# Structuring.pm: extract information about a document structure based on the # document tree. # # Copyright 2010-2023 Free Software Foundation, Inc. # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 3 of the License, # or (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program. If not, see . # # Original author: Patrice Dumas # Parts (also from Patrice Dumas) come from texi2html.pl. package Texinfo::Structuring; use 5.00405; # See comment at start of HTML.pm use if $] >= 5.012, feature => 'unicode_strings'; use strict; # Can be used to check that there is no incorrect autovivfication #no autovivification qw(fetch delete exists store strict); use Carp qw(cluck confess); use Unicode::Normalize; # for %root_commands use Texinfo::Commands; use Texinfo::Common; # for error messages use Texinfo::Convert::Texinfo qw(target_element_to_texi_label link_element_to_texi); # for debugging. Also for index entries sorting. use Texinfo::Convert::Text; # for internal references and misc uses use Texinfo::Convert::NodeNameNormalization; # for new_master_menu translations use Texinfo::Translations; require Exporter; use vars qw($VERSION @ISA @EXPORT_OK %EXPORT_TAGS); @ISA = qw(Exporter); %EXPORT_TAGS = ( 'all' => [ qw( associate_internal_references check_nodes_are_referenced complete_node_tree_with_menus elements_directions elements_file_directions merge_indices nodes_tree number_floats sectioning_structure set_menus_node_directions sort_indices split_by_node split_by_section split_pages warn_non_empty_parts ) ] ); @EXPORT_OK = ( @{ $EXPORT_TAGS{'all'} } ); $VERSION = '7.1'; my %types_to_enter; foreach my $type_to_enter ('brace_command_arg', 'line_arg', 'paragraph') { $types_to_enter{$type_to_enter} = 1; } my %command_structuring_level = %Texinfo::Common::command_structuring_level; my %appendix_commands; my %unnumbered_commands; foreach my $command (keys(%command_structuring_level)) { if ($command =~ /appendix/) { $appendix_commands{$command} = 1; } elsif ($command =~ /unnumbered/) { $unnumbered_commands{$command} = 1; } } $unnumbered_commands{'top'} = 1; $unnumbered_commands{'centerchap'} = 1; $unnumbered_commands{'part'} = 1; # Go through the sectioning commands (e.g. @chapter, not @node), and # set: # 'section_level' # 'section_number' # 'section_childs' # 'section_up' # 'section_prev' # 'section_next' # 'toplevel_next' # 'toplevel_prev' # 'toplevel_up' sub sectioning_structure($$$) { my $registrar = shift; my $customization_information = shift; my $root = shift; my $sec_root; my $previous_section; my $previous_toplevel; my $in_appendix = 0; # lowest level with a number. This is the lowest level above 0. my $number_top_level; my $section_top; my @sections_list; # holds the current number for all the levels. It is not possible to use # something like the last child index, because of @unnumbered. my @command_numbers; # keep track of the unnumbered my @command_unnumbered; foreach my $content (@{$root->{'contents'}}) { if (!$content->{'cmdname'} or $content->{'cmdname'} eq 'node' or $content->{'cmdname'} eq 'bye') { next; } push @sections_list, $content; if ($content->{'cmdname'} eq 'top') { if (! $section_top) { $section_top = $content; } } my $level; $content->{'structure'} = {} if (! $content->{'structure'}); $level = $content->{'structure'}->{'section_level'} = Texinfo::Common::section_level($content); if (!defined($level)) { warn "bug: level not defined for $content->{'cmdname'} " .Texinfo::Common::debug_print_element($content)."\n"; $level = $content->{'structure'}->{'section_level'} = 0; } if ($previous_section) { # new command is below if ($previous_section->{'structure'}->{'section_level'} < $level) { if ($level - $previous_section->{'structure'}->{'section_level'} > 1) { $registrar->line_error($customization_information, sprintf(__("raising the section level of \@%s which is too low"), $content->{'cmdname'}), $content->{'source_info'}); $content->{'structure'}->{'section_level'} = $previous_section->{'structure'}->{'section_level'} + 1; } $previous_section->{'structure'}->{'section_childs'} = [$content]; $content->{'structure'}->{'section_up'} = $previous_section; # if the up is unnumbered, the number information has to be kept, # to avoid reusing an already used number. if (!$unnumbered_commands{$previous_section->{'cmdname'}}) { $command_numbers[$content->{'structure'}->{'section_level'}] = undef; } elsif (!$unnumbered_commands{$content->{'cmdname'}}) { $command_numbers[$content->{'structure'}->{'section_level'}]++; } if ($unnumbered_commands{$content->{'cmdname'}}) { $command_unnumbered[$content->{'structure'}->{'section_level'}] = 1; } else { $command_unnumbered[$content->{'structure'}->{'section_level'}] = 0; } } else { my $up = $previous_section->{'structure'}->{'section_up'}; my $new_upper_part_element; if ($previous_section->{'structure'}->{'section_level'} != $level) { # means it is above the previous command, the up is to be found while ($up->{'structure'}->{'section_up'} and $up->{'structure'}->{'section_level'} >= $level) { $up = $up->{'structure'}->{'section_up'}; } if ($level <= $up->{'structure'}->{'section_level'}) { if ($content->{'cmdname'} eq 'part') { $new_upper_part_element = 1; if ($level < $up->{'structure'}->{'section_level'}) { $registrar->line_warn($customization_information, sprintf(__("no chapter-level command before \@%s"), $content->{'cmdname'}), $content->{'source_info'}); } } else { $registrar->line_warn($customization_information, sprintf(__("lowering the section level of \@%s appearing after a lower element"), $content->{'cmdname'}), $content->{'source_info'}); $content->{'structure'}->{'section_level'} = $up->{'structure'}->{'section_level'} + 1; } } } if ($appendix_commands{$content->{'cmdname'}} and !$in_appendix and $content->{'structure'}->{'section_level'} <= $number_top_level and $up->{'cmdname'} and $up->{'cmdname'} eq 'part') { $up = $up->{'structure'}->{'section_up'}; } if ($new_upper_part_element) { # In that case the root has to be updated because the first # 'part' just appeared $sec_root->{'structure'}->{'section_level'} = $level - 1; push @{$sec_root->{'structure'}->{'section_childs'}}, $content; $content->{'structure'}->{'section_up'} = $sec_root; $number_top_level = $level; $number_top_level++ if (!$number_top_level); } else { push @{$up->{'structure'}->{'section_childs'}}, $content; $content->{'structure'}->{'section_up'} = $up; $content->{'structure'}->{'section_prev'} = $up->{'structure'}->{'section_childs'}->[-2]; $content->{'structure'}->{'section_prev'} ->{'structure'}->{'section_next'} = $content; } if (!$unnumbered_commands{$content->{'cmdname'}}) { $command_numbers[$content->{'structure'}->{'section_level'}]++; $command_unnumbered[$content->{'structure'}->{'section_level'}] = 0; } else { $command_unnumbered[$content->{'structure'}->{'section_level'}] = 1; } } } else { $sec_root = {'structure' => {}}; # first section determines the level of the root. It is # typically -1 when there is a @top. $sec_root->{'structure'}->{'section_level'} = $level - 1; $sec_root->{'structure'}->{'section_childs'} = [$content]; $content->{'structure'}->{'section_up'} = $sec_root; $number_top_level = $level; # if $level of top sectioning element is 0, which means that # it is a @top, $number_top_level is 1 as it is associated to # the level of chapter/unnumbered... $number_top_level++ if (!