As mentioned in the original OClosure commit, OClosures (ab)use the
bytecode's docstring slot to hold the OClosure's type. This currently
prevents OClosures from having their own docstring.
Introduce a new generic function `function-documentation` to fetch the
docstring of a function, which can then be implemented in various
different ways depending on the OClosure's type.
* lisp/simple.el (function-documentation): New generic function.
(bad-package-check): Strength-reduce `eval` to `symbol-value`.
* src/doc.c (Fdocumentation): Use it.
* lisp/emacs-lisp/oclosure.el (oclosure--accessor-docstring): New function.
* test/lisp/emacs-lisp/oclosure-tests.el (oclosure-test):
Add test for accessor's docstrings.
* lisp/emacs-lisp/oclosure.el (oclosure--slot-index)
(oclosure--slot-value, oclosure--set-slot-value): New functions.
* lisp/emacs-lisp/eieio-core.el (eieio-oset, eieio-oref):
Consolidate the type test. Use `oclosure--(set-)slot-value`.
(eieio--validate-slot-value, eieio--validate-class-slot-value):
Don't presume `class` is an EIEIO class.
(eieio--class): Fix bogus `:type` info.
(eieio--object-class): Simplify.
(eieio--known-slot-name-p): New function.
(eieio-oref, eieio-oref-default, eieio-oset-default): Use it.
* test/lisp/emacs-lisp/oclosure-tests.el: Require `eieio`.
(oclosure-test): Make `name` field mutable.
(oclosure-test-slot-value): New test.
Since string-lessp has its own byte-op, using it is much faster than
calling string-greaterp even with the need to bind a temporary
variable.
* lisp/emacs-lisp/byte-opt.el (byte-optimize-string-greaterp): New.
(string-greaterp, string>): Set byte-optimizer.
* lisp/emacs-lisp/easy-mmode.el (easy-mmode--arg-docstring): Allow
saying whether it's a global minor mode or not.
(easy-mmode--mode-docstring): Use it.
(define-minor-mode): Pass in the data.
* lisp/subr.el (ensure-empty-lines, string-lines): Moved from
subr-x so that they can be used in early bootstrap files.
* lisp/emacs-lisp/easy-mmode.el (easy-mmode--mode-docstring):
Don't use string-empty-p because of bootstrap issues.
* lisp/emacs-lisp/oclosure.el (oclosure--class): Add slot `allparents`.
(oclosure--class-make): Add corresponding arg `allparents`.
(oclosure, oclosure--build-class): Pass the new arg to the constructor.
(oclosure--define): Make the predicate function understand subtyping.
* lisp/emacs-lisp/cl-preloaded.el (cl--class-allparents): Move from
`cl-generic.el`.
* lisp/emacs-lisp/cl-generic.el (cl--generic-class-parents): Move to
`cl-preloaded.el` and rename to `cl--class-allparents`.
Adjust all callers.
(cl--generic-oclosure-tag, cl-generic--oclosure-specializers): New functions.
(cl-generic-generalizers) <oclosure-struct>: New generalizer.
* test/lisp/emacs-lisp/oclosure-tests.el (oclosure-test-gen):
New generic function.
(oclosure-test): Add test for dispatch on oclosure types.
* lisp/emacs-lisp/cl-generic.el (cl--generic-describe): Include
the function name in the implementations (bug#54628). This
clarifies what we're talking about here, and avoids getting
(function ...) translated into #'...
Rewrite the handling of `cl-no-next-method` to get rid of the hideous
hack used in `cl--generic-isnot-nnm-p` and also to try and move
some of the cost to the construction of the effective method rather
than its invocation. This speeds up method calls measurably when
there's a `cl-call-next-method` in the body.
* lisp/loadup.el ("emacs-lisp/oclosure"): Load.
* lisp/emacs-lisp/oclosure.el (oclosure-define): Remove workaround now
that we're preloaded.
* lisp/emacs-lisp/cl-generic.el (cl--generic-method): Rename `uses-cnm`
to `call-con` to reflect it's not a boolean any more.
(cl-defmethod): Adjust to the new name and new values.
(cl-generic-define-method): Adjust to the new name.
(cl--generic-lambda): Use the new `curried` calling convention.
(cl--generic-no-next-method-function): Delete function.
(cl--generic-nnm): New type.
(cl-generic-call-method): Rewrite to support the various
calling conventions.
(cl--generic-nnm-sample, cl--generic-cnm-sample): Delete consts.
(cl--generic-isnot-nnm-p): Rewrite using `oclosure-type`.
(cl--generic-method-info): Add support for new calling convention.
We here just add the new type. It is not fully self-contained.
It requires cooperation from `cconv.el` on the one hand, and it
hijacks the docstring info to hold the type of OClosure objects.
This does imply that OClosures can't have docstrings, tho this
limitation will be lifted in subsequent patches.
