* libguile/ports.c (scm_i_set_port_encoding_x): Always initialize
PT->encoding to something non-NULL. This fixes callers of
`scm_encoding_error' such that they always pass a non-NULL encoding
name. Reported by Matei Conovici.
This is so that compiling the same code on environments with different
locale settings yields the same result.
* module/system/base/compile.scm (compile-file): When ENC if #f, default
to "UTF-8" instead of `(fluid-ref %default-port-encoding)'.
* doc/ref/api-evaluation.texi (Compilation): Document the default output
file name and default source file encoding for `compile-file' and
`guile-tools compile'.
* module/language/tree-il/fix-letrec.scm: Replace U+2019 with an ASCII
quote. The iconv implementation on FreeBSD would immediately notice
and bail out when reading it under a US-ASCII locale.
Thanks to Bruno Haible for his suggestions. See
<http://lists.gnu.org/archive/html/bug-libunistring/2010-09/msg00007.html>,
for details.
* libguile/ports.c (register_finalizer_for_port): Always register a
finalizer for PORT.
(finalize_port): Close ENTRY->input_cd and ENTRY->output_cd.
(scm_new_port_table_entry): Initialize the `input_cd' and `output_cd'
fields.
(utf8_to_codepoint): New function.
(get_codepoint): Rewrite to use `iconv' instead of libunistring.
(scm_i_set_port_encoding_x): Initialize the `input_cd' and `output_cd'
fields.
(update_port_lf): Move upward. Use `switch' instead of `if's.
* libguile/ports.h (scm_t_port)[input_cd, output_cd]: New fields.
* libguile/print.c (codepoint_to_utf8, display_string): New functions.
(display_character): Use `display_string'.
(write_combining_character): Likewise.
(iprin1): Use `display_string' instead of `scm_lfwrite_str', and
`display_character' instead of `scm_putc'.
(write_character): Likewise.
(write_character_escaped): New function.
* test-suite/tests/encoding-escapes.test ("display output
escapes")["Rashomon"]: Use lower-case escapes.
["fake escape"]: New test.
* libguile/numbers.c (scm_is_integer): Infinities are not integers, per
the R6RS.
(scm_even_p, scm_odd_p): Passing an infinity to even? or odd? is an
error.
* test-suite/tests/numbers.test ("integer?"): Adapt test.
("expt"): Add tests for +inf.0 and -inf.0 exponents.
* NEWS: Add NEWS entries.
* configure.ac: Check for GC_call_with_gc_active.
* libguile/threads.h (scm_i_thread): Remove "top", as it's not used.
* libguile/threads.c (with_gc_inactive, with_gc_active): Define shims to
GC_do_blocking and GC_call_with_gc_active.
(scm_i_init_thread_for_guile): Don't do thread base adjustment here,
do it in scm_i_with_guile_and_parent. The previous logic would never
be run.
(scm_i_with_guile_and_parent): If we enter Guile mode, leave it too.
Take care of adjusting the thread stack base here too. Also, call
with_gc_active.
(scm_without_guile): Refactor.
* libguile/private-options.h (SCM_HUNGRY_EOL_ESCAPES_P): New private
option.
* libguile/read.c: Define SCM_HUNGRY_EOL_ESCAPES_P, defaulting to #f.
(skip_intraline_whitespace): New helper.
(scm_read_string): If SCM_HUNGRY_EOL_ESCAPES_P,
skip_intraline_whitespace after an escaped EOL.
* test-suite/tests/reader.test ("read-options"): Add test.
* libguile/numbers.c (scm_expt): Fix bug that caused expt to throw an
exception whenever the base was exact and the exponent was an
inexact integer, e.g. (expt 5 6.0).
(scm_expt): Fix bug that caused expt to introduce spurious imaginary
parts in the result when the base was an inexact negative real and
the exponent was an integer, e.g. (expt -1.0 2)
(scm_integer_expt, scm_expt): Change behavior of (integer-expt 0 -1),
and therefore also (expt 0 -1), to return NaN, per R6RS (actually,
R6RS says we should throw an exception or return an "unspecified
number object", but for now we use NaN). Formerly we returned 0, per
R5RS. R5RS claims that 0^x=0 for all non-zero x, but that's
mathematically incorrect, and probably an oversight.
(scm_integer_expt): Consistently throw a wrong-argument-type exception
when the exponent is inexact. Formerly, it didn't always check this
if the base was 0, 1, or -1.
* test-suite/tests/numbers.test ("integer-expt", "expt"): Add tests.
Thanks to Mark Harig for pointing this out and suggesting the text.
* doc/ref/scheme-scripts.texi (Invoking Guile): Mention --autocompile
and --no-autocompile.
Thanks to Mark Harig for pointing these out.
* doc/guile.1: Remove mention of the tutorial, since it no longer
exists (as a separate document). Document the -q option.
* THANKS: Add Mark Harig.