$number_top_level); if ($content->{'cmdname'} ne 'top') { if (!$unnumbered_commands{$content->{'cmdname'}}) { $command_unnumbered[$content->{'structure'}->{'section_level'}] = 0; } else { $command_unnumbered[$content->{'structure'}->{'section_level'}] = 1; } } } if (!defined($command_numbers[$content->{'structure'}->{'section_level'}])) { if ($unnumbered_commands{$content->{'cmdname'}}) { $command_numbers[$content->{'structure'}->{'section_level'}] = 0; } else { $command_numbers[$content->{'structure'}->{'section_level'}] = 1; } } if ($appendix_commands{$content->{'cmdname'}} and !$in_appendix) { $in_appendix = 1; $command_numbers[$content->{'structure'}->{'section_level'}] = 'A'; } if (!$unnumbered_commands{$content->{'cmdname'}}) { # construct the number, if not below an unnumbered if (!$command_unnumbered[$number_top_level]) { my $section_number = $command_numbers[$number_top_level]; for (my $i = $number_top_level+1; $i <= $content->{'structure'}->{'section_level'}; $i++) { $section_number .= ".$command_numbers[$i]"; # If there is an unnumbered above, then no number is added. if ($command_unnumbered[$i]) { $section_number = undef; last; } } if (defined($section_number)) { $content->{'structure'}->{'section_number'} = $section_number; } } } $previous_section = $content; if ($content->{'cmdname'} ne 'part' and $content->{'structure'}->{'section_level'} <= $number_top_level) { if ($previous_toplevel) { $previous_toplevel->{'structure'}->{'toplevel_next'} = $content; $content->{'structure'}->{'toplevel_prev'} = $previous_toplevel; } $previous_toplevel = $content; if ($section_top and $content ne $section_top) { $content->{'structure'}->{'toplevel_up'} = $section_top; } } elsif ($content->{'cmdname'} eq 'part' and not ($content->{'extra'} and $content->{'extra'}->{'part_associated_section'})) { $registrar->line_warn($customization_information, sprintf(__("no sectioning command associated with \@%s"), $content->{'cmdname'}), $content->{'source_info'}); } } return $sec_root, \@sections_list; } # for debugging sub _print_sectioning_tree($); sub _print_sectioning_tree($) { my $current = shift; my $result = ' ' x $current->{'structure'}->{'section_level'} . Texinfo::Convert::Texinfo::root_heading_command_to_texinfo($current)."\n"; foreach my $child (@{$current->{'structure'}->{'section_childs'}}) { $result .= _print_sectioning_tree($child); } return $result; } sub warn_non_empty_parts($$$) { my $registrar = shift; my $customization_information = shift; my $global_commands = shift; if ($global_commands->{'part'}) { foreach my $part (@{$global_commands->{'part'}}) { if (!Texinfo::Common::is_content_empty($part)) { $registrar->line_warn($customization_information, sprintf(__("\@%s not empty"), $part->{'cmdname'}), $part->{'source_info'}); } } } } # $REFERENCE_NODE should always be a target element associated to # a label. sub _check_node_same_texinfo_code($$) { my $reference_node = shift; my $node_extra = shift; my $reference_node_texi; if (defined($reference_node->{'extra'}->{'normalized'})) { my $label_element = Texinfo::Common::get_label_element($reference_node); $reference_node_texi = Texinfo::Convert::Texinfo::convert_to_texinfo( {'contents' => $label_element->{'contents'}}); $reference_node_texi =~ s/\s+/ /g; } else { $reference_node_texi = ''; } my $node_texi; if ($node_extra and $node_extra->{'node_content'}) { my @contents_node = @{$node_extra->{'node_content'}}; pop @contents_node if ($contents_node[-1]->{'type'} and $contents_node[-1]->{'type'} eq 'space_at_end_menu_node'); $node_texi = Texinfo::Convert::Texinfo::convert_to_texinfo( {'contents' => \@contents_node}); $node_texi =~ s/\s+/ /g; } else { $node_texi = ''; } return ($reference_node_texi eq $node_texi); } my @node_directions = ('next', 'prev', 'up'); # No translation of those special Info keywords. my %direction_texts = ( 'prev' => 'Prev', 'next' => 'Next', 'up' => 'Up' ); sub _check_menu_entry($$$$$$) { my $registrar = shift; my $customization_information = shift; my $labels = shift; my $command = shift; my $menu_content = shift; my $menu_entry_node = shift; if (defined($menu_entry_node->{'extra'}->{'normalized'})) { my $normalized_menu_node = $menu_entry_node->{'extra'}->{'normalized'}; my $menu_node = $labels->{$normalized_menu_node}; if (!$menu_node) { $registrar->line_error($customization_information, sprintf(__("\@%s reference to nonexistent node `%s'"), $command, link_element_to_texi($menu_entry_node)), $menu_content->{'source_info'}); } else { if (!_check_node_same_texinfo_code($menu_node, $menu_entry_node->{'extra'})) { $registrar->line_warn($customization_information, sprintf(__("\@%s entry node name `%s' different from %s name `%s'"), $command, link_element_to_texi($menu_entry_node), $menu_node->{'cmdname'}, target_element_to_texi_label($menu_node)), $menu_content->{'source_info'}); } } } } # In general should be called only after complete_node_tree_with_menus # to try to generate menus automatically before checking. sub check_nodes_are_referenced { my ($registrar, $customization_information, $nodes_list, $top_node, $labels, $refs) = @_; return undef unless ($nodes_list and scalar(@{$nodes_list})); my %referenced_nodes = ($top_node => 1); foreach my $node (@{$nodes_list}) { next if (!$node->{'structure'}); # gather referenced nodes based on node pointers foreach my $direction (@node_directions) { if ($node->{'structure'}->{'node_'.$direction} and not $node->{'structure'}->{'node_'.$direction} ->{'extra'}->{'manual_content'}) { $referenced_nodes{$node->{'structure'}->{'node_'.$direction}} = 1; } } if ($node->{'structure'}->{'menu_up_hash'}) { $referenced_nodes{$node} = 1; } # If an automatic menu can be setup, consider that all # the nodes appearing in the automatic menu are referenced. # Note that the menu may not be actually setup, but # it is better not to warn for nothing. my $automatic_directions = (not ($node->{'args'} and scalar(@{$node->{'args'}}) > 1)); if ($automatic_directions and (not $node->{'extra'}->{'menus'} or not scalar(@{$node->{'extra'}->{'menus'}}))) { my @node_childs = get_node_node_childs_from_sectioning($node); foreach my $node_child (@node_childs) { $referenced_nodes{$node_child} = 1; } } } # consider nodes in @*ref commands to be referenced if (defined($refs)) { foreach my $ref (@$refs) { if ($ref->{'args'} and scalar(@{$ref->{'args'}}) and $ref->{'args'}->[0]->{'extra'} and defined($ref->{'args'}->[0]->{'extra'}->{'normalized'})) { my $normalized = $ref->{'args'}->[0]->{'extra'}->{'normalized'}; my $node_target = $labels->{$normalized}; if ($node_target) { $referenced_nodes{$node_target} = 1; } } } } foreach my $node (@{$nodes_list}) { if (not exists($referenced_nodes{$node})) { $registrar->line_warn($customization_information, sprintf(__("node `%s' unreferenced"), target_element_to_texi_label($node)), $node->{'source_info'}); } } } # set menu directions sub set_menus_node_directions($$$$$$) { my $registrar = shift; my $customization_information = shift; my $parser_information = shift; my $global_commands = shift; my $nodes_list = shift; my $labels = shift; return undef unless ($nodes_list and scalar(@{$nodes_list})); my $check_menu_entries = (!