of type int8.The syntax of the int8 operation is:
[ $destination ] int8 value
The $destination token is optional; if it is omitted, the converted value is
assigned to _. Exactly one operand must follow int8. If you write int8
with no operand, it is treated as int8 _.
Converting an integer literal:
$small int8 42 # $small is int8 value 42
Converting a hexadecimal string:
$hex let "0x7F"
$byte int8 $hex # $byte is int8 value 127
Converting the current _ in place:
let -5
int8 # same as int8 _
Overflow raises an error:
int8 200 # error: overflow outside -128..127
A float with a fractional part also raises an error:
int8 3.14 # error: fractional part not allowed
int8 always returns a fresh int8 value that follows single-assignment rules;
it never mutates its operand and never produces more than one result.