* doc/ref/goops.texi (The Metaobject Protocol): Bring forward to
before `Class Options', and add intro text to explain why. All of
the sections now remaining are ones where the MOP is more likely to
be relevant than not.
* doc/ref/goops.texi (GOOPS Object Miscellany): New section, combining
previous `Object Comparison', `Cloning Objects' and `Write and
Display'. Replace `Object Comparison' text with something that
makes sense. Add a snippet to the end of the write/display section.
* doc/ref/goops.texi (Introspection): Make a bit snappier.
(Classes): Add intro. Remove "the" and "metaobject" everywhere.
Remove `class-environment', since I don't know what it's useful for.
(Instances): Remove implementation notes.
(Built-in classes): Merge with `Instances'. Reorder text points.
(Generic Functions): Add intro. Remove "the" and "metaobject"
everywhere.
(Generic Function Methods): Merge with `Generic Functions'.
* doc/ref/goops.texi (Introspection): Move to after `Generic Functions
and Methods'. The idea is to have all of that sections that make
sense without needing to understand the MOP, before sections that
really depend on the MOP.
* doc/ref/goops.texi (Methods and Generic Functions): More explanation
of generic function invocation. Created (or moved) subsections here
to cover material on the following that was previously spread elsewhere:
accessors; extending primitives; merging generics; generic function
examples; handling invocation errors. Edited for clarity throughout.
(Generic functions and methods, Example): Nodes deleted, with their
material incorporated above.
(Class Precedence List): Edited to improve clarity.
(Sorting Methods): New subsection.
* module/web/request.scm (build-request): Make URI a positional
argument.
* module/web/response.scm: Remove extend-response.
(read-response): Fix a docstring.
* module/web/server.scm (extend-response): Include extend-response here,
but not exported.
* module/web/request.scm (read-request-body, write-request-body): Rename
from read-request-body/bytevector and
write-request-body/bytevector. Remove the /latin-1 variants, as they
were unused and a bad idea.
* module/web/response.scm (read-response-body, write-response-body):
Likewise.
* module/web/server/http.scm (http-read, http-write): Adapt to
request/response change.
* test-suite/tests/web-request.test:
* test-suite/tests/web-response.test: Update tests.
* module/web/http.scm (declare-symbol-list-header!): New helper.
("Connection"): Redefine as a header list.
("Allow", "Content-Encoding", "Accept-Ranges"): Redefine as symbol
lists.
* test-suite/tests/web-http.test:
* test-suite/tests/web-response.test: Adapt tests.
* module/web/http.scm (parse-media-type): Parse media types as symbols.
(parse-key-value-list, parse-param-component, parse-param-list):
Change kons to val-parser. Always parse keys as symbols, and always
either cons, if there is a val, or just have the key, if there is no
val. Easier to explain and just as correct.
(declare-param-list-header!, declare-key-value-list-header!): Adapt to
key-list and param-list kons change.
("Cache-Control", "Pragma", "Transfer-Encoding", "Accept", "Expect")
("TE"): Likewise, adapt.
("Content-Type"): Param keys are symbols.
* module/web/http.scm: Change to not expose the header-decl objects,
instead exposing header-parse, header-validator, header-writer et al.
Explaining header decls in the manual was too complicated.
(string->header, header->string): New helpers.
(<header-decl>): Remove the `sym' field.
(declare-header!): Adapt to header-decl change, and use
string->header.
(known-header?, header-parser, header-validator, header-writer): New
procedures.
Adapt to use the new procedures internally.
* module/web/http.scm (declare-header!): No need to specify `sym', as it
can be derived from `name'. Change to take parser, validator, and
writer as positional arguments, and multiple? as a keyword.
(parse-header): Change to take the header as a symbol already, and
just return the parsed value. All headers are symbols now, including
unknown headers. I feel OK doing this given that the symbol GC works
now.
(lookup-header-decl): Only look up headers by symbol.
(read-header): Adapt to parse-header change.
(valid-header?, write-header): Adapt to all headers being symbols.
(split-header-names, list-of-header-names?, write-header-list):
Represent all header names as symbols.
(declare-opaque-header!, declare-date-header!)
(declare-string-list-header!, declare-header-list-header!)
(declare-integer-header!, declare-uri-header!)
(declare-quality-list-header!, declare-param-list-header!)
(declare-key-value-list-header!, declare-entity-tag-list-header!):
Change to be functions instead of syntax, and no need to specify the
symbolic name. Update all header declarations accordingly.
* module/web/request.scm (validate-headers):
* module/web/response.scm (validate-headers): Adapt to all headers being
symbols.
* test-suite/tests/web-http.test (pass-if-parse, pass-if-any-error)
(pass-if-parse-error): Update for parse-header change.
("general headers"): Update header list examples to be all symbols.
* libguile/hashtab.c: Fix remaining vector hash table support.
* module/ice-9/session.scm (apropos-fold): Use a hash table instead of a
vector for the completions.
