You can easily hack the readtable in Lisp, and rewrite Emacs Lisp sexps as Common Lisp sexps.
Some interesting things to consider are file-local variables, buffer-local variables, dynamic-scope/lexical-scope, how to handle floating point (-0.0e+NaN in Emacs Lisp, custom floating point rouding modes/traps in SBCL for example), but this looks like a reasonable approach overall.
For example (just a draft, this might be incorrect):
USER> (defun make-local-variable (symbol)
(eval `(define-symbol-macro ,symbol (bvar (quote ,symbol)))))
MAKE-LOCAL-VARIABLE
USER> (make-local-variable 'foo)
FOO
USER> (macroexpand-all '(list foo))
(LIST (BVAR 'FOO))
T
T
BVAR could expand into code that calls FFI functions, for compatibility. Using an FFI approach would allow to progressively rewrite some parts of the runtime into CL.
Just to clarify, I know this is a huge work to undertake, but this comes from the assumption that we want to keep all existing Elisp files running identically.
The funny thing about lisp is that all things are easy, but yet, other eco systems offer more finished libraries.
As for compilation to Common Lisp, I don't know how feasible it would be. Apparently they tried something similar with Guile and it ran into problems.
You maybe have some more inside info, but back in 2016 I remember a lot of complaints and that guile was a liability for the much bigger Emacs project. Maybe even people threatening to retire as maintainers
This implies you have either to convert back and forward everything or you just can't use the native datastructures of the new programming language (making the whole operation often quite pointless).
I must confess that most of the comments in this thread seems to start from the assumption that people working on this in the last two+ decades are probably dumb, and this is sad.
For example, there's Hemlock from CMUCL, and its descendants built into Lispworks and Clozure Common Lisp.
Those implementations aren't likely to work well as a substitute for GNU Emacs. For one thing, there's a substantial ecosystem of software that depends on specific APIs and other characteristics of GNU Emacs. Writing code to bridge GNU Emacs APIs with those available in the Hemlock descendants would be a lot of work.
There are some other obstacles as well. CCL's implementation of Hemlock uses the Cocoa text architecture, so it's not portable to platforms other than macOS.
The Lispworks implementation is portable across Windows and numerous UNIXEN, but the Lispworks license will not allow delivery of a proper substitute for GNU Emacs (it forbids building an application that can be construed as a Lisp development system, and when I pressed them about exactly what limits that policy implies, they explicitly used Emacs as an example of an application that would be forbidden).
The original CMUCL Hemlock and its portable version are designed to work with CLX. It could probably be made to work in a modern X environment, but making it fit well into modern GUI environments and porting it to all the platforms GNU Emacs works on would be a huge amount of work.
I'm probably overlooking some other Emacsen, but I don't know of any off the top of my head for which the situation is any easier.
So far, nobody -that I know of- has tried to move GNU Emacs to Common Lisp. GNU Emacs is a well-known, well-used platform and moving to Common Lisp could very well be worth the upfront costs.
Wishful thinking aside, the GNU Emacs development community has hunkered down behind Emacs Lisp, so improving it will bring immediate and tangible benefits to every GNU Emacs user. I love Common Lisp and do use SBCL for a lot of personal projects, but I also love GNU Emacs and I will take any improvements there that I can get. Watching Emacs Lisp become a more viable general purpose language (improved performance means an expanded set of problems that it can now address) is a great development. Stefan Monnier on emacs-devel:
The main benefit of such a compiler is not to run existing Elisp
code faster (99% of existing Elisp code runs fast enough that the
user won't notice if it runs faster) but to make it practical to
write other Elisp code which would otherwise be too slow. But for
that to work well, you want the new compiler to be
available "everywhere", rather than just on some platforms. So it
will only start being useful when it works on GNU/Linux, macOS,
Windows, ARM, RISC-V, x86, amd64, MIPS, younameit (AFAIK
lbgccjit's CPU coverage is already good enough for that, so the
main barrier here is the OS support), and when it's not just an
option at compile-time but when it's included in all builds.Hemlock is not a reimplementation of GNU Emacs; it's derived from earlier Emacsen, just as GNU Emacs is. Hemlock in particular is based on ZWEI, and was in version 0.99 around the time that work started on the GNU Emacs project. It's a slightly older sibling of GNU Emacs, rather than a descendant of it.
GNU Emacs has undoubtedly become the de facto standard Emacs, but I don't see why that disqualifies other branches of the Emacs family tree from using the name that all of them inherit from the venerable TECO implementation.
My point was that Emacsen have already been built in Common Lisp, but that they do not serve as substitutes for GNU Emacs--a point with which you appear to agree.
Maybe we even agree on the reason why: because no Emacs implementation can substitute for GNU Emacs without supporting its ecosystem. The amount of work needed to do that would be enormous, whether the proposed substitute is written in Common Lisp or in something else.
As you said, the value lies primarily in the ecosystem. Common Lisp is by far the better language, but GNU Emacs is by far the most empowering/practical environment. I don't think enormous work would be needed, there are plenty of powerful shortcuts one could take (some ideas are in this thread) and Emacs Lisp can be made to run on top of Common Lisp but it's safe to say that for whatever reasons the will/intent is simply not there.
Emacs Lisp is what we have to go on with and it's nice to see that it is not stagnating.
Any kind of rewrite of Emacs is going to be a lot of work, I am not denying that. But if we assume there is a layer where everything works as it currently does (the C primitives), the Emacs Lisp defined on top of that needs not be defined in C, as far as I know.
Also, the last time I checked, there were a ton of legacy stuff in the configure script for different platforms.
Refactor the existing emacs to work headless and become the legacy-core, while also establishing channels to attatch any other core and frontend.
Additionally, when you free up your architecture and loose up the internal bindings between different areas, you sometimes can open chances for significant improvments which prior where not possible, thus still be faster with a "worse" solution.
Considering how old GNU Emacs is, how much historical problems it has and how often you hear that this and that can't be optimized for this and that reason, this seems like such a situation where loosing up would open faster roads long term.
There are small syntactic differences between Emacs lisp and Common Lisp so a lexer that handled those would suffice.
There are some small semantic differences but they could be handled by porting code. And the Emacs c code could be rewritten in lisp or ported into the Common Lisp implementation, depending on your taste.
> There are some small semantic differences but they could be handled by porting code. I suspect there are more than just small differences. And someone would have to do the work to port the code. I suspect that if packages that people use don't work, they will just avoid using the new version of Emacs, as opposed to abandoning the package.
It's not an impossible project, but it's not clear to me it's easier/better than the proposed solution.