Two small bugfixes to argument/parameter parsing

rxi / fe

A tiny, embeddable language implemented in ANSI C

MIT License

1.3k stars 81 forks source link

Two small bugfixes to argument/parameter parsing #8

Closed prez closed 4 years ago

prez commented 4 years ago

I also have various patches to add miscellaneous features, like:

GT, GTE (very trivial)
variadic comparison operators and "is"
read (fe_string -> AST)
eval (again, very trivial)
load (for files)
a rather big patch to make the bindings in recursive functions lexical instead of dynamic, using the Y combinator (no gensym) - implemented as either a new Scheme-style syntax, or alternatively, changing how P_SET behaves when the second argument is a regular function declaration.

Let me know if you would accept or like to see any of those so I can make a PR.

rxi commented 4 years ago

The way functions behave concerning extra arguments is by design and not something I intend to change, the other change provides very little in functionality for the amount of code it takes, additionally I believe it breaks variadic functions. Regarding the other changes: the lack of those features was a decision rather than an oversight, see the response to #1

prez commented 4 years ago

>the other change provides very little in functionality for the amount of code it takes Agreed. It could be compressed further by moving the checks towards the end of eval, saving a couple of lines. >additionally I believe it breaks variadic functions. It does not, attention has been paid to that.

That said, I was anticipating this response - your code is incredibly well thought-out and those were the only "bugs" I was able to find. So it is not surprising that this decision was deliberate. Oh and thank you for fe. Reading and understanding it was a great learning opportunity.