$customization_information->get_conf('novalidate') and $customization_information->get_conf('FORMAT_MENU') eq 'menu'); # First go through all the menus and set menu_up, menu_next and menu_prev, # and warn for unknown nodes. # Remark: since the @menu are only checked if they are in @node, # menu entries before the first node, or @menu nested inside # another command such as @format, may be treated slightly # differently; at least, there are no error messages for them. # foreach my $node (@{$nodes_list}) { if ($node->{'extra'}->{'menus'}) { if (@{$node->{'extra'}->{'menus'}} > 1) { foreach my $menu (@{$node->{'extra'}->{'menus'}}[1 .. $#{$node->{'extra'}->{'menus'}}]) { $registrar->line_warn($customization_information, sprintf(__("multiple \@%s"), $menu->{'cmdname'}), $menu->{'source_info'}); } } foreach my $menu (@{$node->{'extra'}->{'menus'}}) { my $previous_node; my $menu_contents = $menu; # simple menu if (scalar(@{$menu->{'contents'}}) > 0 and $menu->{'contents'}->[0]->{'type'} and $menu->{'contents'}->[0]->{'type'} eq 'preformatted') { $menu_contents = $menu->{'contents'}->[0]; } foreach my $menu_content (@{$menu_contents->{'contents'}}) { if ($menu_content->{'type'} and $menu_content->{'type'} eq 'menu_entry') { my $menu_node; foreach my $arg (@{$menu_content->{'contents'}}) { if ($arg->{'type'} eq 'menu_entry_node') { if ($arg->{'extra'}) { if (!$arg->{'extra'}->{'manual_content'}) { if ($check_menu_entries) { _check_menu_entry($registrar, $customization_information, $labels, 'menu', $menu_content, $arg); } if (defined($arg->{'extra'}->{'normalized'})) { # this may happen more than once for a given node if the node # is in more than one menu. Therefore all the menu up node # are kept in $menu_node->{'structure'}->{'menu_up_hash'} $menu_node = $labels->{$arg->{'extra'}->{'normalized'}}; if ($menu_node) { $menu_node->{'structure'} = {} if (!$menu_node->{'structure'}); $menu_node->{'structure'}->{'menu_up'} = $node; $menu_node->{'structure'}->{'menu_up_hash'} = {} if (!$menu_node->{'structure'}->{'menu_up_hash'}); $menu_node->{'structure'}->{'menu_up_hash'} ->{$node->{'extra'}->{'normalized'}} = 1; } } } else { $menu_node = $arg; } } last; } } if ($menu_node) { if ($previous_node) { if (!$menu_node->{'extra'}->{'manual_content'}) { $menu_node->{'structure'} = {} if (!$menu_node->{'structure'}); $menu_node->{'structure'}->{'menu_prev'} = $previous_node; } if (!$previous_node->{'extra'}->{'manual_content'}) { $previous_node->{'structure'} = {} if (!$previous_node->{'structure'}); $previous_node->{'structure'}->{'menu_next'} = $menu_node; } } else { $node->{'structure'} = {} if (!$node->{'structure'}); $node->{'structure'}->{'menu_child'} = $menu_node; } $previous_node = $menu_node; } } } # end menu } } } # Check @detailmenu if ($check_menu_entries) { if ($global_commands->{'detailmenu'}) { foreach my $detailmenu (@{$global_commands->{'detailmenu'}}) { foreach my $menu_content (@{$detailmenu->{'contents'}}) { if ($menu_content->{'type'} and $menu_content->{'type'} eq 'menu_entry') { foreach my $arg (@{$menu_content->{'contents'}}) { if ($arg->{'type'} eq 'menu_entry_node' and $arg->{'extra'} and !$arg->{'extra'}->{'node_manual'}) { _check_menu_entry($registrar, $customization_information, $labels, 'detailmenu', $menu_content, $arg); } } } } } } } } # determine node found through section directions, usually # from section_$direction. It could also be from # toplevel_$direction if going through parts, except for @top # as prev or next. sub _section_direction_associated_node($$) { my $section = shift; my $direction = shift; foreach my $direction_base ('section', 'toplevel') { if ($section->{'structure'}->{$direction_base.'_'.$direction} and $section->{'structure'}->{$direction_base.'_'.$direction}->{'extra'} and ($direction_base ne 'toplevel' or $direction eq 'up' or $section->{'structure'}->{$direction_base.'_'.$direction} ->{'cmdname'} ne 'top') and $section->{'structure'}->{$direction_base.'_'.$direction} ->{'extra'}->{'associated_node'}) { return $section->{'structure'}->{$direction_base.'_'.$direction} ->{'extra'}->{'associated_node'}; } } return undef; } # complete automatic directions with menus (and first node # for Top node). # Checks on structure related to menus. sub complete_node_tree_with_menus($$$$) { my $registrar = shift; my $customization_information = shift; my $nodes_list = shift; my $top_node = shift; return undef unless ($nodes_list and @{$nodes_list}); # Go through all the nodes foreach my $node (@{$nodes_list}) { my $automatic_directions = (not ($node->{'args'} and scalar(@{$node->{'args'}}) > 1)); if ($automatic_directions) { if ($node->{'extra'}->{'normalized'} ne 'Top') { foreach my $direction (@node_directions) { # prev already defined for the node first Top node menu entry if ($direction eq 'prev' and $node->{'structure'} and $node->{'structure'}->{'node_'.$direction} and $node->{'structure'}->{'node_'.$direction}->{'extra'} and $node->{'structure'}->{'node_'.$direction}->{'extra'} ->{'normalized'} and $node->{'structure'}->{'node_'.$direction}->{'extra'} ->{'normalized'} eq 'Top') { next; } if ($node->{'extra'}->{'associated_section'}) { my $section = $node->{'extra'}->{'associated_section'}; # Prefer the section associated with a @part for node directions. if ($section->{'extra'}->{'part_associated_section'}) { $section = $section->{'extra'}->{'part_associated_section'}; } # Check consistency with section and menu structure my $direction_associated_node = _section_direction_associated_node($section, $direction); if ($direction_associated_node) { if ($customization_information->get_conf( 'CHECK_NORMAL_MENU_STRUCTURE')) { if ($section->{'structure'}->{'section_up'}->{'extra'} and $section->{'structure'}->{'section_up'} ->{'extra'}->{'associated_node'} and $section->{'structure'}->{'section_up'} ->{'extra'}->{'associated_node'}->{'extra'} and $section->{'structure'}->{'section_up'} ->{'extra'}->{'associated_node'}->{'extra'}->{'menus'} and @{$section->{'structure'}->{'section_up'} ->{'extra'}->{'associated_node'}->{'extra'}->{'menus'}} and !$node->{'structure'}->{'menu_'.$direction}) { $registrar->line_warn($customization_information, sprintf(__("node %s for `%s' is `%s' in sectioning but not in menu"), $direction, target_element_to_texi_label($node), target_element_to_texi_label($direction_associated_node)), $node->{'source_info'}); } } } } # no direction was found using sections, use menus. This allows # using only automatic direction for manuals without sectioning # commands. if ($node->{'structure'} and !$node->{'structure'}->{'node_'.$direction} and $node->{'structure'}->{'menu_'.$direction} and !$node->{'structure'}->{'menu_'.$direction}->{'extra'} ->{'manual_content'}) { if ($customization_information->get_conf( 'CHECK_NORMAL_MENU_STRUCTURE') and $node->{'extra'}->{'associated_section'}) { $registrar->line_warn($customization_information, sprintf(__("node `%s' is %s for `%s' in menu but not in sectioning"), target_element_to_texi_label( $node->{'structure'}->{'menu_'.$direction}), $direction, target_element_to_texi_label($node), ), $node->{'source_info'}); } $node->{'structure'}->{'node_'.$direction} = $node->{'structure'}->{'menu_'.$direction}; } } } elsif (not $node->{'structure'} or not $node->{'structure'}->{'node_next'}) { # use first menu entry if available as next for Top if ($node->{'structure'} and $node->{'structure'}->{'menu_child'}) { $node->{'structure'}->{'node_next'} = $node->{'structure'}->{'menu_child'}; if (!$node->{'structure'}->{'menu_child'}->{'extra'}->{'manual_content'} and !