Fix a typo in close_functions.

[zhenya1007]

The change fixes what is pretty obviously a typo in close_functions: the uncurried_functions list is a list of (id, params, body, fundesc) tuples, rather than a list of (id, params, arity, body) tuples, as can be seen from the form that creates it as well as from several places it's used.

The typo isn't a bug since the variables arity and body are never actually referenced in the body of the function, but it does (negatively) affect readability.

[samoht]

If they are not used, what about using _ instead for both ?

[thelema]

Good to have names as documentation, but since unused, how about _arity and _body?

[zhenya1007]

I have considered that. The rest of the code in close_functions explicitly names/binds each element in the (id, params, body, fundesc) tuple when referencing the uncurried_defs list regardless of whether the variables are actually used or not: take a look at how fenv_rec and clos_offsets are computed, and observe that params and body are named but not used in either computation. To be consistent with the rest of the code in the function, my preference would be to either keep every element of the tuple named (as my current patch is doing) or replace unused variables with underscore throughout the body of close_functions (which is a patch I can submit, if you like), but to avoid a mish-mash of two styles.

To Edgar: reinterpreting your suggestion as, "how about _body, and _fundesc" (since arity and body are misleading names): I haven't (yet) seen such usage anywhere else in OCaml compiler code, so I would prefer to either keep the variables named or replace them with just plain underscores (per Thomas's suggestion).

[gasche]

Patch applied in trunk, thanks!

I think you're right to make the patch non-invasive. The benefits of tweaking code for readability are in tension with the "if it ain't broke, don't fix it" principle, and changing one line for a very good reason is better than touching 20 lines with a good initial reason, and some marginal improvement. If/when we decide to activate unused variables warning on this part of the code, prefixing variables with _ will be easy.

[mcclurmc]

Should these pull requests be closed after a push to trunk?

[gasche]

I don't have the rights to close PR, so we should wait for someone else, e.g. @avsm or @samoht , to do so.