[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[no subject]
|
From: |
Patrice Dumas |
|
Date: |
Wed, 10 Jan 2024 17:21:19 -0500 (EST) |
branch: master
commit 43e3278325f3c250b9c75332a98d376aeb236549
Author: Patrice Dumas <pertusus@free.fr>
AuthorDate: Wed Jan 10 10:57:21 2024 +0100
* tp/Texinfo/Structuring.pm (setup_sortable_index_entries): remove the
$silent argument, not used.
* tp/Texinfo/XS/main/utils.c (set_expanded_formats_from_options): add
to be able to set expanded formats based on EXPANDED_FORMATS
customization option, without a converter information.
* tp/Texinfo/XS/convert/indices_in_conversion.c
(index_entry_element_sort_string), tp/Texinfo/XS/main/get_perl_info.c
(find_index_entry_sv, get_sv_index_entries_sorted_by_letter),
tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
(index_entry_element_sort_string): return as reference the INDEX in
find_index_entry_sv. Pass in_code argument to
index_entry_element_sort_string and use it to modify the text options
directly in the function.
* tp/Texinfo/XS/main/convert_to_text.c (copy_options_for_convert_text)
(copy_converter_options_for_convert_text, destroy_text_options),
tp/Texinfo/XS/main/convert_to_text.h (TEXT_OPTIONS),
tp/Texinfo/XS/main/get_perl_info.c (converter_initialize): use OPTIONS in
copy_options_for_convert_text argument, and add another variant with a
converter in argument, copy_converter_options_for_convert_text.
* tp/Texinfo/XS/main/document.c
(register_document_convert_index_text_options)
(destroy_document_information_except_tree),
tp/Texinfo/XS/convert/converter.c (free_generic_converter),
tp/Texinfo/XS/main/document_types.h (DOCUMENT),
tp/Texinfo/XS/main/converter_types.h (CONVERTER): add
convert_index_text_options to register text options used for index
sorting keys conversion.
* tp/Texinfo/Structuring.pm (%XS_overrides),
tp/Texinfo/XS/main/convert_to_text.c
(set_additional_index_entry_keys_options)
(setup_index_entry_keys_formatting)
(setup_converter_index_entry_keys_formatting),
tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
(setup_index_entry_keys_formatting, index_entry_element_sort_string):
add an override for setup_index_entry_keys_formatting that triggers
initializing TEXT_OPTIONS for a converter or a document to be used for
index sorting in index_entry_element_sort_string.
* tp/t/test_sort.t: call register_XS_document_main_configuration to
have the customization available in document for the
setup_index_entry_keys_formatting override.
---
ChangeLog | 49 +++++++++++++
tp/TODO | 4 +
tp/Texinfo/Convert/Text.pm | 5 +-
tp/Texinfo/Structuring.pm | 21 +++---
tp/Texinfo/XS/convert/converter.c | 3 +
tp/Texinfo/XS/convert/indices_in_conversion.c | 6 +-
tp/Texinfo/XS/convert/indices_in_conversion.h | 2 +-
tp/Texinfo/XS/main/convert_to_text.c | 72 ++++++++++++++----
tp/Texinfo/XS/main/convert_to_text.h | 12 ++-
tp/Texinfo/XS/main/converter_types.h | 1 +
tp/Texinfo/XS/main/document.c | 12 +++
tp/Texinfo/XS/main/document.h | 5 ++
tp/Texinfo/XS/main/document_types.h | 5 ++
tp/Texinfo/XS/main/get_perl_info.c | 27 +++----
tp/Texinfo/XS/main/get_perl_info.h | 4 +-
tp/Texinfo/XS/main/utils.c | 23 +++++-
tp/Texinfo/XS/main/utils.h | 7 +-
tp/Texinfo/XS/parsetexi/end_line.c | 7 +-
tp/Texinfo/XS/parsetexi/indices.c | 2 +-
tp/Texinfo/XS/parsetexi/indices.h | 2 +-
.../XS/structuring_transfo/StructuringTransfoXS.xs | 85 +++++++++++++++++++---
tp/t/test_sort.t | 5 +-
22 files changed, 288 insertions(+), 71 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index d4300b4055..f9fb5a33b4 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,52 @@
+2024-01-10 Patrice Dumas <pertusus@free.fr>
+
+ * tp/Texinfo/Structuring.pm (setup_sortable_index_entries): remove the
+ $silent argument, not used.
+
+ * tp/Texinfo/XS/main/utils.c (set_expanded_formats_from_options): add
+ to be able to set expanded formats based on EXPANDED_FORMATS
+ customization option, without a converter information.
+
+ * tp/Texinfo/XS/convert/indices_in_conversion.c
+ (index_entry_element_sort_string), tp/Texinfo/XS/main/get_perl_info.c
+ (find_index_entry_sv, get_sv_index_entries_sorted_by_letter),
+ tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
+ (index_entry_element_sort_string): return as reference the INDEX in
+ find_index_entry_sv. Pass in_code argument to
+ index_entry_element_sort_string and use it to modify the text options
+ directly in the function.
