Re:Abstract Syntax Tree for Patching Code and Assessing Code...
by thelopa on Monday June 07, 2021 @04:40AM
(#27420371)
A naive AST traversal might also interpret some bit of fancy macro DSL as being “regular” code and have a spurious match. Any reliable whole-program source transformation in CL needs to leverage compiler-specific extensions (code walker, environment, etc.).
Re:Abstract Syntax Tree for Patching Code and Assessing Code...
by dasyatidprime on Monday June 07, 2021 @06:16AM
(#27420866)
I was assuming that at minimum the forms would be fully macroexpanded first (though macrolet is its own problem there). I also have the sense this might be easier in some Scheme dialects, but I don't know any of them well enough to say (thus that's part of my question).
Re:Abstract Syntax Tree for Patching Code and Assessing Code...
by foldr on Monday June 07, 2021 @06:56AM
(#27421101)
IIRC it's difficult to walk macroexpanded code in Common Lisp because you have to deal with lots of implementation-specific special forms.
(Edit) This article has some good info: https://zenodo.org/record/3254669#.YL3-WzYzb2w
Re:Abstract Syntax Tree for Patching Code and Assessing Code...
by thelopa on Monday June 07, 2021 @02:09PM
(#27425947)
Exactly! If you walk the expanded form you have implementation-specific junk to deal with. Also, it may not be possible to map a form post-expansion to a source location. If you walk the pre-expansion form you’re going to get bad results. There’s no winning here unless you have help from the implementation.
¹ For instance, I believe CL forbids lexically binding symbols from the CL package in ways that might conflict with the original definitions. But if you want to find uses of a particular library call from some other non-special package in a sound way, then you have to assume someone might flet that name, due to the way locals work. The underlying codewalker issue is roughly dual to the functioning of non-hygienic macros.