$node->{'structure'}->{'menu_child'}->{'structure'}->{'node_prev'}) { $node->{'structure'}->{'menu_child'}->{'structure'}->{'node_prev'} = $node; } } else { # use the first non top node as next for Top foreach my $first_non_top_node (@{$nodes_list}) { if ($first_non_top_node ne $node) { $node->{'structure'} = {} if (! $node->{'structure'}); $node->{'structure'}->{'node_next'} = $first_non_top_node; my $first_non_top_automatic = (not ($first_non_top_node->{'args'} and scalar(@{$first_non_top_node->{'args'}}) > 1)); if ($first_non_top_automatic) { $first_non_top_node->{'structure'} = {} if (! $first_non_top_node->{'structure'}); $first_non_top_node->{'structure'}->{'node_prev'} = $node; } last; } } } } } # check consistency between node pointer and node entries menu order if ($node->{'extra'}->{'normalized'} ne 'Top') { foreach my $direction (@node_directions) { if ($customization_information->get_conf('CHECK_NORMAL_MENU_STRUCTURE') and $node->{'structure'}->{'node_'.$direction} and $node->{'structure'}->{'menu_'.$direction} and $node->{'structure'}->{'menu_'.$direction} ne $node->{'structure'}->{'node_'.$direction} and not $node->{'structure'}->{'menu_'.$direction} ->{'extra'}->{'manual_content'}) { $registrar->line_warn($customization_information, sprintf(__("node %s pointer for `%s' is `%s' but %s is `%s' in menu"), $direction, target_element_to_texi_label($node), target_element_to_texi_label( $node->{'structure'}->{'node_'.$direction}), $direction, target_element_to_texi_label( $node->{'structure'}->{'menu_'.$direction})), $node->{'source_info'}); } } } # check for node up / menu up mismatch if ($customization_information->get_conf('CHECK_MISSING_MENU_ENTRY') and $node->{'structure'}->{'node_up'} # No check if node up is an external manual and (!$node->{'structure'}->{'node_up'}->{'extra'}->{'manual_content'}) and (!$node->{'structure'}->{'menu_up_hash'} or !$node->{'structure'}->{'menu_up_hash'}->{$node->{'structure'} ->{'node_up'}->{'extra'}->{'normalized'}})) { # check if up node has a menu if ($node->{'structure'}->{'node_up'}->{'extra'}->{'menus'} and @{$node->{'structure'}->{'node_up'}->{'extra'}->{'menus'}}) { $registrar->line_warn($customization_information, sprintf( __("node `%s' lacks menu item for `%s' despite being its Up target"), target_element_to_texi_label($node->{'structure'}->{'node_up'}), target_element_to_texi_label($node)), $node->{'structure'}->{'node_up'}->{'source_info'}); } # FIXME check that the menu_up_hash is not empty (except for Top)? # FIXME check that node_up is not an external node (except for Top)? } } } # set node directions based on sectioning and @node explicit directions sub nodes_tree($$$$$) { my $registrar = shift; my $customization_information = shift; my $parser_information = shift; my $nodes_list = shift; my $labels = shift; return undef unless ($nodes_list and @{$nodes_list}); my $top_node; # Go through all the nodes and set directions. foreach my $node (@{$nodes_list}) { if ($node->{'extra'}->{'normalized'} eq 'Top') { $top_node = $node; } my $automatic_directions = (not ($node->{'args'} and scalar(@{$node->{'args'}}) > 1)); if ($automatic_directions) { if ($node->{'extra'}->{'normalized'} ne 'Top') { foreach my $direction (@node_directions) { # prev already defined for the node first Top node menu entry if ($direction eq 'prev' and $node->{'node_'.$direction} and $node->{'structure'}->{'node_'.$direction}->{'extra'} and $node->{'structure'}->{'node_'.$direction}->{'extra'} ->{'normalized'} and $node->{'structure'}->{'node_'.$direction}->{'extra'} ->{'normalized'} eq 'Top') { next; } if ($node->{'extra'}->{'associated_section'}) { my $section = $node->{'extra'}->{'associated_section'}; # Prefer the section associated with a @part for node directions. if ($section->{'extra'}->{'part_associated_section'}) { $section = $section->{'extra'}->{'part_associated_section'}; } my $direction_associated_node = _section_direction_associated_node($section, $direction); if ($direction_associated_node) { $node->{'structure'} = {} if (!$node->{'structure'}); $node->{'structure'}->{'node_'.$direction} = $direction_associated_node; } } } } else { # Special case for Top node, use first section if ($node->{'extra'}->{'associated_section'} and $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'} and $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}->[0] and $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}->[0]->{'extra'} and $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}->[0] ->{'extra'}->{'associated_node'}) { my $top_node_section_child = $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}->[0] ->{'extra'}->{'associated_node'}; $node->{'structure'} = {} if (! $node->{'structure'}); $node->{'structure'}->{'node_next'} = $top_node_section_child; my $first_section_child_automatic = (not ($top_node_section_child->{'args'} and scalar(@{$top_node_section_child->{'args'}}) > 1)); if ($first_section_child_automatic) { $top_node_section_child->{'structure'} = {} if (! $top_node_section_child->{'structure'}); $top_node_section_child->{'structure'}->{'node_prev'} = $node; } } } } else { # explicit directions for (my $i = 1; $i < scalar(@{$node->{'args'}}); $i++) { my $direction_element = $node->{'args'}->[$i]; next if (!$direction_element->{'extra'}); my $node_direction = $direction_element->{'extra'}; my $direction = $node_directions[$i-1]; $node->{'structure'} = {} if (! defined($node->{'structure'})); # external node if ($node_direction->{'manual_content'}) { $node->{'structure'}->{'node_'.$direction} = { 'extra' => $node_direction }; } elsif (defined($node_direction->{'normalized'})) { if ($labels->{$node_direction->{'normalized'}}) { my $node_target = $labels->{$node_direction->{'normalized'}}; $node->{'structure'}->{'node_'.$direction} = $node_target; if (!$customization_information->get_conf('novalidate') and !_check_node_same_texinfo_code($node_target, $node_direction)) { $registrar->line_warn($customization_information, sprintf( __("%s pointer `%s' (for node `%s') different from %s name `%s'"), $direction_texts{$direction}, link_element_to_texi($direction_element), target_element_to_texi_label($node), $node_target->{'cmdname'}, target_element_to_texi_label($node_target)), $node->{'source_info'}); } } else { if ($customization_information->get_conf('novalidate')) { $node->{'structure'}->{'node_'.$direction} = { 'extra' => $node_direction }; } else { $registrar->line_error($customization_information, sprintf(__("%s reference to nonexistent `%s'"), $direction_texts{$direction}, link_element_to_texi($direction_element)), $node->{'source_info'}); } } } } } } $top_node = $nodes_list->[0] if (!