+
+ * tp/Texinfo/XS/main/convert_to_text.c (copy_options_for_convert_text)
+ (copy_converter_options_for_convert_text, destroy_text_options),
+ tp/Texinfo/XS/main/convert_to_text.h (TEXT_OPTIONS),
+ tp/Texinfo/XS/main/get_perl_info.c (converter_initialize): use OPTIONS
in
+ copy_options_for_convert_text argument, and add another variant with a
+ converter in argument, copy_converter_options_for_convert_text.
+
+ * tp/Texinfo/XS/main/document.c
+ (register_document_convert_index_text_options)
+ (destroy_document_information_except_tree),
+ tp/Texinfo/XS/convert/converter.c (free_generic_converter),
+ tp/Texinfo/XS/main/document_types.h (DOCUMENT),
+ tp/Texinfo/XS/main/converter_types.h (CONVERTER): add
+ convert_index_text_options to register text options used for index
+ sorting keys conversion.
+
+ * tp/Texinfo/Structuring.pm (%XS_overrides),
+ tp/Texinfo/XS/main/convert_to_text.c
+ (set_additional_index_entry_keys_options)
+ (setup_index_entry_keys_formatting)
+ (setup_converter_index_entry_keys_formatting),
+ tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
+ (setup_index_entry_keys_formatting, index_entry_element_sort_string):
+ add an override for setup_index_entry_keys_formatting that triggers
+ initializing TEXT_OPTIONS for a converter or a document to be used for
+ index sorting in index_entry_element_sort_string.
+
+ * tp/t/test_sort.t: call register_XS_document_main_configuration to
+ have the customization available in document for the
+ setup_index_entry_keys_formatting override.
+
2024-01-09 Patrice Dumas <pertusus@free.fr>
* tp/Texinfo/XS/main/utils.c (copy_strings, set_conf_string)
diff --git a/tp/TODO b/tp/TODO
index 8d18307060..7a3454bdbc 100644
--- a/tp/TODO
+++ b/tp/TODO
@@ -19,6 +19,10 @@ Any directory part in the file name argument of the source
cross reference comma
Document that Texinfo::Document::rebuild_document or
Texinfo::Document::rebuild_tree should be called after tree modifications
if the parser is XS but converters are perl.
+Also if conversion with XS, and there is a MainConfig, the configuration
+should be registered in the document if it is needed in structuring and
+indices sorting.
+$main_configuration->register_XS_document_main_configuration($document);
Bugs
diff --git a/tp/Texinfo/Convert/Text.pm b/tp/Texinfo/Convert/Text.pm
index 317e08f555..d2cc173c34 100644
--- a/tp/Texinfo/Convert/Text.pm
+++ b/tp/Texinfo/Convert/Text.pm
@@ -386,7 +386,7 @@ my @text_indicator_converter_options = ('NUMBER_SECTIONS',
'ASCII_GLYPH', 'TEST'
# for error registering,
'DEBUG');
-# TODO not documented, used in many converters
+# TODO not documented
# $SELF is typically a converter object.
# Setup options as used by Texinfo::Convert::Text::convert_to_text
# based on the converter information.
@@ -442,8 +442,7 @@ sub set_options_encoding_if_not_ascii($$)
my $self = shift;
my $options = shift;
my $output_encoding_name = $self->get_conf('OUTPUT_ENCODING_NAME');
- if (defined($output_encoding_name)
- and $output_encoding_name ne 'us-ascii') {
+ if (defined($output_encoding_name) and $output_encoding_name ne 'us-ascii') {
if (defined($options->{'_saved_enabled_encoding'})) {
print STDERR "BUG: if_not_ascii _saved_enabled_encoding set: "
.$options->{'_saved_enabled_encoding'}." / ".
diff --git a/tp/Texinfo/Structuring.pm b/tp/Texinfo/Structuring.pm
index 5ffb9e17e2..8d1bbac0da 100644
--- a/tp/Texinfo/Structuring.pm
+++ b/tp/Texinfo/Structuring.pm
@@ -114,10 +114,10 @@ our %XS_overrides = (
"Texinfo::Structuring::_XS_unsplit"
=> "Texinfo::StructTransfXS::unsplit",
- # TODO the text options are read
- # from perl for each entry instead of once for each index.
"Texinfo::Structuring::index_entry_element_sort_string"
=> "Texinfo::StructTransfXS::index_entry_element_sort_string",
+ "Texinfo::Structuring::setup_index_entry_keys_formatting",
+ => "Texinfo::StructTransfXS::setup_index_entry_keys_formatting",
# Not useful for HTML as functions, as the calling functions are
# already overriden
@@ -2410,14 +2410,19 @@ sub _converter_or_registrar_line_warn($$$$)
# information. It should already be set if it is a converter based
# on Texinfo::Convert::Converter, but otherwise it should be set by
# the caller, setting 'document_descriptor' to document->document_descriptor().
-sub setup_sortable_index_entries($$$$$;$)
+# If $PRESET_KEYS is set, the entries sort keys are set with a collator help
+# and the default sort function can be directly used. If unset, no collator
+# is passed to the functions setting the sort key, but a collator is used for
+# sorting. In practice $PRESET_KEYS is always set, as the default
+# $default_preset_keys, which set to 1 is always used in calling functions
+# to determine the value of $PRESET_KEYS.
