No cons cell dot notation support

[mmiszczyk]

Usually, an s-expression is defined recursively as either an atom or a cons cell, with the cons cell consisting of a pair of symbolic expressions. Lists are defined as nested cons cells ending with a NIL/empty list atom. Cons cells are usually presented using dot notation: (1 . 2).

In Common Lisp:

* (equal '(1 . (2 . nil)) '(1 2))

T

In Hy:

=> (= '(1 . (2 . ())) '(1 2))
True

sexpr doesn't understand it, and treats dot as a floating point atom:

In [6]: parse('(1 . (2 . ())') == parse('(1 2)')
Out[6]: False

In [7]: parse('(1 . (2 . ())')
Out[7]: [SExpr([Atom int('1'), Atom float('.'), SExpr([Atom int('2'), Atom float('.'), SExpr([])])])]

While using dot notation for normal list is fairly obscure given the convenient syntactic sugar of (1 2), it's pretty useful for the alist data structure: ((key1 . value1) (key2 . value2)).