$top_node); return $top_node; } sub section_level_adjusted_command_name($) { my $element = shift; my $heading_level = $element->{'structure'}->{'section_level'}; my $command; if ($heading_level ne $Texinfo::Common::command_structuring_level{ $element->{'cmdname'}}) { $command = $Texinfo::Common::level_to_structuring_command{$element->{'cmdname'}} ->[$heading_level]; } else { $command = $element->{'cmdname'}; } return $command; } # Return a list of tree units to be converted into pages. Each tree unit # starts with a @node as its first child (except possibly the first one). # It is important that this function reassociates all the root commands # such that the result does not depend on the previous association (if any). sub split_by_node($) { my $root = shift; my $tree_units; my $current = { 'type' => 'unit' }; push @$tree_units, $current; my @pending_parts = (); foreach my $content (@{$root->{'contents'}}) { if ($content->{'cmdname'} and $content->{'cmdname'} eq 'part') { push @pending_parts, $content; next; } if ($content->{'cmdname'} and $content->{'cmdname'} eq 'node') { if (not $current->{'extra'} or not $current->{'extra'}->{'unit_command'}) { $current->{'extra'} = {} if (! $current->{'extra'}); $current->{'extra'}->{'unit_command'} = $content; } else { $current = { 'type' => 'unit', 'extra' => {'unit_command' => $content}, 'structure' => {'unit_prev' => $tree_units->[-1]}}; $tree_units->[-1]->{'structure'} = {} if (! $tree_units->[-1]->{'structure'}); $tree_units->[-1]->{'structure'}->{'unit_next'} = $current; push @$tree_units, $current; } } if (@pending_parts) { foreach my $part (@pending_parts) { push @{$current->{'contents'}}, $part; $part->{'structure'}->{'associated_unit'} = $current; } @pending_parts = (); } push @{$current->{'contents'}}, $content; #if (defined($content->{'structure'}) # and defined($content->{'structure'}->{'associated_unit'})) { # print STDERR "Resetting node associated_unit for $content\n"; #} $content->{'structure'} = {} if (! $content->{'structure'}); $content->{'structure'}->{'associated_unit'} = $current; } if (@pending_parts) { foreach my $part (@pending_parts) { push @{$current->{'contents'}}, $part; $part->{'structure'}->{'associated_unit'} = $current; } @pending_parts = (); } return $tree_units; } # Return a list of tree units to be converted into pages. Each tree unit # starts with the @node associated with a sectioning command or with the # sectioning command if there is no associated node. # It is important that this function reassociates all the root commands # such that the result does not depend on the previous association (if any). sub split_by_section($) { my $root = shift; my $tree_units; my $current = { 'type' => 'unit' }; push @$tree_units, $current; foreach my $content (@{$root->{'contents'}}) { if ($content->{'cmdname'} and (($content->{'cmdname'} eq 'node' and $content->{'extra'} and $content->{'extra'}->{'associated_section'}) or ($content->{'cmdname'} eq 'part' and $content->{'extra'} and $content->{'extra'}->{'part_associated_section'}))) { my $new_section; if ($content->{'cmdname'} eq 'node') { $new_section = $content->{'extra'}->{'associated_section'}; } else { $new_section = $content->{'extra'}->{'part_associated_section'}; } if (not defined($current->{'extra'}) or not defined($current->{'extra'}->{'unit_command'})) { $current->{'extra'} = {} if (!$current->{'extra'}); $current->{'extra'}->{'unit_command'} = $new_section; } elsif (!$current->{'extra'}->{'unit_command'} or $new_section ne $current->{'extra'}->{'unit_command'}) { $current = { 'type' => 'unit', 'extra' => {'unit_command' => $new_section}, 'structure' => {'unit_prev' => $tree_units->[-1]}}; $tree_units->[-1]->{'structure'} = {} if (! $tree_units->[-1]->{'structure'}); $tree_units->[-1]->{'structure'}->{'unit_next'} = $current; push @$tree_units, $current; } } elsif ($content->{'cmdname'} and $content->{'cmdname'} ne 'node' and $Texinfo::Commands::root_commands{$content->{'cmdname'}}) { if (not defined($current->{'extra'}) or not defined($current->{'extra'}->{'unit_command'})) { $current->{'extra'} = {} if (! $current->{'extra'}); $current->{'extra'}->{'unit_command'} = $content; } elsif ($current->{'extra'}->{'unit_command'} ne $content) { $current = {'type' => 'unit', 'extra' => {'unit_command' => $content}, 'structure' => {'unit_prev' => $tree_units->[-1]}}; $tree_units->[-1]->{'structure'} = {} if (! $tree_units->[-1]->{'structure'}); $tree_units->[-1]->{'structure'}->{'unit_next'} = $current; push @$tree_units, $current; } } push @{$current->{'contents'}}, $content; #if (defined($content->{'structure'}) # and defined($content->{'structure'}->{'associated_unit'})) { # print STDERR "Resetting section associated_unit for $content\n"; #} $content->{'structure'} = {} if (! $content->{'structure'}); $content->{'structure'}->{'associated_unit'} = $current; } return $tree_units; } # Associate top-level units with pages according to the splitting # specification. Set 'first_in_page' on each unit to the unit # that is the first in the output page. sub split_pages ($$) { my $tree_units = shift; my $split = shift; return undef if (!$tree_units or !@$tree_units); my $split_level; if (!$split) { foreach my $tree_unit (@$tree_units) { $tree_unit->{'extra'} = {} if (! $tree_unit->{'extra'}); $tree_unit->{'extra'}->{'first_in_page'} = $tree_units->[0]; } return; } elsif ($split eq 'chapter') { $split_level = 1; } elsif ($split eq 'section') { $split_level = 2; } elsif ($split ne 'node') { warn "Unknown split specification: $split\n"; } my $current_first_in_page; foreach my $tree_unit (@$tree_units) { my $level; my $section = _tree_unit_section($tree_unit); if (defined($section)) { $level = $section->{'structure'}->{'section_level'}; } #print STDERR "level($split_level) $level " # .root_or_external_element_cmd_texi($tree_unit)."\n"; if (!defined($split_level) or (defined($level) and $split_level >= $level) or !$current_first_in_page) { $current_first_in_page = $tree_unit; } $tree_unit->{'extra'} = {} if (! $tree_unit->{'extra'}); $tree_unit->{'extra'}->{'first_in_page'} = $current_first_in_page; } } # Returns a unit element associated to a label that can be used to setup a target # to the label. If the target is an external node, create such element here, # if it is a node return the parent element that is supposed to be the # target for links to the node. Otherwise there is no such element (yet), # for floats and anchor, return undef. sub _label_target_unit_element($) { my $label = shift; if ($label->{'extra'} and $label->{'extra'}->{'manual_content'}) { my $external_node = { 'type' => 'external_node', 'extra' => {'manual_content' => $label->{'extra'}->{'manual_content'}}}; if ($label->{'extra'}->{'node_content'}) { $external_node->{'extra'}->{'node_content'} = $label->{'extra'}->{'node_content'}; $external_node->{'extra'}->{'normalized'} = Texinfo::Convert::NodeNameNormalization::normalize_node( {'contents' => $label->{'extra'}->{'node_content'}}); } return $external_node; } elsif ($label->{'cmdname'} and $label->{'cmdname'} eq 'node') { return $label->{'structure'}->{'associated_unit'}; } else { # case of a @float or an @anchor, no target element defined at this stage return undef; } } sub _tree_unit_section($) { my $tree_unit = shift; if (not defined($tree_unit->{'extra'}) or not defined($tree_unit->{'extra'}->{'unit_command'})) { return undef; } my $tree_unit_command_element = $tree_unit->{'extra'}->{'unit_command'}; if ($tree_unit_command_element->{'cmdname'} eq 'node') { if ($tree_unit_command_element->{'extra'} and $tree_unit_command_element->{'extra'}->{'associated_section'}) { return $tree_unit_command_element->{'extra'}->{'associated_section'}; } else { return undef; } } else { return $tree_unit_command_element; } } sub _tree_unit_node($) { my $tree_unit = shift; if (not defined($tree_unit->{'extra'}) or not defined($tree_unit->{'extra'}->{'unit_command'})) { return undef; } my $tree_unit_command_element = $tree_unit->{'extra'}->{'unit_command'}; if ($tree_unit_command_element->{'cmdname'} eq 'node') { return $tree_unit_command_element; } else { if ($tree_unit_command_element->{'extra'} and $tree_unit_command_element->{'extra'}->{'associated_node'}) { return $tree_unit_command_element->{'extra'}->{'associated_node'} } else { return undef; } } } # Do element directions (like in texi2html) and store them # in 'structure->'directions'. # The directions are only created if pointing to other 'unit' elements. # In practice there are only tree unit passed to the function, but # other root elements could probably be used, in theory. sub elements_directions($$$) { my $customization_information = shift; my $labels = shift; my $tree_units = shift; return if (!$tree_units or !@$tree_units); my $node_top = $labels->{'Top'}; foreach my $tree_unit (@$tree_units) { my $directions = {}; $directions->{'This'} = $tree_unit; $directions->{'Forward'} = $tree_unit->{'structure'}->{'unit_next'} if ($tree_unit->{'structure'} and $tree_unit->{'structure'}->{'unit_next'} and defined($tree_unit->{'structure'}->{'unit_next'}->{'type'}) and $tree_unit->{'structure'}->{'unit_next'}->{'type'} eq 'unit'); $directions->{'Back'} = $tree_unit->{'structure'}->{'unit_prev'} if ($tree_unit->{'structure'} and $tree_unit->{'structure'}->{'unit_prev'} and defined($tree_unit->{'structure'}->{'unit_prev'}->{'type'}) and $tree_unit->{'structure'}->{'unit_prev'}->{'type'} eq 'unit'); my $node = _tree_unit_node($tree_unit); if (defined($node)) { foreach my $direction(['NodeUp', 'node_up'], ['NodeNext', 'node_next'], ['NodePrev', 'node_prev']) { $directions->{$direction->[0]} = _label_target_unit_element($node->{'structure'}->{$direction->[1]}) if ($node->{'structure'}->{$direction->[1]}); } # Now do NodeForward which is something like the following node. my $automatic_directions = (not ($node->{'args'} and scalar(@{$node->{'args'}}) > 1)); if ($node->{'structure'}->{'menu_child'}) { $directions->{'NodeForward'} = _label_target_unit_element($node->{'structure'}->{'menu_child'}); } elsif ($automatic_directions and $node->{'associated_section'} and $node->{'associated_section'}->{'structure'}->{'section_childs'} and $node->{'associated_section'}->{'structure'}->{'section_childs'}->[0]) { $directions->{'NodeForward'} = $node->{'associated_section'}->{'structure'} ->{'section_childs'}->[0]->{'structure'}->{'associated_unit'}; } elsif ($node->{'structure'}->{'node_next'}) { $directions->{'NodeForward'} = _label_target_unit_element($node->{'structure'}->{'node_next'}); } elsif ($node->{'structure'}->{'node_up'}) { my $up = $node->{'structure'}->{'node_up'}; my @up_list = ($node); # the condition with the up_list avoids infinite loops # the last condition stops when the Top node is reached. while (not (grep {$up eq $_} @up_list or ($node_top and $up eq $node_top))) { if ($up->{'structure'} and defined($up->{'structure'}->{'node_next'})) { $directions->{'NodeForward'} = _label_target_unit_element($up->{'structure'}->{'node_next'}); last; } push @up_list, $up; last if (not $up->{'structure'} or not $up->{'structure'}->{'node_up'}); $up = $up->{'structure'}->{'node_up'}; } } if ($directions->{'NodeForward'} and $directions->{'NodeForward'}->{'type'} eq 'unit' and (!$directions->{'NodeForward'}->{'structure'} or !$directions->{'NodeForward'}->{'structure'}->{'directions'} or !$directions->{'NodeForward'}->{'structure'}->{'directions'} ->{'NodeBack'})) { $directions->{'NodeForward'}->{'structure'} = {} if (! $directions->{'NodeForward'}->{'structure'}); $directions->{'NodeForward'}->{'structure'}->{'directions'} = {} if (! $directions->{'NodeForward'}->{'structure'}->{'directions'}); $directions->{'NodeForward'}->{'structure'}->{'directions'} ->{'NodeBack'} = $tree_unit; } } my $section = _tree_unit_section($tree_unit); if (not defined($section)) { # If there is no associated section, find the previous element section. # Use the FastForward of this element. # Use it as FastBack if the section is top level, or use the FastBack. my $section_element; my $current = $tree_unit; while ($current->{'structure'} and $current->{'structure'}->{'unit_prev'}) { $current = $current->{'structure'}->{'unit_prev'}; $section = _tree_unit_section($current); if (defined($section)) { $section_element = $current; last; } } if ($section_element) { if ($section_element->{'structure'}->{'directions'}->{'FastForward'}) { $directions->{'FastForward'} = $section_element->{'structure'}->{'directions'}->{'FastForward'}; } if ($section->{'structure'}->{'section_level'} <= 1) { $directions->{'FastBack'} = $section_element; } elsif ($section_element->{'structure'}->{'directions'}->{'Fastback'}) { $directions->{'FastBack'} = $section_element->{'structure'}->{'directions'}->{'Fastback'}; } } } else { foreach my $direction(['Up', 'section_up'], ['Next', 'section_next'], ['Prev', 'section_prev']) { # in most cases $section->{$direction->[1]} # ->{'structure'}->{'associated_unit'} is defined # but it may not be the case for the up of @top. # The section may be its own up in cases like # @part part # @chapter chapter # in that cas the direction is not set up $directions->{$direction->[0]} = $section->{'structure'}->{$direction->[1]} ->{'structure'}->{'associated_unit'} if ($section->{'structure'}->{$direction->[1]} and $section->{'structure'}->{$direction->[1]} ->{'structure'}->{'associated_unit'} and $section->{'structure'}->{$direction->[1]} ->{'structure'}->{'associated_unit'} ne $section->{'structure'}->{'associated_unit'}); } my $up = $section; while ($up->{'structure'}->{'section_level'} > 1 and $up->{'structure'}->{'section_up'}) { $up = $up->{'structure'}->{'section_up'}; } # fastforward is the next element on same level than the upper parent # element. if ($up->{'structure'}->{'section_level'} < 1 and $up->{'cmdname'} and $up->{'cmdname'} eq 'top' and $up->{'structure'}->{'section_childs'} and @{$up->{'structure'}->{'section_childs'}}) { $directions->{'FastForward'} = $up->{'structure'}->{'section_childs'}->[0] ->{'structure'}->{'associated_unit'}; } elsif ($up->{'structure'}->{'toplevel_next'}) { $directions->{'FastForward'} = $up->{'structure'}->{'toplevel_next'} ->{'structure'}->{'associated_unit'}; } elsif ($up->{'structure'}->{'section_next'}) { $directions->{'FastForward'} = $up->{'structure'}->{'section_next'} ->{'structure'}->{'associated_unit'}; } # if the element isn't at the highest level, fastback is the # highest parent element if ($up and $up ne $section) { $directions->{'FastBack'} = $up->{'structure'}->{'associated_unit'}; } elsif ($section->{'structure'}->{'section_level'} <= 1 and $directions->{'FastForward'}) { # the element is a top level element, we adjust the next # toplevel element fastback $directions->{'FastForward'}->{'structure'} = {} if (! $directions->{'FastForward'}->{'structure'}); $directions->{'FastForward'}->{'structure'}->{'directions'} = {} if (! $directions->{'FastForward'}->{'structure'}->{'directions'}); $directions->{'FastForward'}->{'structure'}->{'directions'}->{'FastBack'} = $tree_unit if ($directions and $directions->{'FastForward'}); } } # Use node up for Up if there is no section up. # Not done in the default case. if ($customization_information->get_conf('USE_UP_NODE_FOR_ELEMENT_UP') and !$directions->{'Up'} and defined($node) and $node->{'structure'}->{'node_up'} and (!$node_top or ($node ne $node_top))) { #print STDERR "Node for up: ".root_or_external_element_cmd_texi($tree_unit)."