The jsl quasiquoter converts Jaskell syntax into Haskell syntax.

A Jaskell expression is a sequence of commands. An command is one of the following:

• A Haskell identifier, optionally preceded by $, #, ?, !, or &. The identifier can be qualified or unqualified, and can be lowercase (function) or uppercase (data constructor). It cannot be an operator.

  • An identifier x translates to x if x is a function, and push x if x is a data constructor.
  • $x translates to liftS x. For example, $reverse will reverse the list on top of the stack.
  • #x translates to liftS2 x. For example, #gcd will pop the top two values and push their gcd.
  • ?x translates to pushM x. For example, ?getLine will execute getLine and push the result.
  • !x translates to popM x. For example, !putStrLn will print the string on top of the stack.
  • &x translates to liftSM x. For example, &readFile will pop a file path and push the contents of that file.
• A Haskell operator. This translates to liftS2 applied to the operator.
• A list of zero or more expressions, surrounded in square brackets and separated by commas. For example, [ 1, 3, 4 1 + ] pushes the list [ 1, 3, 5 ] onto the stack.
• A tuple of two expressions. For example, ( 0, "a" "b" ++ ) pushes the tuple ( 0, "ab" ) onto the stack.
• An empty tuple. This translates to push ().
• An integer, floating-point, character, or string literal. For example, 5 translates to push 5.
• A (potentially empty) expression surrounded in curly brackets. This pushes the translation of the expression onto the stack, allowing for higher-order programming.

Jaskell programs can be preceded by zero or more definitions, each of the form DEF name = expr ;. Definitions can be recursive and even mutually recursive.