+sub setup_sortable_index_entries($$$$$)
{
my $registrar = shift;
my $customization_information = shift;
my $index_entries = shift;
my $indices_information = shift;
my $preset_keys = shift;
- my $silent = shift;
# The 'Non-Ignorable' for variable collation elements means that they are
# treated as normal characters. This allows to have spaces and punctuation
@@ -2495,13 +2500,11 @@ sub setup_sortable_index_entries($$$$$;$)
$entry_cmdname
= $main_entry_element->{'extra'}->{'original_def_cmdname'}
if (!defined($entry_cmdname));
- if (!$silent) {
- _converter_or_registrar_line_warn($registrar,
+ _converter_or_registrar_line_warn($registrar,
$customization_information,
sprintf(__("empty index key in \@%s"),
$entry_cmdname),
$main_entry_element->{'source_info'});
- }
push @entry_keys, '';
push @sort_entry_keys, '';
} else {
@@ -2522,13 +2525,11 @@ sub setup_sortable_index_entries($$$$$;$)
$entry_cmdname
= $main_entry_element->{'extra'}->{'original_def_cmdname'}
if (!defined($entry_cmdname));
- if (!$silent) {
- _converter_or_registrar_line_warn($registrar,
+ _converter_or_registrar_line_warn($registrar,
$customization_information,
sprintf(__("empty index sub entry %d key in \@%s"),
$subentry_nr, $entry_cmdname),
$main_entry_element->{'source_info'});
- }
push @entry_keys, '';
push @sort_entry_keys, '';
} else {
diff --git a/tp/Texinfo/XS/convert/converter.c
b/tp/Texinfo/XS/convert/converter.c
index 55cee45dc2..37308d0250 100644
--- a/tp/Texinfo/XS/convert/converter.c
+++ b/tp/Texinfo/XS/convert/converter.c
@@ -866,6 +866,9 @@ free_generic_converter (CONVERTER *self)
free_options (self->conf);
free (self->conf);
+ if (self->convert_index_text_options)
+ destroy_text_options (self->convert_index_text_options);
+
free_output_files_information (&self->output_files_information);
free_output_unit_files (&self->output_unit_files);
diff --git a/tp/Texinfo/XS/convert/indices_in_conversion.c
b/tp/Texinfo/XS/convert/indices_in_conversion.c
index cd8b43097c..8f05f58e61 100644
--- a/tp/Texinfo/XS/convert/indices_in_conversion.c
+++ b/tp/Texinfo/XS/convert/indices_in_conversion.c
@@ -164,7 +164,7 @@ index_content_element (const ELEMENT *element, int
prefer_reference_element)
char *
index_entry_element_sort_string (INDEX_ENTRY *main_entry,
ELEMENT *index_entry_element,
- TEXT_OPTIONS *options,
+ TEXT_OPTIONS *options, int in_code,
int prefer_reference_element)
{
char *sort_string;
@@ -183,7 +183,11 @@ index_entry_element_sort_string (INDEX_ENTRY *main_entry,
entry_tree_element = index_content_element (index_entry_element,
prefer_reference_element);
+ if (in_code)
+ options->code_state++;
sort_string = convert_to_text (entry_tree_element, options);
+ if (in_code)
+ options->code_state--;
index_ignore_chars = lookup_extra_string (main_entry->entry_element,
"index_ignore_chars");
diff --git a/tp/Texinfo/XS/convert/indices_in_conversion.h
b/tp/Texinfo/XS/convert/indices_in_conversion.h
index 7d9c7f7ab0..71787d165d 100644
--- a/tp/Texinfo/XS/convert/indices_in_conversion.h
+++ b/tp/Texinfo/XS/convert/indices_in_conversion.h
@@ -16,6 +16,6 @@ ELEMENT *index_content_element (const ELEMENT *element,
char *index_entry_element_sort_string (INDEX_ENTRY *main_entry,
ELEMENT *index_entry_element,
- TEXT_OPTIONS *options,
+ TEXT_OPTIONS *options, int in_code,
int prefer_reference_element);
#endif
diff --git a/tp/Texinfo/XS/main/convert_to_text.c
b/tp/Texinfo/XS/main/convert_to_text.c
index 4a009dc1bc..1178fac039 100644
--- a/tp/Texinfo/XS/main/convert_to_text.c
+++ b/tp/Texinfo/XS/main/convert_to_text.c
@@ -127,7 +127,8 @@ destroy_text_options (TEXT_OPTIONS *text_options)
free (text_options->expanded_formats);
free_strings_list (&text_options->include_directories);
if (text_options->other_converter_options
- && !text_options->converter)
+ && !text_options->converter
+ && !text_options->other_options)
{
free_options (text_options->other_converter_options);
free (text_options->other_converter_options);
@@ -148,34 +149,73 @@ destroy_text_options (TEXT_OPTIONS *text_options)
/* the string and strlist options need to be copied, in case they are
deallocated if options are reset */
