In this language there are only expressions to compute integers or
strings. An expression can be type-checked using infer to determine
its type or the line of a type error. A well-formed expression can be
simplified using eval to produce its unique normal form which always
exists. Based on the first example from the book Practical Foundations
for Programming Languages.
T = Integer | String
E = integer
| string
| variable
| let variable = E in E
| E + E
| E - E
| E * E
| E ++ E
| len(E)
| decimal(E)
| (E)
parse : Text -> E || Error
infer : E -> T || Error
eval : E -> E
./calc
let x = 2 in x + x
<ctrl+D>
4 : Integer
If file prog.x contains the expression
let s1 = "fruit" in
let s2 = "cake" in
s1 ++ " " ++ s2
then you can feed it directly to the interpreter with
./calc prog.x
"fruit cake" : String
Not all syntactically valid expressions make sense.
./calc
let s1 = "fruit" in
let s2 = "cake" in
s1 ++ " " ++ len(s2)
<ctrl+D>
calc: line 3: right operand of ++ not a string
Output C source code.
./calc -c prox.x | gcc -x c -
./a.out
fruit cake
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent