* doc/ref/api-modules.texi (Creating Guile Modules): Document
#:re-export-and-replace.
* module/ice-9/boot-9.scm (module-replacements): New module field.
(make-module, make-autoload-interface): Initialize replacements to an
empty hash table.
(resolve-interface): Propagate replacement info when making custom
interfaces.
(define-module): Parse a #:re-export-and-replace keyword arg.
(define-module*): Handle #:re-export-and-replace.
(module-export!, module-re-export!): Add a keyword arg to indicate
whether to replace or not.
(module-replace!): Call module-export! with #:replace? #t.
(duplicate-handlers): Update replace duplicate handler to look for
replacement info on the interfaces.
* module/srfi/srfi-18.scm (srfi):
* module/srfi/srfi-34.scm (srfi): Update to #:re-export-and-replace
raise-continuable as raise.
* module/language/cps/types.scm (div-result-range): It is possible for a
max value to be less than a minimum. In this bug from zig:
(define (benchmark x)
(let loop ((count 0)
(sum 0))
(if (= count 10)
(exact->inexact (/ sum 10)))
(loop (+ count 1) x)))
Here the first iteration gets peeled, and thus the first "if" can't be
true, because "count" is zero. However on the true branch of the if,
range inference produces bogus ranges -- notably, the variable bound
to 10 is inferred to have a min of 10 and a max of 0. This is fine,
because it's unreachable; but that then infects the division, because
the same variable bound to 10 is used there, resulting in division by
zero.
* module/system/vm/assembler.scm (<arity>): Add new "has-closure?"
flag.
(begin-kw-arity, pack-arity-flags, write-arities): Write
"elided-closure?" flag into binary. A negative flag for compat
reasons.
* module/system/vm/debug.scm (elided-closure?, arity-has-closure?): Add
arity-has-closure? accessor.
* module/system/vm/frame.scm (frame-call-representation): Count from 0
for callees with elided closures.
* module/Makefile.am (SOURCES): Add new files.
* module/scheme/base.scm:
* module/scheme/case-lambda.scm:
* module/scheme/char.scm:
* module/scheme/complex.scm:
* module/scheme/cxr.scm:
* module/scheme/eval.scm:
* module/scheme/file.scm:
* module/scheme/inexact.scm:
* module/scheme/lazy.scm:
* module/scheme/load.scm:
* module/scheme/process-context.scm:
* module/scheme/r5rs.scm:
* module/scheme/read.scm:
* module/scheme/repl.scm:
* module/scheme/time.scm:
* module/scheme/write.scm: New files. Thanks to Göran Weinholt for
akku-scm and OKUMURA Yuki for yuni, off of which some of these files
were based.
* module/ice-9/boot-9.scm (exception-kind, exception-args): Export.
* module/ice-9/exceptions.scm (exception-kind, exception-args):
Re-export.
* module/srfi/srfi-18.scm: Rewrite exception support in terms of core
exceptions, not SRFI-34/35.
* test-suite/tests/srfi-18.test: Since Guile doesn't expose the current
exception handler as such, SRFI-18 captures it using delimited
continuations. This means that we can't compare the result
of (current-exception-handler) with the installed handler using eq?,
even though the procedures are indeed equivalent. So, instead test
handler behavior.
* libguile/exceptions.c:
* libguile/exceptions.h: New files.
* libguile.h: Add exceptions.h.
* libguile/Makefile.am (libguile_@GUILE_EFFECTIVE_VERSION@_la_SOURCES):
(DOT_X_FILES, DOT_DOC_FILES, modinclude_HEADERS): Add exceptions.c and
exceptions.h.
* libguile/init.c (scm_i_init_guile): Initialize exceptions.
* libguile/threads.c (scm_spawn_thread): Use new names for
scm_i_make_catch_handler and scm_c_make_thunk.
* libguile/throw.c: Rewrite to be implemented in terms of
with-exception-handler / raise-exception.
* libguile/throw.h: Use data types from exceptions.h. Move
scm_report_stack_overflow and scm_report_out_of_memory to
exceptions.[ch].
* module/ice-9/boot-9.scm (&error, &programming-error)
(&non-continuable, make-exception-from-throw, raise-exception)
(with-exception-handler): New top-level definitions.
(throw, catch, with-throw-handler): Rewrite in terms of
with-exception-handler and raise-exception.
: New top-level definitions.
* module/ice-9/exceptions.scm: Adapt to re-export &error,
&programming-error, &non-continuable, raise-exception, and
with-exception-handler from boot-9.
(make-quit-exception, guile-quit-exception-converter): New exception
converters.
(make-exception-from-throw): Override core binding.
* test-suite/tests/eval.test ("inner trim with prompt tag"): Adapt to
"with-exception-handler" being the procedure on the stack.
("outer trim with prompt tag"): Likewise.
* test-suite/tests/exceptions.test (throw-test): Use pass-if-equal.
* module/srfi/srfi-34.scm: Reimplement in terms of core exceptions, and
make "guard" actually re-raise continuations with the original "raise"
continuation.
* module/ice-9/exceptions.scm: New file, derived from (rnrs
conditions). Perhaps unadvisedly, in this file I've renamed a number
of the identifiers. I have never found that the R6RS identifiers made
sense to me. For now this is an internal module that R6RS and SRFI-35
will be based on.
* module/Makefile.am (SOURCES): Add the new file.
* module/rnrs/conditions.scm (rnrs): Export renamed identifiers
from (ice-9 exceptions).
* module/rnrs/conditions.scm: Use core record facilities to define the
base condition types, define-condition-type, and the standard
condition hierarchy.
(simple-condition?): Rename from condition-internal?.
* module/rnrs/exceptions.scm: Move `raise' definition here, out from the
procedural records layer.
(format-simple-condition): Reimplement in a simpler way, hopefully
producing the same output.
* module/rnrs/records/procedural.scm:
* module/rnrs/records/inspection.scm: Import the exceptions and
conditions modules, and use the normal raise function.
* module/ice-9/boot-9.scm (record-type-uid): New accessor.
(make-record-type): Record UID in record type properties.
* module/rnrs/conditions.scm (define-condition-type): Fix invalid
invocation of make-record-type.
* module/rnrs/records/inspection.scm: Rewrite to use core record
inspection facilities.
* module/rnrs/records/procedural.scm: Rewrite to use core
make-record-type. Incidentally the result is that instances of
derived R6RS record types are now flat instead of nested.
* test-suite/tests/r6rs-records-procedural.test
("make-record-type-descriptor"): Relax a couple condition type checks,
while we redo the exception system.
* module/ice-9/boot-9.scm (make-record-type): Allow (mutable NAME)
or (immutable NAME) as a field name, and record field mutability in a
bitfield.
(record-modifier): Throw an error if the field isn't mutable.
* test-suite/tests/records.test ("records"): Add tests.
* doc/ref/api-data.texi (Records): Update.
* module/srfi/srfi-35.scm: Import (ice-9 match), and remove now-unused
srfi-1 import.
(print-condition): Print more like records, as appears to be the
intention.
(&condition): Define using make-record-type. Adapt all callers.
Also, compound conditions are now a disjoint type, handled specially
by condition-ref, condition?, and so on.
* test-suite/tests/srfi-35.test (v3): Fix an error in which a
subcondition was initialized without initializers for all of its
fields.
* module/ice-9/boot-9.scm (make-record-type): Validate that the fields
are a unique list of symbols. Deprecate passing a string as a type
name.
* module/system/base/syntax.scm (define-record): Update to pass a symbol
as a type name.
* test-suite/tests/records.test (rtd-foo, rtd-fŏŏ, "records"): Adapt to
make record types with symbol names.
* module/ice-9/boot-9.scm (make-record-type): Don't allow subtyping of
final types.
(%record-type-error): Remove helper.
(record-accessor, record-modifier): Use computed record type
predicate, to allow for subtyping.
(define-record-type): Adapt to %record-type-error going away; these
types are final so no accessor adaptation is needed.
* test-suite/tests/records.test: Add tests.
* doc/ref/api-data.texi (Records): Update.
* module/ice-9/boot-9.scm (record-type-vtable): Add slots for "flags"
and a parent vector.
(record-type-name, record-type-fields): Move up in the file.
(record-type-constructor, record-type-flags, record-type-parents): New
accessors.
(make-record-type): Take #:final? and #:parent keyword arguments.
(record-constructor): Delegate to record-type-constructor.
(record-predicate): For non-final types --types that can be extended
by subtyping -- implement an O(1) type predicate.
(define-record-type): Initialize the new fields.
* module/srfi/srfi-9.scm (%define-record-type): Initialize flags and
parent fields.
* module/ice-9/channel.scm: Remove. This file has had a bug since
2.0 or so that prevented loading the module; I can only conclude that
it hasn't worked for years.
* module/Makefile.am (SOURCES): Remove ice-9/channel.scm.
* module/ice-9/peg/using-parsers.scm (prec): Use SRFI-9 to define PEG
record types.
(peg:start, peg:end, peg:string, peg:tree, peg:substring): Implement
in a more efficient way.
* module/ice-9/safe-r5rs.scm: Define local versions of `case' and `cond'
that assume aux syntax is unbound. If this doesn't work, we can
switch to exporting aux syntax.
* module/ice-9/top-repl.scm (top-repl): Don't add (ice-9 r5rs) to the
REPL environment.
* module/ice-9/boot-9.scm (else, =>, ..., _): New definitions. These
are specified by the r6rs and the r7rs.
* module/ice-9/sandbox.scm (core-bindings): Include the aux syntax
definitions.
* module/rnrs/base.scm:
* module/rnrs.scm: Re-export aux syntax.
Since the change in 2.2 noted in the NEWS as "Fix literal matching for
module-bound literals", defining `_' makes `syntax-rules' and `match'
fail to recognize `_' as the catch-all literal. This change adapts the
recommendations to current practice in 2.2, as users have had to adapt
to this change.
* doc/ref/api-i18n.texi (Gettext Support): Update documentation.
* module/language/tree-il/analyze.scm (proc-ref?, gettext?): G_ is the
conventional abbreviation, not _.
* test-suite/tests/tree-il.test: Adapt.
* module/ice-9/command-line.scm: Use G_ instead of _.
* module/language/tree-il/letrectify.scm (compute-private-toplevels):
New function; computes the subset of declarative bindings that are
private to a module. If the module exports a macro, all bindings are
public, as we have no way to know what binding might be exported.
(letrectify): Add #:seal-private-bindings? keyword arg. If true, avoid
making boxes for private definitions.
* module/language/tree-il/optimize.scm (optimize): Add
-Oseal-private-bindings, enabled at -O3.
* module/language/cps/cse.scm (compute-equivalent-subexpressions): When
CSE sees a definition like `(cons a b)', it will also record an
"auxiliary definition" for `(car x)', where x is the variable defined
by the cons, whereby calling `(car x)' can reduce to `a' if there is
no intervening effect that clobbers the definitions. However, when
the successor of the cons is a control-flow join, then any variables
defined there have multiple definitions. It's incorrect to add the
aux definition in that case.
* test-suite/tests/compiler.test ("cse auxiliary definitions"): New
test.