TEXT_OPTIONS *
-copy_options_for_convert_text (CONVERTER *self)
+copy_options_for_convert_text (OPTIONS *options)
{
- TEXT_OPTIONS *options = new_text_options ();
+ TEXT_OPTIONS *text_options = new_text_options ();
int text_indicator_option;
- if (self->conf->ENABLE_ENCODING.integer > 0
- && self->conf->OUTPUT_ENCODING_NAME.string)
+ if (options->ENABLE_ENCODING.integer > 0
+ && options->OUTPUT_ENCODING_NAME.string)
{
- options->encoding = strdup (self->conf->OUTPUT_ENCODING_NAME.string);
+ text_options->encoding = strdup (options->OUTPUT_ENCODING_NAME.string);
}
#define tico_option_name(name) \
- text_indicator_option = self->conf->name.integer; \
- if (text_indicator_option > 0) { options->name = 1; } \
- else if (text_indicator_option >= 0) { options->name = 0; }
+ text_indicator_option = options->name.integer; \
+ if (text_indicator_option > 0) { text_options->name = 1; } \
+ else if (text_indicator_option >= 0) { text_options->name = 0; }
TEXT_INDICATOR_CONVERTER_OPTIONS
#undef tico_option_name
- memcpy (options->expanded_formats, self->expanded_formats,
- expanded_formats_number () * sizeof (EXPANDED_FORMAT));
+ set_expanded_formats_from_options (text_options->expanded_formats, options);
- copy_strings (&options->include_directories,
- self->conf->INCLUDE_DIRECTORIES.strlist);
+ copy_strings (&text_options->include_directories,
+ options->INCLUDE_DIRECTORIES.strlist);
- options->other_converter_options = self->conf;
- options->converter = self;
+ text_options->other_converter_options = options;
- return options;
+ return text_options;
+}
+
+TEXT_OPTIONS *
+copy_converter_options_for_convert_text (CONVERTER *self)
+{
+ TEXT_OPTIONS *text_options = copy_options_for_convert_text (self->conf);
+ text_options->converter = self;
+ return text_options;
+}
+
+/* In Structuring.pm */
+static void
+set_additional_index_entry_keys_options (OPTIONS *options,
+ TEXT_OPTIONS *text_options)
+{
+ if (options->ENABLE_ENCODING.integer <= 0
+ || !(options->OUTPUT_ENCODING_NAME.string
+ && !strcasecmp (options->OUTPUT_ENCODING_NAME.string, "utf-8")))
+ {
+ text_options->sort_string = 1;
+ }
+}
+
+/* there are two variants, to setup the text options used for index
+ index entries formatting as text in case this is done with a
+ converter or without. */
+TEXT_OPTIONS *
+setup_index_entry_keys_formatting (OPTIONS *options)
+{
+ TEXT_OPTIONS *text_options = copy_options_for_convert_text (options);
+ text_options->other_options = 1;
+ set_additional_index_entry_keys_options (options, text_options);
+ return text_options;
+}
+
+TEXT_OPTIONS *
+setup_converter_index_entry_keys_formatting (CONVERTER *self)
+{
+ TEXT_OPTIONS *text_options = copy_converter_options_for_convert_text (self);
+ set_additional_index_entry_keys_options (self->conf, text_options);
+ return text_options;
}
void
diff --git a/tp/Texinfo/XS/main/convert_to_text.h
b/tp/Texinfo/XS/main/convert_to_text.h
index 82b44bbfe0..92d1d01d13 100644
--- a/tp/Texinfo/XS/main/convert_to_text.h
+++ b/tp/Texinfo/XS/main/convert_to_text.h
@@ -17,14 +17,15 @@ typedef struct TEXT_OPTIONS {
int ASCII_GLYPH;
int TEST;
int NUMBER_SECTIONS;
+ int other_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 */
+ OPTIONS *other_converter_options; /* same than converter if converter
+ is set, but can be set directly from perl or from another source,
+ in which case other_options is set */
OPTIONS *self_converter_options; /* text converter options available
as the OPTIONS structure */
int document_descriptor;
@@ -33,7 +34,10 @@ typedef struct TEXT_OPTIONS {
char *convert_to_text (const ELEMENT *root, TEXT_OPTIONS *text_options);
TEXT_OPTIONS *new_text_options (void);
void destroy_text_options (TEXT_OPTIONS *text_options);
-TEXT_OPTIONS *copy_options_for_convert_text (CONVERTER *self);
+TEXT_OPTIONS *copy_options_for_convert_text (OPTIONS *options);
+TEXT_OPTIONS *copy_converter_options_for_convert_text (CONVERTER *self);
+TEXT_OPTIONS *setup_index_entry_keys_formatting (OPTIONS *options);
+TEXT_OPTIONS *setup_converter_index_entry_keys_formatting (CONVERTER *self);
void text_set_options_encoding_if_not_ascii (CONVERTER *self,