\n"; my $up_node_element = _label_target_unit_element($node->{'structure'}->{'node_up'}); $directions->{'Up'} = $up_node_element if ($up_node_element); } $tree_unit->{'structure'} = {} if (! $tree_unit->{'structure'}); if ($tree_unit->{'structure'}->{'directions'}) { %{$tree_unit->{'structure'}->{'directions'}} = (%{$tree_unit->{'structure'}->{'directions'}}, %$directions); } else { $tree_unit->{'structure'}->{'directions'} = $directions; } } if ($customization_information->get_conf('DEBUG')) { foreach my $tree_unit (@$tree_units) { print STDERR 'Directions' # uncomment to show the perl object name # . "($tree_unit)" . ': '.print_element_directions($tree_unit)."\n"; } } } # for now, the elements can only be tree units. It could probably # be other root elements for which file directions make sense, in theory. sub elements_file_directions($) { my $tree_units = shift; return if (!$tree_units or !@$tree_units); my $current_filename; my $first_element_in_file; # need to gather the directions before the FirstInFile* directions # are added to the first element in the file. my @first_element_in_file_directions; foreach my $tree_unit (@$tree_units) { my $directions; my $filename; if (defined($tree_unit->{'structure'}->{'unit_filename'})) { $filename = $tree_unit->{'structure'}->{'unit_filename'}; my $current_tree_unit = $tree_unit; if (not defined($current_filename) or $filename ne $current_filename) { $first_element_in_file = $tree_unit; @first_element_in_file_directions = keys %{$tree_unit->{'structure'}->{'directions'}}; $current_filename = $filename; } while ($current_tree_unit->{'structure'}->{'unit_prev'}) { $current_tree_unit = $current_tree_unit->{'structure'}->{'unit_prev'}; if (defined($current_tree_unit->{'structure'}->{'unit_filename'})) { if ($current_tree_unit->{'structure'}->{'unit_filename'} ne $filename) { $tree_unit->{'structure'}->{'directions'}->{'PrevFile'} = $current_tree_unit; last; } } else { last; } } $current_tree_unit = $tree_unit; while ($current_tree_unit->{'structure'}->{'unit_next'}) { $current_tree_unit = $current_tree_unit->{'structure'}->{'unit_next'}; if (defined($current_tree_unit->{'structure'}->{'unit_filename'})) { if ($current_tree_unit->{'structure'}->{'unit_filename'} ne $filename) { $tree_unit->{'structure'}->{'directions'}->{'NextFile'} = $current_tree_unit; last; } } else { last; } } } # set the directions of the first elements in file, to # be used in footers for example if (defined($first_element_in_file)) { foreach my $first_in_file_direction (@first_element_in_file_directions) { $tree_unit->{'structure'}->{'directions'} ->{'FirstInFile'.$first_in_file_direction} = $first_element_in_file->{'structure'}->{'directions'} ->{$first_in_file_direction}; } } } } # used in debug messages sub root_or_external_element_cmd_texi($) { my $element = shift; if (!$element) { return "UNDEF ELEMENT"; } if (!$element->{'type'}) { return "element $element without type: ". Texinfo::Common::debug_print_element_details($element, 1); } if ($element->{'type'} eq 'external_node') { my $command = {'contents' => [{'text' => '('}, @{$element->{'extra'}->{'manual_content'}}, {'text' => ')'}]}; if ($element->{'extra'}->{'node_content'}) { unshift @{$command->{'contents'}}, @{$element->{'extra'}->{'node_content'}}; } return Texinfo::Convert::Texinfo::convert_to_texinfo($command); } my $command_element; if ($element->{'extra'} and $element->{'extra'}->{'unit_command'}) { $command_element = $element->{'extra'}->{'unit_command'}; } else { # happens when there are only nodes and sections are used as elements my $result = "No associated command "; $result .= "(type $element->{'type'})" if (defined($element->{'type'})); return $result; } return Texinfo::Convert::Texinfo::root_heading_command_to_texinfo($command_element); } # Used for debugging and in test suite, but not generally useful. Not # documented in pod section and not exportable as it should not, in # general, be used. # In general would be called with tree unit elements, but is more generic # to account for other situations. sub print_element_directions($) { my $element = shift; my $result = 'element: '.root_or_external_element_cmd_texi($element)."\n"; if ($element->{'extra'} and $element->{'structure'}->{'directions'}) { foreach my $direction (sort(keys(%{$element->{'structure'}->{'directions'}}))) { $result .= " $direction: ". root_or_external_element_cmd_texi( $element->{'structure'}->{'directions'}->{$direction})."\n"; } } else { $result .= " NO DIRECTION"; } return $result; } # For each internal reference command, set the 'normalized' key, in the # @*ref first argument or in 'menu_entry_node' extra. sub associate_internal_references($$$$$) { my $registrar = shift; my $customization_information = shift; my $parser_information = shift; my $labels = shift; my $refs = shift; return if (!defined($refs)); foreach my $ref (@$refs) { my $label_element; if ($ref->{'type'} and $ref->{'type'} eq 'menu_entry_node') { $label_element = $ref; } else { $label_element = $ref->{'args'}->[0]; } if ($label_element->{'extra'} and $label_element->{'extra'}->{'node_content'}) { my $normalized = Texinfo::Convert::NodeNameNormalization::normalize_node( {'contents' => $label_element->{'extra'}->{'node_content'} }); $label_element->{'extra'}->{'normalized'} = $normalized if (defined($normalized) and $normalized ne ''); } if ($ref->{'type'} and $ref->{'type'} eq 'menu_entry_node') { # similar messages are output in _check_menu_entry next; } elsif ($label_element->{'extra'}) { my $label_info = $label_element->{'extra'}; if (!defined($label_info->{'normalized'}) or !defined($labels->{$label_info->{'normalized'}})) { if (!$customization_information->get_conf('novalidate')) { $registrar->line_error($customization_information, sprintf(__("\@%s reference to nonexistent node `%s'"), $ref->{'cmdname'}, link_element_to_texi($label_element)), $ref->{'source_info'}); } } else { my $node_target = $labels->{$label_info->{'normalized'}}; if (!$customization_information->get_conf('novalidate') and !_check_node_same_texinfo_code($node_target, $label_info)) { $registrar->line_warn($customization_information, sprintf(__("\@%s to `%s', different from %s name `%s'"), $ref->{'cmdname'}, link_element_to_texi($label_element), $node_target->{'cmdname'}, target_element_to_texi_label($node_target)), $ref->{'source_info'}); } } } } } sub number_floats($) { my $floats = shift; return if (!defined($floats)); foreach my $style (keys(%$floats)) { my %nr_in_chapter; my $float_index = 0; foreach my $float (@{$floats->{$style}}) { next if (!$float->{'extra'} or !defined($float->{'extra'}->{'normalized'})); $float_index++; my $number; if ($float->{'extra'}->{'float_section'}) { my $up = $float->{'extra'}->{'float_section'}; while ($up->{'structure'}->{'section_up'} and defined($up->{'structure'}->{'section_up'}->{'cmdname'}) and $command_structuring_level{ $up->{'structure'}->{'section_up'}->{'cmdname'}}) { $up = $up->{'structure'}->{'section_up'}; } if (!