[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[no subject]
|
From: |
Patrice Dumas |
|
Date: |
Tue, 9 Jan 2024 08:08:15 -0500 (EST) |
branch: master
commit d68472208fe0eb65fe6ec7b24db90d13914c7388
Author: Patrice Dumas <pertusus@free.fr>
AuthorDate: Mon Jan 8 11:12:47 2024 +0100
* tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
(index_entry_element_sort_string): free sort_string.
* tp/Texinfo/Convert/Text.pm (text_accents, brace_no_arg_command)
(_convert): use 'set_case' for case, that can be positive for upper
casing and negative for lower casing instead of 'sc', as in XS.
* tp/Texinfo/Structuring.pm (setup_sortable_index_entries): rename
$options as $convert_text_options and set code directly and pass
$convert_text_options directly instead of copying.
* tp/Texinfo/Convert/LaTeX.pm (_prepare_indices, _index_entry): set
index_formatting_text_options state once for all in _prepare_indices
instead of for each index entry. Set code directly in the state and
pass the state insetad of copying it in _index_entry.
* tp/Texinfo/XS/main/convert_to_text.c (destroy_text_options),
tp/Texinfo/XS/main/get_perl_info.c (copy_sv_options_for_convert_text):
if an XS converter is associated to the text optins, set it in
TEXT_OPTIONS and directly use its conf for other_converter_options. Do
not destroy other_converter_options if converter is set.
Set self_converter_options only if other_converter_options is not set.
* tp/Texinfo/XS/main/convert_to_text.c (text_accents_options)
(text_options): use static text_accents_options in text_options
instead of allocating and freeing.
---
ChangeLog | 29 ++++++++++++++++++++
tp/TODO | 1 +
tp/Texinfo/Convert/LaTeX.pm | 8 +++---
tp/Texinfo/Convert/Text.pm | 32 ++++++++++++----------
tp/Texinfo/Structuring.pm | 18 ++++++------
tp/Texinfo/XS/main/convert_to_text.c | 13 +++++----
tp/Texinfo/XS/main/convert_to_text.h | 4 ++-
tp/Texinfo/XS/main/get_perl_info.c | 30 ++++++++++++++------
.../XS/structuring_transfo/StructuringTransfoXS.xs | 5 +++-
9 files changed, 96 insertions(+), 44 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index 417ffd4736..eaab480d69 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,32 @@
+2024-01-08 Patrice Dumas <pertusus@free.fr>
+
+ * tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
+ (index_entry_element_sort_string): free sort_string.
+
+ * tp/Texinfo/Convert/Text.pm (text_accents, brace_no_arg_command)
+ (_convert): use 'set_case' for case, that can be positive for upper
+ casing and negative for lower casing instead of 'sc', as in XS.
+
+ * tp/Texinfo/Structuring.pm (setup_sortable_index_entries): rename
+ $options as $convert_text_options and set code directly and pass
+ $convert_text_options directly instead of copying.
+
+ * tp/Texinfo/Convert/LaTeX.pm (_prepare_indices, _index_entry): set
+ index_formatting_text_options state once for all in _prepare_indices
+ instead of for each index entry. Set code directly in the state and
+ pass the state insetad of copying it in _index_entry.
+
+ * tp/Texinfo/XS/main/convert_to_text.c (destroy_text_options),
+ tp/Texinfo/XS/main/get_perl_info.c (copy_sv_options_for_convert_text):
+ if an XS converter is associated to the text optins, set it in
+ TEXT_OPTIONS and directly use its conf for other_converter_options. Do
+ not destroy other_converter_options if converter is set.
+ Set self_converter_options only if other_converter_options is not set.
+
+ * tp/Texinfo/XS/main/convert_to_text.c (text_accents_options)
+ (text_options): use static text_accents_options in text_options
+ instead of allocating and freeing.
+
2024-01-07 Patrice Dumas <pertusus@free.fr>
* tp/Texinfo/Structuring.pm (setup_index_entry_keys_formatting):
diff --git a/tp/TODO b/tp/TODO
index 179ba95fe8..9a0f5f639f 100644
--- a/tp/TODO
+++ b/tp/TODO
@@ -657,6 +657,7 @@ tidy -qe *.html
profiling: package on debian:
libdevel-nytprof-perl
In doc:
+perl -d:NYTProf ../tp/texi2any.pl texinfo.texi --html
perl -d:NYTProf ../tp/texi2any.pl texinfo.texi
nytprofhtml
# firefox nytprof/index.html
diff --git a/tp/Texinfo/Convert/LaTeX.pm b/tp/Texinfo/Convert/LaTeX.pm
index e178016a0e..57fc65b6ab 100644
--- a/tp/Texinfo/Convert/LaTeX.pm
+++ b/tp/Texinfo/Convert/LaTeX.pm
@@ -940,6 +940,8 @@ sub _prepare_indices($)
my $index_names = $self->{'indices_information'};
if ($index_names) {
+ $self->{'index_formatting_text_options'}
+ = Texinfo::Structuring::setup_index_entry_keys_formatting($self);
my $merged_index_entries
= Texinfo::Structuring::merge_indices($index_names);
# select non empty indices
@@ -2429,8 +2431,6 @@ sub _index_entry($$)
if ($index_info->{'in_code'}) {
$in_code = 1;
}
- my $options
- = Texinfo::Structuring::setup_index_entry_keys_formatting($self);
my @subindex_commands = ($element);
my $current_element = $element;
while ($current_element->{'extra'}
@@ -2452,12 +2452,12 @@ sub _index_entry($$)
pop @{$self->{'formatting_context'}->[-1]->{'code'}};
}
# always setup a string to sort with as we may use commands
- my $convert_to_text_options = {%$options, 'code' => $in_code};
+ $self->{'index_formatting_text_options'}->{'code'} = $in_code;
my $sort_string
= Texinfo::Structuring::index_entry_element_sort_string(
$self, $entry,
$subindex_command,
- $convert_to_text_options, 1);
+ $self->{'index_formatting_text_options'}, 1);
my $result = '';
if (defined($sort_string)) {
# | in sort key breaks with hyperref
diff --git a/tp/Texinfo/Convert/Text.pm b/tp/Texinfo/Convert/Text.pm
index b98766b677..17082d1881 100644
--- a/tp/Texinfo/Convert/Text.pm
+++ b/tp/Texinfo/Convert/Text.pm
@@ -269,10 +269,9 @@ sub text_accents($;$$)
my ($contents_element, $stack)
= Texinfo::Convert::Utils::find_innermost_accent_contents($accent);
-
my $options = {};
$options->{'enabled_encoding'} = $encoding if (defined($encoding));
- $options->{'sc'} = $set_case if (defined($set_case));
+ $options->{'set_case'} = $set_case if (defined($set_case));
my $text = convert_to_text($contents_element, $options);
my $result = Texinfo::Convert::Unicode::encoded_accents(undef, $text,
@@ -321,13 +320,14 @@ sub brace_no_arg_command($;$)
}
}
if ($options and $Texinfo::Commands::letter_no_arg_commands{$command}) {
- if ($options->{'sc'}) {
- $result = uc($result);
- # NOTE does not seems to be set anywhere. Not a big deal to keep it,
- # but if it become used, it should be checked whether code elsewhere
- # should be changed to look for lc too. XS should be ok.
- } elsif ($options->{'lc'}) {
- $result = lc($result);
+ if ($options->{'set_case'}) {
+ if ($options->{'set_case'} > 0) {
+ $result = uc($result);
+ # NOTE does not seems to be decreased/set to negative anywhere, but
+ # should work ok if it is.
+ } else {
+ $result = lc($result);
+ }
}
}
return $result;
@@ -531,8 +531,12 @@ sub _convert($;$)
if ((! defined($element->{'type'})
or $element->{'type'} ne 'raw')
and !$options->{'_raw_state'}) {
- if ($options->{'sc'}) {
- $result = uc($result);
+ if ($options->{'set_case'}) {
+ if ($options->{'set_case'} > 0) {
+ $result = uc($result);
+ } else {
+ $result = lc($result);
+ }
}
if (!$options->{'_code_state'}) {
$result =~ s/``/"/g;
@@ -569,7 +573,7 @@ sub _convert($;$)
# commands with braces
} elsif ($accent_commands{$element->{'cmdname'}}) {
my $result = text_accents($element, $options->{'enabled_encoding'},
- $options->{'sc'});
+ $options->{'set_case'});
return $result;
} elsif ($element->{'cmdname'} eq 'image') {
$options->{'_code_state'}++;
@@ -638,7 +642,7 @@ sub _convert($;$)
$element->{'cmdname'}})))) {
my $result;
my $in_code;
- $options->{'sc'}++ if ($element->{'cmdname'} eq 'sc');
+ $options->{'set_case'}++ if ($element->{'cmdname'} eq 'sc');
if ($Texinfo::Commands::brace_code_commands{$element->{'cmdname'}}
or $Texinfo::Commands::math_commands{$element->{'cmdname'}}) {
$in_code = 1;
@@ -646,7 +650,7 @@ sub _convert($;$)
$options->{'_code_state'}++ if ($in_code);
$result = _convert($element->{'args'}->[0], $options);
$options->{'_code_state'}-- if ($in_code);
- $options->{'sc'}-- if ($element->{'cmdname'} eq 'sc');
+ $options->{'set_case'}-- if ($element->{'cmdname'} eq 'sc');
return $result;
# block commands
} elsif ($element->{'cmdname'} eq 'quotation'
diff --git a/tp/Texinfo/Structuring.pm b/tp/Texinfo/Structuring.pm
index abbb3fb4e8..bb92d614e7 100644
--- a/tp/Texinfo/Structuring.pm
+++ b/tp/Texinfo/Structuring.pm
@@ -114,11 +114,10 @@ our %XS_overrides = (
"Texinfo::Structuring::_XS_unsplit"
=> "Texinfo::StructTransfXS::unsplit",
- # TODO the XS override is slower than the perl function.
- # One possible reason could be that the text options are read
+ # TODO the text options are read
# from perl for each entry instead of once for each index.
- #"Texinfo::Structuring::index_entry_element_sort_string"
- # => "Texinfo::StructTransfXS::index_entry_element_sort_string",
+ "Texinfo::Structuring::index_entry_element_sort_string"
+ => "Texinfo::StructTransfXS::index_entry_element_sort_string",
# Not useful for HTML as functions, as the calling functions are
# already overriden
@@ -2471,19 +2470,20 @@ sub setup_sortable_index_entries($$$$$;$)
return $index_sortable_index_entries, $collator, $index_entries_sort_strings
unless ($index_entries);
- my $options = setup_index_entry_keys_formatting($customization_information);
+ my $convert_text_options
+ = setup_index_entry_keys_formatting($customization_information);
$index_sortable_index_entries = {};
foreach my $index_name (keys(%$index_entries)) {
my $sortable_index_entries = [];
foreach my $index_entry (@{$index_entries->{$index_name}}) {
my $entry_index_name = $index_entry->{'index_name'};
my $main_entry_element = $index_entry->{'entry_element'};
- my $convert_to_text_options = {%$options,
- 'code' => $indices_information->{$entry_index_name}->{'in_code'}};
+ $convert_text_options->{'code'}
+ = $indices_information->{$entry_index_name}->{'in_code'};
my ($entry_key, $sort_entry_key)
= _index_entry_element_sort_string_key($customization_information,
$index_entry, $main_entry_element,
- $convert_to_text_options, $entries_collator);
+ $convert_text_options, $entries_collator);
my @entry_keys;
my @sort_entry_keys;
if ($entry_key !~ /\S/) {
@@ -2511,7 +2511,7 @@ sub setup_sortable_index_entries($$$$$;$)
$subentry = $subentry->{'extra'}->{'subentry'};
my ($subentry_key, $sort_subentry_key)
=
_index_entry_element_sort_string_key($customization_information,
- $index_entry, $subentry, $convert_to_text_options,
+ $index_entry, $subentry, $convert_text_options,
$entries_collator);
if ($subentry_key !~ /\S/) {
my $entry_cmdname = $main_entry_element->{'cmdname'};
diff --git a/tp/Texinfo/XS/main/convert_to_text.c
b/tp/Texinfo/XS/main/convert_to_text.c
index 46004fc579..313093bad0 100644
--- a/tp/Texinfo/XS/main/convert_to_text.c
+++ b/tp/Texinfo/XS/main/convert_to_text.c
@@ -126,7 +126,8 @@ destroy_text_options (TEXT_OPTIONS *text_options)
free (text_options->encoding);
free (text_options->expanded_formats);
free_strings_list (&text_options->include_directories);
- if (text_options->other_converter_options)
+ if (text_options->other_converter_options
+ && !text_options->converter)
{
free_options (text_options->other_converter_options);
free (text_options->other_converter_options);
@@ -144,7 +145,7 @@ destroy_text_options (TEXT_OPTIONS *text_options)
tico_option_name(ASCII_GLYPH) \
tico_option_name(TEST)
-/* note that nothing is copied */
+/* note that nothing is allocated, except for the TEXT_OPTIONS themselves */
TEXT_OPTIONS *
copy_options_for_convert_text (CONVERTER *self,
int enable_encoding_if_not_ascii)
@@ -181,6 +182,8 @@ copy_options_for_convert_text (CONVERTER *self,
return options;
}
+static TEXT_OPTIONS text_accents_options;
+
/* format an accent command and nested accents within as Text. */
char *
text_accents (const ELEMENT *accent, char *encoding, int set_case)
@@ -188,10 +191,9 @@ text_accents (const ELEMENT *accent, char *encoding, int
set_case)
ACCENTS_STACK *accent_stack = find_innermost_accent_contents (accent);
char *arg_text;
char *result;
- TEXT_OPTIONS *text_options = new_text_options ();
+ TEXT_OPTIONS *text_options = &text_accents_options;
- if (encoding)
- text_options->encoding = strdup (encoding);
+ text_options->encoding = encoding;
text_options->set_case = set_case;
if (accent_stack->argument)
@@ -206,7 +208,6 @@ text_accents (const ELEMENT *accent, char *encoding, int
set_case)
result = ascii_accents_internal (arg_text, &accent_stack->stack, set_case);
free (arg_text);
destroy_accent_stack (accent_stack);
- destroy_text_options (text_options);
return result;
}
diff --git a/tp/Texinfo/XS/main/convert_to_text.h
b/tp/Texinfo/XS/main/convert_to_text.h
index f3c12dcee9..7d72347ba2 100644
--- a/tp/Texinfo/XS/main/convert_to_text.h
+++ b/tp/Texinfo/XS/main/convert_to_text.h
@@ -8,7 +8,7 @@
#include "converter_types.h"
typedef struct TEXT_OPTIONS {
- int set_case; /* sc and lc */
+ int set_case;
char *encoding; /* enabled_encoding */
int code_state; /* code */
int raw_state;
@@ -19,6 +19,8 @@ typedef struct TEXT_OPTIONS {
EXPANDED_FORMAT *expanded_formats;
STRING_LIST include_directories;
CONVERTER *converter;
+ /* other_converter_options is used if set. If not set and in some cases
+ self_converter_options, if set, is used. */
OPTIONS *other_converter_options; /* corresponds to converter passed
to convert_to_text text options.
same than converter if set, but can be set directly from perl */
diff --git a/tp/Texinfo/XS/main/get_perl_info.c
b/tp/Texinfo/XS/main/get_perl_info.c
index 984bca7ac0..373a7b9fb8 100644
--- a/tp/Texinfo/XS/main/get_perl_info.c
+++ b/tp/Texinfo/XS/main/get_perl_info.c
@@ -911,7 +911,7 @@ force_conf (CONVERTER *converter, const char *conf, SV
*value)
/* output format specific */
/* map hash reference of Convert::Text options to TEXT_OPTIONS */
-/* TODO more to do? */
+/* TODO more to do? set_case? */
#define FETCH(key) key##_sv = hv_fetch (hv_in, #key, strlen(#key), 0);
TEXT_OPTIONS *
copy_sv_options_for_convert_text (SV *sv_in)
@@ -955,15 +955,27 @@ copy_sv_options_for_convert_text (SV *sv_in)
FETCH(converter)
if (converter_sv)
{
- HV *converter_hv = (HV *) SvRV (*converter_sv);
- SV **conf_sv = hv_fetch (converter_hv, "conf", strlen ("conf"), 0);
- if (conf_sv)
- text_options->other_converter_options
- = init_copy_sv_options (*conf_sv, 0, 1);
+ CONVERTER *converter = get_sv_converter (*converter_sv, 0);
+ if (converter)
+ {
+ text_options->other_converter_options
+ = converter->conf;
+ text_options->converter = converter;
+ }
+ else
+ {
+ HV *converter_hv = (HV *) SvRV (*converter_sv);
+ SV **conf_sv = hv_fetch (converter_hv, "conf", strlen ("conf"), 0);
+ if (conf_sv)
+ text_options->other_converter_options
+ = init_copy_sv_options (*conf_sv, 0, 1);
+ }
+ }
+ else
+ {
+ text_options->self_converter_options
+ = init_copy_sv_options (sv_in, 0, 1);
}
-
- text_options->self_converter_options
- = init_copy_sv_options (sv_in, 0, 1);
return text_options;
}
diff --git a/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
b/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
index 900f9e9c22..80d48356e6 100644
--- a/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
+++ b/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
@@ -468,7 +468,10 @@ index_entry_element_sort_string (SV *document_in, SV
*main_entry_sv, SV *element
if (!sort_string)
RETVAL = newSV (0);
else
- RETVAL = newSVpv_utf8 (sort_string, 0);
+ {
+ RETVAL = newSVpv_utf8 (sort_string, 0);
+ free (sort_string);
+ }
OUTPUT:
RETVAL
- master updated (ae85f7133c -> 03c9181681), Patrice Dumas, 2024/01/09
- [no subject],
Patrice Dumas <=
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09
- [no subject], Patrice Dumas, 2024/01/09