TEXT_OPTIONS *text_options);
diff --git a/tp/Texinfo/XS/main/converter_types.h
b/tp/Texinfo/XS/main/converter_types.h
index c6c9e3c119..7903997e3c 100644
--- a/tp/Texinfo/XS/main/converter_types.h
+++ b/tp/Texinfo/XS/main/converter_types.h
@@ -712,6 +712,7 @@ typedef struct CONVERTER {
int document_units_descriptor;
struct TEXT_OPTIONS *convert_text_options;
+ struct TEXT_OPTIONS *convert_index_text_options;
/* output unit files API */
FILE_NAME_PATH_COUNTER_LIST output_unit_files;
diff --git a/tp/Texinfo/XS/main/document.c b/tp/Texinfo/XS/main/document.c
index b85b536f54..2568fd1eed 100644
--- a/tp/Texinfo/XS/main/document.c
+++ b/tp/Texinfo/XS/main/document.c
@@ -30,6 +30,7 @@
#include "debug.h"
/* for delete_global_info and wipe_index */
#include "utils.h"
+#include "convert_to_text.h"
#include "document.h"
/* note that each time the document list is reallocated, pointers
@@ -132,6 +133,13 @@ register_document_options (DOCUMENT *document, OPTIONS
*options)
document->options = options;
}
+void
+register_document_convert_index_text_options (DOCUMENT *document,
+ TEXT_OPTIONS *text_options)
+{
+ document->convert_index_text_options = text_options;
+}
+
void
destroy_document_information_except_tree (DOCUMENT *document)
{
@@ -162,6 +170,10 @@ destroy_document_information_except_tree (DOCUMENT
*document)
free_options (document->options);
free (document->options);
}
+ if (document->convert_index_text_options)
+ {
+ destroy_text_options (document->convert_index_text_options);
+ }
}
}
diff --git a/tp/Texinfo/XS/main/document.h b/tp/Texinfo/XS/main/document.h
index d883d34319..f9dbd64e06 100644
--- a/tp/Texinfo/XS/main/document.h
+++ b/tp/Texinfo/XS/main/document.h
@@ -14,6 +14,8 @@ typedef struct TREE_AND_STRINGS {
STRING_LIST *small_strings;
} TREE_AND_STRINGS;
+struct TEXT_OPTIONS;
+
DOCUMENT *retrieve_document (int document_descriptor);
size_t register_document (ELEMENT *root, INDEX **index_names,
@@ -30,6 +32,9 @@ void register_document_nodes_list (DOCUMENT *document,
void register_document_sections_list (DOCUMENT *document,
ELEMENT_LIST *sections_list);
void register_document_options (DOCUMENT *document, OPTIONS *options);
+void register_document_convert_index_text_options (DOCUMENT *document,
+ struct TEXT_OPTIONS *text_options);
+
void remove_document_descriptor (int document_descriptor);
TREE_AND_STRINGS *unregister_document_descriptor_tree (int
document_descriptor);
void clear_document_errors (int document_descriptor);
diff --git a/tp/Texinfo/XS/main/document_types.h
b/tp/Texinfo/XS/main/document_types.h
index 92c1884f9f..7ce19de6ae 100644
--- a/tp/Texinfo/XS/main/document_types.h
+++ b/tp/Texinfo/XS/main/document_types.h
@@ -22,12 +22,15 @@
#include "tree_types.h"
/*
#include "options_types.h"
+#include "convert_to_text.h"
*/
#include "global_commands_types.h"
/* to avoid interdependency with options_types.h, including for
other include files */
struct OPTIONS;
+/* same with convert_to_text.h */
+struct TEXT_OPTIONS;
enum error_type { MSG_error, MSG_warning,
MSG_document_error, MSG_document_warning };
@@ -75,6 +78,8 @@ typedef struct DOCUMENT {
ELEMENT_LIST *sections_list;
ERROR_MESSAGE_LIST *error_messages;
struct OPTIONS *options; /* for options used in structuring */
+ struct TEXT_OPTIONS *convert_index_text_options; /* for index
+ sorting without converter */
} DOCUMENT;
/* not in document, but used in parser */
diff --git a/tp/Texinfo/XS/main/get_perl_info.c
b/tp/Texinfo/XS/main/get_perl_info.c
index bf12fda410..d137193fec 100644
--- a/tp/Texinfo/XS/main/get_perl_info.c
+++ b/tp/Texinfo/XS/main/get_perl_info.c
@@ -607,7 +607,7 @@ converter_initialize (SV *converter_sv)
set_output_encoding (converter->conf, converter->document);
converter->convert_text_options
- = copy_options_for_convert_text (converter);
+ = copy_converter_options_for_convert_text (converter);
converter->hv = hv_in;
@@ -639,15 +639,16 @@ reset_output_init_conf (SV *sv_in)
}
INDEX_ENTRY *
-find_index_entry_sv (SV *index_entry_sv, INDEX **index_names,
- const char *warn_string, char **entry_index_name,
+find_index_entry_sv (const SV *index_entry_sv, INDEX **index_names,