* lisp/emacs-lisp/oclosure.el: New file.
* test/lisp/emacs-lisp/oclosure-tests.el: New file.
* doc/lispref/functions.texi (OClosures): New section.
* src/eval.c (Ffunction): Accept symbols instead of strings for docstrings.
* src/doc.c (store_function_docstring): Avoid overwriting an OClosure type.
* lisp/emacs-lisp/cconv.el (cconv--convert-function): Tweak ordering of
captured variables.
(cconv-convert): Add case for `oclosure--fix-type`.
Co-authored-by: João Távora <joaotavora@gmail.com>
* lisp/emacs-lisp/eldoc.el (eldoc--echo-area-substring,
eldoc-display-in-echo-area): Take invisible text into consideration
when counting lines to crop an echo-area message.
(Version): Bump.
* lisp/emacs-lisp/comp.el (comp-func): Add a command-modes slot.
(comp-spill-lap-function, comp-intern-func-in-ctxt): Fill it.
(comp-emit-for-top-level, comp-emit-lambda-for-top-level): Use it.
* src/alloc.c (mark_object): Mark the command_modes slot.
* src/comp.c (make_subr): Add a command_modes parameter.
(Fcomp__register_lambda): Use it.
(Fcomp__register_subr): Ditto.
* src/data.c (Fcommand_modes): Output the command_modes data for subrs
(bug#54437).
* src/lisp.h (GCALIGNED_STRUCT): Add a command_modes slot.
* src/pdumper.c (dump_subr): Update hash.
(dump_subr): Dump the command_modes slot.
The dependencies between `advice`, cl-generic`, `bytecomp`, `cl-lib`,
`simple`, `help`, ... were becoming unmanageable.
Break the reliance on `advice` (which includes making sure the
compiler is not needed during the early bootstrap).
* lisp/simple.el (pre-redisplay-function): Set without using `add-function`.
* lisp/loadup.el (advice, simple): Move to after `cl-generic`.
* lisp/help.el (command-error-function): Set without using `add-function`.
(help-command-error-confusable-suggestions): Explicitly call
`command-error-default-function` instead.
* lisp/emacs-lisp/cl-macs.el (pcase--mutually-exclusive-p): Don't
optimize during early-bootstrap.
* lisp/emacs-lisp/cl-generic.el (cl--generic-lambda): Tiny simplification.
(cl-defmethod): Label the obsolescence warning as it should.
(cl--generic-compiler): New variable.
(cl--generic-get-dispatcher): Use it.
(cl--generic-prefill-dispatchers): Make freshly made dispatchers.
This fixes bug #54079.
* lisp/emacs-lisp/bytecomp.el (byte-compile-initial-macro-environment): Change
the position of 'byte-run-strip-symbol-positions' in the eval-when-compile
entry. Add a call to `byte-run-strip-symbol-positions' in the
eval-and-compile entry.
For example, 'xref-item' uses the list form of ':noinline'.
* lisp/emacs-lisp/cl-macs.el (cl-defstruct): Accept ':noinline' and
':named' with an argument.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-defstruct/edebug): New
unit test.
* lisp/emacs-lisp/cl-macs.el (cl-macro-list): Add the &whole and the
dotted tail cases from `cl-macro-list1`.
(cl-macro-list1, cl-define-compiler-macro-list): Delete edebug elem specs.
(cl-destructuring-bind, cl-define-compiler-macro): Use `cl-macro-list`
instead, now that it covers all the cases we need.
Also increment a loop counter. This should fix bug #54248.
* lisp/emacs-lisp/byte-run.el (byte-run--strip-vector/record): increment the
loop counter 'i' in the main loop.
* lisp/emacs-lisp/bytecomp.el (byte-compile-out): call
byte-run-strip-symbol-positions on operands which are one-element lists.
Print newlines appearing inside values (i.e. inside strings) as \n
rather than actual newlines so as to avoid messing up the main
structure of the trace buffer.
* lisp/emacs-lisp/trace.el (trace-entry-message)
(trace-exit-message, trace--read-args): Quote newlines
* lisp/emacs-lisp/cl-macs.el (cl-define-compiler-macro-list): New
Edebug element specification.
(cl-define-compiler-macro): Use it.
* test/lisp/emacs-lisp/cl-macs-tests.el
(cl-define-compiler-macro/edebug): New unit test.
This fixes bug #54098.
* lisp/emacs-lisp/byte-run.el (byte-run--strip-list)
(byte-run--strip-vector/record): New functions. These alter a list or
vector/record structure only where a symbol with position gets replaced by a
bare symbol.
(byte-run-strip-symbol-positions): Reformulate to use the two new functions.
(function-put): No longer strip positions from the second and third arguments.
* lisp/emacs-lisp/bytecomp.el (byte-compile-out): Remove the senseless
"stripping" of putative symbol positions from OPERAND, which is nil or a
number.
