One thing that reading _The Little Prover_ did for me was correct a sort of lazy preconception I had that you cannot “reason about” code in dynamically typed la

One thing that reading _The Little Prover_ did for me was correct a sort of lazy preconception I had that you cannot “reason about” code in dynamically typed languages. And it sort of crystallized my experience of large dynamically typed codebases not being as bad as one might expect. The patterns in OP are one case here where the boolean expressions are trivially analyzable and there are other classes of such analysis: e.g. it’s possible to use the condition of an id statement to detect useless expressions in both the then and else clauses.