+ const char *warn_string, const INDEX **entry_idx,
int *entry_number)
{
HV *index_entry_hv;
SV **index_name_sv;
SV **entry_number_sv;
int entry_idx_in_index;
- INDEX *idx;
+ char *entry_index_name = 0;
+ const INDEX *idx;
dTHX;
@@ -657,9 +658,6 @@ find_index_entry_sv (SV *index_entry_sv, INDEX
**index_names,
entry_number_sv = hv_fetch (index_entry_hv, "entry_number",
strlen ("entry_number"), 0);
- *entry_index_name = 0;
- *entry_number = 0;
-
if (!index_name_sv || !entry_number_sv)
{
char *msg;
@@ -669,13 +667,13 @@ find_index_entry_sv (SV *index_entry_sv, INDEX
**index_names,
xasprintf (&msg, "%s: no entry info\n", warn_str);
fatal (msg);
}
- *entry_index_name = (char *) SvPVutf8_nolen (*index_name_sv);
+ entry_index_name = (char *) SvPVutf8_nolen (*index_name_sv);
*entry_number = SvIV (*entry_number_sv);
entry_idx_in_index = *entry_number - 1;
idx = indices_info_index_by_name (index_names,
- *entry_index_name);
-
+ entry_index_name);
+ *entry_idx = idx;
if (idx)
{
if (entry_idx_in_index < idx->entries_number)
@@ -849,7 +847,7 @@ get_sv_index_entries_sorted_by_letter (INDEX **index_names,
for (k = 0; k < entries_nr; k++)
{
SV** index_entry_sv = av_fetch (entries_av, k, 0);
- char *entry_index_name;
+ const INDEX *entry_idx = 0;
int entry_number;
char *warn_string;
INDEX_ENTRY *index_entry = 0;
@@ -866,7 +864,7 @@ get_sv_index_entries_sorted_by_letter (INDEX **index_names,
"get_sv_index_entries_sorted_by_letter: %s: %d: %s:
%d",
idx_name, i, letter_entries->letter, k);
index_entry = find_index_entry_sv (*index_entry_sv,
index_names,
- warn_string,
&entry_index_name,
+ warn_string, &entry_idx,
&entry_number);
free (warn_string);
@@ -875,6 +873,9 @@ get_sv_index_entries_sorted_by_letter (INDEX **index_names,
if (!letter_entries->entries[k])
{
char *msg;
+ char *entry_index_name = 0;
+ if (entry_idx)
+ entry_index_name = entry_idx->name;
xasprintf (&msg,
"BUG: index %s letter %s position %d: %s entry %d not found\n",
idx_name, letter_string, k,
@@ -1254,7 +1255,7 @@ find_document_index_entry_extra_index_entry_sv (DOCUMENT
*document,
AV *extra_index_entry_av;
SV **index_name_sv;
char *index_name = 0;
- INDEX *idx = 0;
+ const INDEX *idx = 0;
dTHX;
diff --git a/tp/Texinfo/XS/main/get_perl_info.h
b/tp/Texinfo/XS/main/get_perl_info.h
index 5be575eff5..5e61413dd8 100644
--- a/tp/Texinfo/XS/main/get_perl_info.h
+++ b/tp/Texinfo/XS/main/get_perl_info.h
@@ -42,8 +42,8 @@ CONVERTER *get_sv_converter (SV *sv_in, const char
*warn_string);
int converter_initialize (SV *converter_sv);
void reset_output_init_conf (SV *sv_in);
-INDEX_ENTRY *find_index_entry_sv (SV *index_entry_sv, INDEX **index_names,
- const char *warn_string, char **entry_index_name,
+INDEX_ENTRY *find_index_entry_sv (const SV *index_entry_sv, INDEX
**index_names,
+ const char *warn_string, const INDEX **entry_idx,
int *entry_number);
INDEX_SORTED_BY_LETTER *get_sv_index_entries_sorted_by_letter
diff --git a/tp/Texinfo/XS/main/utils.c b/tp/Texinfo/XS/main/utils.c
index f3e9de84d6..dc0cdf9dd9 100644
--- a/tp/Texinfo/XS/main/utils.c
+++ b/tp/Texinfo/XS/main/utils.c
@@ -518,7 +518,7 @@ clear_expanded_formats (EXPANDED_FORMAT *formats)
}
void
-add_expanded_format (EXPANDED_FORMAT *formats, char *format)
+add_expanded_format (EXPANDED_FORMAT *formats, const char *format)
{
int i;
for (i = 0; i < sizeof (expanded_formats)/sizeof (*expanded_formats);
@@ -563,6 +563,22 @@ expanded_formats_number (void)
return sizeof (expanded_formats)/sizeof (*expanded_formats);
}
+void
+set_expanded_formats_from_options (EXPANDED_FORMAT *formats,
+ const OPTIONS *options)
+{
+ if (options->EXPANDED_FORMATS.strlist
+ && (options->EXPANDED_FORMATS.strlist->number))
+ {
+ size_t i;
+ for (i = 0; i < options->EXPANDED_FORMATS.strlist->number; i++)
+ {
+ add_expanded_format (formats,
+ options->EXPANDED_FORMATS.strlist->list[i]);
+ }
+ }
+}
+
/* Return the parent if in an item_line command, @*table */
ELEMENT *
@@ -594,9 +610,10 @@ get_label_element (const ELEMENT *e)
/* index related code used both in parsing and conversion */
/* NAME is the name of an index, e.g. "cp" */
INDEX *
-indices_info_index_by_name (INDEX **indices_information, char *name)
