Modern PEG supports inversed class like `[^a-z]` that would get any
character not in the `a-z` range. This commit adds support for that and
also for a new `not-in-range` PEG pattern for scheme.
* module/ice-9/peg/codegen.scm (cg-not-in-range): New function.
* module/ice-9/peg/string-peg.scm: Add support for `[^...]`
* test-suite/tests/peg.test: Test it.
* doc/ref/api-peg.texi: Document accordingly.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
This commit adds support for PEG as described in:
<https://bford.info/pub/lang/peg.pdf>
It adds support for the missing features (comments, underscores in
identifiers and escaping) while keeping the extensions (dashes in
identifiers, < and <--).
The naming system tries to be as close as possible to the one proposed
in the paper.
* module/ice-9/peg/string-peg.scm: Rewrite PEG parser.
* test-suite/tests/peg.test: Fix import
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
This module replaces the method and define-method bindings with their
method* and define-method* counterparts, for use by users who prefer not
to use both kinds of syntactic forms.
* module/oop/goops/keyword-formals.scm: New module.
* am/bootstrap.am: Added
* doc/ref/goops.texi: Document this change.
* module/oop/goops.scm (compute-keyword-formal-ids): Renamed from
->keyword-formal-ids; modified to do work both on the list of formals
and the list of formal ids in the next-method call.
(compute-make-procedure): Use compute-keyword-formal-ids.
* module/oop/goops (compute-procedure, compute-make-procedure): Emit
lambda or lambda* as appropriate. This doesn't matter now since all
will boil down to lambda-case, but to be future-proof...
Also add some clarifying comments.
* module/oop/goops.scm (method-keyword-formals?): Remov method slot
keyword-formals? as well as exported procedure method-keyword-formals?
introduced in 765f1d49 by partially reverting that commit.
* module/oop/goops.scm: Export method* and define-method*.
(define-method): Extract definitions of helper procedures and place
them in an eval-when at top level.
(define-method*): Renamed from last commits define-method and modified
to invoke method*.
(define-method): New syntax.
(parse-keyword-formals): Renamed from parse-formals and modified to
give keyword methods a specialzers list with tail <top>.
(parse-formals): Re-introduce the code of previous parse-args.
(%compute-applicable-methods): Revert change of previous
commit. Giving keyword methods a specializer tail <top> naturally
makes original %compute-applicable-methods work also with keyword
methods (which kind of shows that we have made the correct choices).
(method*): Renamed from last commit's "method".
(method): New syntax.
* module/oop/goops.scm (keyword-formals?): New slot in <method>.
(method-keyword-formals?): New exported <method> getter.
(%compute-applicable-methods): Treat method as applicable if having
matched all specializers, still have further arguments and have
keyword-formals.
(%compute-applicable-methods): Remove unused local variable n.
(define-syntax method): Rename parse-args to parse-formals.
(parse-formals): Return formals, specializers and keyword-formals.
(compute-procedure): Make a lambda* with possibly keyword formals.
(->formal-ids): Renamed from ->proper and now returns formal-ids.
(->keyword-formal-ids): New procedure. Filter out formal ids from
a keyword formal specification.
(compute-make-procedure): Adapted for keyword formals. Needs
->formal-ids and ->keyword-formal-ids to compute the
real-next-method call.
(compute-procedures): Pass on keyword-formals.
(syntax method): Adapted for keyword formals.
* module/ice-9/psyntax.scm (define/override, define*/override): Use
instead of set! on globals.
($sc-dispatch): Renest. Will compile to the same thing as before.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/psyntax.scm (resolve-module*, resolve-variable): New
helpers.
(free-id=?, resolve-identifier): Use new helpers.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/psyntax.scm (expand-let, expand-letrec, ...): Name these
expanders, then install them. Allows for better code evolution and
decreases the indent.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/psyntax.scm (build-lexical-reference): No "type"
parameter. Adapt callers.
(valid-bound-ids?, distinct-bound-ids?, bound-id-member?): Use match.
(expand-sequence, expand-top-sequence): Use match. For
expand-top-sequence, ensure that both phases of expansion are run in
order; was the case before, but by accident. Don't accumulate results
in reverse.
(parse-when-list): Use match.
* module/ice-9/psyntax.scm (id-var-name): No need for `search` to return
the marks. Simplify to use scope instead of repeating, and use match.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/psyntax.scm (maybe-name-value): Return a fresh lambda
instead of mutating the given lambda.
(define-expansion-accessors): No need to define setters.
* module/ice-9/psyntax.scm (gen-unique): Instead of making a string with
an embedded hex counter, make a vector. A little less work than making
a string, and slightly smaller binaries.
(gen-label, gen-mark): Use gen-unique.
(resolve-identifier): Adapt case that recognizes labels denoting
lexicals to be less strict.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/psyntax.scm (expand-top-sequence): Remove needless
gen-label uses, and replace one use with gen-lexical (which is what is
needed).
* module/ice-9/psyntax-pp.scm: Regenerate.
The specification mandates a string, but with rationale suggesting symbols
would be a more natural fit.
> In some ways using symbols would be preferable. However, we want
> human-readable names, and standard Scheme does not provide a way to include
> spaces or mixed-case text in literal symbols.
Add support for symbols as an implementation extension and for backwards
compatibility with the reference implementation.
* module/srfi/srfi-64.scm (%cmp-group-name): New procedure.
(test-end): Use it.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
Reported-by: Daniel Llorens <lloda@sarc.name>
Fixes <https://bugs.gnu.org/30600>.
* module/system/repl/command.scm (define-meta-command): Flush all
remaining input after handling a read error.
* module/system/repl/common.scm (flush-all-input): New public procedure.
* module/system/repl/repl.scm: Remove local flush-all-input definition.
The bundled (reference) implementation was of somewhat mixed quality and
it failed to follow standard in multiple places. This commit replaces
it with a new one, written from scratch to follow the standard as close
as possible.
* module/srfi/srfi-64/testing.scm: Delete file.
* module/srfi/srfi-64.scm: Replace with new implementation.
* am/bootstrap.am (srfi/srfi-64.go): Remove extra dependencies.
(NOCOMP_SOURCES): Remove srfi/srfi-64/testing.scm.
* test-suite/tests/srfi-64-test.scm
("8.6.1. Simple (form 1) test-apply")
("8.6.2. Simple (form 2) test-apply"): Adjust tests to follow the
specification.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
* module/language/cps/specialize-numbers.scm (sigbits-ref): New helper.
(invert-graph*): New helper.
(compute-significant-bits): When visiting a term changes computed
needed-bits for one of its definitions, we need to revisit the variables
that contributed to its result (the uses), because they might need more
bits as well. Previously we were doing this by enqueueing predecessors
to the term, which worked if the uses were defined in predecessors, or
if all defining terms were already in the worklist, which is the case
without loops. But with loops, when revisiting a term, you could see
that it causes sigbits to change, enqueue its predecessors, but then the
predecessors don't change anything and the fixpoint stops before
reaching the definitions of the variables we need. So instead we
compute the use-def graph and enqueue defs directly.
