Support for chained comparisons

[marcoeilers]

This PR adds support for chained comparisons as syntactic sugar.
I.e., e1 < e2 <= e3 > e4 can now be parsed and will be desugared to (e1 < e2) && (e2 <= e3) && (e3 > e4) .
The AST and ParseAST are unchanged.
Note that the in operator has the same behavior as the usual comparisons, so you can also write s1 in s2 in s3.
Parentheses can be used to force different parsing; e.g. (e1 < e2) <= e3 would not be seen as a chained comparison and would thus result in a type error (since there is a boolean value on the left hand side of a <= operator).

Additionally, this PR fixes the positions reported for other suffixed expressions. E.g., the position Viper reported for the expression e + e2 used to be the position of just the + e2 part (i.e., the suffix), without the first operand. Same for e.g. e.f.

Additionally, this PR adds a deprecation warning when the Rational type is used.

[JonasAlaif]

This will cause a merge conflict for my IDE changes where I fixed the position of suffixed stuff too 😢

[marcoeilers]

This will cause a merge conflict for my IDE changes where I fixed the position of suffixed stuff too cry

Ah crap :/

[JonasAlaif]

At least I've essentially already done a good review for that part and I can compare our two solutions :P

[JonasAlaif]

@marcoeilers looks good, I cleaned up some things up in the PR. If you're happy with my changes then we can merge