+indices_info_index_by_name (INDEX **indices_information, const char *name)
{
- INDEX **i, *idx;
+ INDEX **i;
+ INDEX *idx;
for (i = indices_information; (idx = *i); i++)
if (!strcmp (idx->name, name))
diff --git a/tp/Texinfo/XS/main/utils.h b/tp/Texinfo/XS/main/utils.h
index 4e987e3685..a3657badf8 100644
--- a/tp/Texinfo/XS/main/utils.h
+++ b/tp/Texinfo/XS/main/utils.h
@@ -166,7 +166,8 @@ void delete_global_commands (GLOBAL_COMMANDS
*global_commands_ref);
char *normalize_encoding_name (const char *text, int *possible_encoding);
ELEMENT *item_line_parent (ELEMENT *current);
ELEMENT *get_label_element (const ELEMENT *e);
-INDEX *indices_info_index_by_name (INDEX **indices_information, char *name);
+INDEX *indices_info_index_by_name (INDEX **indices_information,
+ const char *name);
INDEX *ultimate_index (INDEX *index);
size_t index_number_index_by_name (const SORTED_INDEX_NAMES *sorted_indices,
const char *name);
@@ -215,9 +216,11 @@ char *encode_string (char *input_string, char *encoding,
int *status,
EXPANDED_FORMAT *new_expanded_formats (void);
void clear_expanded_formats (EXPANDED_FORMAT *formats);
-void add_expanded_format (EXPANDED_FORMAT *formats, char *format);
+void add_expanded_format (EXPANDED_FORMAT *formats, const char *format);
int format_expanded_p (EXPANDED_FORMAT *formats, const char *format);
int expanded_formats_number (void);
+void set_expanded_formats_from_options (EXPANDED_FORMAT *formats,
+ const OPTIONS *options);
char *enumerate_item_representation (char *specification, int number);
diff --git a/tp/Texinfo/XS/parsetexi/end_line.c
b/tp/Texinfo/XS/parsetexi/end_line.c
index ef4371aa93..ed61982df7 100644
--- a/tp/Texinfo/XS/parsetexi/end_line.c
+++ b/tp/Texinfo/XS/parsetexi/end_line.c
@@ -367,7 +367,8 @@ parse_line_command_args (ELEMENT *line_command)
{
/* synindex FROM TO */
char *index_name_from = 0, *index_name_to = 0;
- INDEX *from_index, *to_index;
+ INDEX *from_index;
+ INDEX *to_index;
char *p = line;
if (!isascii_alnum (*p))
@@ -386,8 +387,8 @@ parse_line_command_args (ELEMENT *line_command)
if (*p)
goto synindex_invalid; /* More at end of line. */
- from_index = indices_info_index_by_name (index_names,index_name_from);
- to_index = indices_info_index_by_name (index_names,index_name_to);
+ from_index = indices_info_index_by_name (index_names, index_name_from);
+ to_index = indices_info_index_by_name (index_names, index_name_to);
if (!from_index)
line_error ("unknown source index in @%s: %s",
command_name(cmd), index_name_from);
diff --git a/tp/Texinfo/XS/parsetexi/indices.c
b/tp/Texinfo/XS/parsetexi/indices.c
index b31653f97c..25319fcf06 100644
--- a/tp/Texinfo/XS/parsetexi/indices.c
+++ b/tp/Texinfo/XS/parsetexi/indices.c
@@ -124,7 +124,7 @@ add_index_internal (char *name, int in_code)
/* Add a user defined index with the name NAME */
void
-add_index (char *name, int in_code)
+add_index (const char *name, int in_code)
{
INDEX *idx = indices_info_index_by_name (index_names, name);
char *cmdname;
diff --git a/tp/Texinfo/XS/parsetexi/indices.h
b/tp/Texinfo/XS/parsetexi/indices.h
index 623cee4320..926acd8da3 100644
--- a/tp/Texinfo/XS/parsetexi/indices.h
+++ b/tp/Texinfo/XS/parsetexi/indices.h
@@ -8,7 +8,7 @@ extern INDEX **index_names;
void init_index_commands (void);
-void add_index (char *name, int in_code);
+void add_index (const char *name, int in_code);
INDEX *index_of_command (enum command_id cmd);
void enter_index_entry (enum command_id index_type_cmd,
ELEMENT *current);
diff --git a/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
b/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
index 80d48356e6..899fd77531 100644
--- a/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
+++ b/tp/Texinfo/XS/structuring_transfo/StructuringTransfoXS.xs
@@ -438,31 +438,98 @@ split_pages (SV *output_units_in, char *split)
if (output_units)
split_pages (output_units, split);
+# This function triggers setting the information needed for calls
+# to index_entry_element_sort_string in C, either in a document
+# or in a converter, depending whether index sorting is done with a converter
+# or without.
+# A returned hash reference is needed as some information is set in the
+# hash in perl afterwards, but this information is not used as the hash
+# is not used by any function not overriden, so there is no need to
+# return information corresponding to the text options for perl. An empty
+# hash reference is therefore returned.
SV *
-index_entry_element_sort_string (SV *document_in, SV *main_entry_sv, SV
*element_sv, SV *options_sv, SV *prefer_reference_element_sv=0)
+setup_index_entry_keys_formatting (SV *customization_info_sv)
PREINIT:
+ CONVERTER *self;
+ HV *hv;
+ CODE:
+ self = get_sv_converter (customization_info_sv, 0);
+ if (self)
+ {
+ TEXT_OPTIONS *text_options
+ = setup_converter_index_entry_keys_formatting (self);
+ self->convert_index_text_options = text_options;
+ }
+ else
+ {
+ DOCUMENT *document;
+ document = get_sv_document_document (customization_info_sv,
+ "setup_index_entry_keys_formatting");
+ if (document && document->options)
+ {
+ TEXT_OPTIONS *text_options
+ = setup_index_entry_keys_formatting (document->options);
+ register_document_convert_index_text_options (document,
+ text_options);
+ }
+ }
+ hv = newHV ();
+ RETVAL = newRV_noinc ((SV *) hv);
+ OUTPUT:
+ RETVAL
+
+SV *
+index_entry_element_sort_string (SV *customization_info_sv, SV *main_entry_sv,
SV *element_sv, SV *options_sv, SV *prefer_reference_element_sv=0)
+ PREINIT:
+ CONVERTER *self;
DOCUMENT *document;
char *sort_string = 0;
+ TEXT_OPTIONS *convert_index_text_options = 0;
CODE:
- document = get_sv_document_document (document_in,
- "index_entry_element_sort_string");
+ self = get_sv_converter (customization_info_sv, 0);
+ if (self)
+ {
+ document = self->document;
+ convert_index_text_options = self->convert_index_text_options;
+ }
+ else
+ {
+ document = get_sv_document_document (customization_info_sv,
+ "index_entry_element_sort_string");
+ if (document)
+ convert_index_text_options
+ = document->convert_index_text_options;
+ }
if (document)
{
- char *entry_index_name;
+ const INDEX *entry_idx = 0;
int entry_number;
int prefer_reference_element = 0;
+ int in_code;
+ int allocated_text_options = 0;
ELEMENT *element = find_element_from_sv (0, document,
element_sv, 0);
INDEX_ENTRY *main_entry = find_index_entry_sv (main_entry_sv,
document->index_names, 0,
- &entry_index_name, &entry_number);
- TEXT_OPTIONS *options
- = copy_sv_options_for_convert_text (options_sv);
+ &entry_idx, &entry_number);
+ in_code = entry_idx->in_code;
+
+ /* should not be possible */
+ if (!convert_index_text_options)
+ {
+ allocated_text_options = 1;
+ convert_index_text_options
+ = copy_sv_options_for_convert_text (options_sv);
+ }
+
if (prefer_reference_element_sv && SvOK
(prefer_reference_element_sv))
prefer_reference_element = SvIV (prefer_reference_element_sv);
+
sort_string = index_entry_element_sort_string (main_entry,
- element, options, prefer_reference_element);
- destroy_text_options (options);
+ element, convert_index_text_options, in_code,
+ prefer_reference_element);
+ if (allocated_text_options)
+ destroy_text_options (convert_index_text_options);
}
if (!sort_string)
diff --git a/tp/t/test_sort.t b/tp/t/test_sort.t
index eef08d15c0..872120b0b0 100644
--- a/tp/t/test_sort.t
+++ b/tp/t/test_sort.t
@@ -51,6 +51,7 @@ my $main_configuration =
Texinfo::MainConfig::new({'ENABLE_ENCODING' => 1});
Texinfo::Common::set_output_encodings($main_configuration,
$document_information);
$main_configuration->{'document_descriptor'}
= $document->document_descriptor();
+$main_configuration->register_XS_document_main_configuration($document);
my ($sorted_index_entries, $index_entries_sort_strings)
= Texinfo::Structuring::sort_indices_by_index($registrar,
$main_configuration,
$index_entries,
$indices_information);
@@ -62,7 +63,7 @@ foreach my $entry (@{$sorted_index_entries->{'cp'}}) {
use utf8;
-# e is before E because they are equal and e appears before E in the document
+# e is before E because they are equal and e appears before E in the document
my @entries_ref = ('!', '"', 'aaaaaaaaaaaa', 'e', 'E', 'ẽ', 'ł');
cmp_deeply (\@entries, \@entries_ref, 'sorted index entries');
@@ -81,7 +82,7 @@ my @letter_entries_ref = (
{'E' => [ 'e', 'E', 'ẽ' ]},
{'Ł' => [ 'ł' ]},
);
-
+
my @letter_entries;
foreach my $letter (@{$sorted_index_entries_by_letter->{'cp'}}) {
my $letter_entry = {};