$unnumbered_commands{$up->{'cmdname'}}) { $nr_in_chapter{$up->{'structure'}->{'section_number'}}++; $number = $up->{'structure'}->{'section_number'} . '.' . $nr_in_chapter{$up->{'structure'}->{'section_number'}}; } } $number = $float_index if (!defined($number)); $float->{'structure'} = {} if (! $float->{'structure'}); $float->{'structure'}->{'float_number'} = $number; } } } sub get_node_node_childs_from_sectioning { my ($node) = @_; my @node_childs; if ($node->{'extra'} and $node->{'extra'}->{'associated_section'} and $node->{'extra'}->{'associated_section'}->{'structure'} and $node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}) { foreach my $child (@{$node->{'extra'}->{'associated_section'} ->{'structure'}->{'section_childs'}}) { if ($child->{'extra'} and $child->{'extra'}->{'associated_node'}) { push @node_childs, $child->{'extra'}->{'associated_node'}; } } } # Special case for @top. Gather all the children of the @part following # @top. if ($node->{'extra'} and $node->{'extra'}->{'associated_section'} and $node->{'extra'}->{'associated_section'}->{'cmdname'} eq 'top') { my $current = $node->{'extra'}->{'associated_section'}; while ($current->{'structure'} and $current->{'structure'}->{'section_next'}) { $current = $current->{'structure'}->{'section_next'}; if ($current->{'cmdname'} and $current->{'cmdname'} eq 'part' and $current->{'structure'}->{'section_childs'}) { foreach my $child (@{$current->{'structure'}->{'section_childs'}}) { if ($child->{'extra'} and $child->{'extra'}->{'associated_node'}) { push @node_childs, $child->{'extra'}->{'associated_node'}; } } } elsif ($current->{'extra'} and $current->{'extra'}->{'associated_node'}) { # for @appendix, and what follows, as it stops a @part, but is # not below @top push @node_childs, $current->{'extra'}->{'associated_node'}; } } } return @node_childs; } # returns the texinfo tree corresponding to a single menu entry pointing # to $NODE. # if $USE_SECTIONS is set, use the section name as menu entry name. sub new_node_menu_entry { my ($node, $use_sections) = @_; my $node_contents; if ($node->{'extra'} and defined($node->{'extra'}->{'normalized'})) { $node_contents = $node->{'args'}->[0]->{'contents'}; } # can happen with node without argument or with empty argument return undef if (not defined($node_contents)); my ($name_contents, $menu_entry_name); if ($use_sections) { if (defined $node->{'extra'}->{'associated_section'}) { $name_contents = $node->{'extra'}->{'associated_section'}->{'args'}->[0]->{'contents'}; } else { $name_contents = $node_contents; # shouldn't happen } } if ($use_sections) { $menu_entry_name = {'type' => 'menu_entry_name'}; $menu_entry_name->{'contents'} = Texinfo::Common::copy_contents($name_contents); foreach my $content (@{$menu_entry_name->{'contents'}}) { $content->{'parent'} = $menu_entry_name; } # colons could be doubly protected, but it is probably better # than not protected at all. Texinfo::Common::protect_colon_in_tree($menu_entry_name); } my $entry = {'type' => 'menu_entry', 'extra' => {}}; my $menu_entry_node = {'type' => 'menu_entry_node'}; $menu_entry_node->{'contents'} = Texinfo::Common::copy_contents($node_contents); foreach my $content (@{$menu_entry_node->{'contents'}}) { $content->{'parent'} = $menu_entry_node; } # do not protect here, as it could already be protected, and # the menu entry should be the same as the node #Texinfo::Common::protect_colon_in_tree($menu_entry_node); my $description = {'type' => 'menu_entry_description', 'contents' => []}; $description->{'contents'}->[0] = {'type' => 'preformatted', 'parent' => $description, 'contents' => []}; $description->{'contents'}->[0]->{'contents'}->[0] = {'text' => "\n", 'parent' => $description->{'contents'}->[0]}; if ($use_sections) { $entry->{'contents'} = [{'text' => '* ', 'type' => 'menu_entry_leading_text'}, $menu_entry_name, {'text' => ': ', 'type' => 'menu_entry_separator'}, $menu_entry_node, {'text' => '.', 'type' => 'menu_entry_separator'}, $description]; } else { $entry->{'contents'} = [{'text' => '* ', 'type' => 'menu_entry_leading_text'}, $menu_entry_node, {'text' => '::', 'type' => 'menu_entry_separator'}, $description]; } foreach my $arg(@{$entry->{'contents'}}) { $arg->{'parent'} = $entry; } my $parsed_node_manual = Texinfo::Common::parse_node_manual($menu_entry_node, 1); if (defined($parsed_node_manual)) { foreach my $label_info (keys(%$parsed_node_manual)) { $menu_entry_node->{'extra'} = {} if (!$menu_entry_node->{'extra'}); $menu_entry_node->{'extra'}->{$label_info} = [@{$parsed_node_manual->{$label_info}}]; if ($label_info eq 'node_content') { $menu_entry_node->{'extra'}->{'normalized'} = Texinfo::Convert::NodeNameNormalization::normalize_node( {'contents' => $menu_entry_node->{'extra'}->{$label_info}}); } } } return $entry; } sub new_block_command($$$) { my $block_contents = shift; my $parent = shift; my $command_name = shift; my $new_block = {'cmdname' => $command_name, 'parent' => $parent}; $new_block->{'args'} = [{'type' => 'block_line_arg', 'parent' => $new_block, 'info' => { 'spaces_after_argument' => {'text' => "\n",}}}]; foreach my $content (@$block_contents) { confess("new_block_command: undef \$block_contents content") if (!defined($content)); $content->{'parent'} = $new_block; } my $end = {'cmdname' => 'end', 'parent' => $new_block, 'extra' => {'text_arg' => $command_name}}; $end->{'info'} = {'spaces_before_argument' => {'text' => ' '}}; $end->{'args'} = [{'type' => 'line_arg', 'parent' => $end, 'info' => {'spaces_after_argument' => {'text' => "\n"}}}]; push @{$end->{'args'}->[0]->{'contents'}}, {'text' => $command_name, 'parent' => $end->{'args'}->[0]}; $new_block->{'contents'} = [@$block_contents, $end]; return $new_block; } sub new_complete_node_menu { my ($node, $use_sections) = @_; my @node_childs = get_node_node_childs_from_sectioning($node); if (not scalar(@node_childs)) { return; } my @pending; foreach my $child (@node_childs) { my $entry = new_node_menu_entry($child, $use_sections); push @pending, $entry if defined($entry); } my $section = $node->{'extra'}->{'associated_section'}; my $new_menu = new_block_command(\@pending, $section, 'menu'); return $new_menu; } sub _sort_string($$) { my $a = shift; my $b = shift; return (($a =~ /^[[:alpha:]]/ and $b =~ /^[[:alpha:]]/) or ($a !~ /^[[:alpha:]]/ and $b !~ /^[[:alpha:]]/)) ? ($a cmp $b) : (($a =~ /^[[:alpha:]]/ && 1) || -1); } # Return ($NORMALIZED_ENTRY_NODE, $NODE) where $NODE is the node referred to # by menu entry $ENTRY, and $NORMALIZED_ENTRY_NODE is the name of this node. sub _normalized_entry_associated_internal_node($;$) { my $entry = shift; my $labels = shift; foreach my $arg (@{$entry->{'contents'}}) { if ($arg->{'type'} eq 'menu_entry_node') { if (! $arg->{'extra'}->{'manual_content'}) { my $normalized_entry_node = $arg->{'extra'}->{'normalized'}; if (defined($normalized_entry_node)) { if ($labels) { return ($normalized_entry_node, $labels->{$normalized_entry_node}); } else { return ($normalized_entry_node, undef); } } } last; } } return (undef, undef); } # used in Plaintext converter and tree transformations sub new_master_menu($$$) { my $self = shift; my $labels = shift; my $menus = shift; my @master_menu_contents; if (defined($menus) and @$menus) { foreach my $menu (@$menus) { foreach my $entry (@{$menu->{'contents'}}) { if ($entry->{'type'} and $entry->{'type'} eq 'menu_entry') { my ($normalized_entry_node, $node) = _normalized_entry_associated_internal_node($entry, $labels); if (defined($node) and $node->{'extra'}) { push @master_menu_contents, _print_down_menus($node, $labels); } } } } } if (scalar(@master_menu_contents)) { my $first_preformatted = $master_menu_contents[0]->{'contents'}->[0]; my $master_menu_title = Texinfo::Translations::gdt($self, ' --- The Detailed Node Listing ---'); my @master_menu_title_contents; foreach my $content (@{$master_menu_title->{'contents'}}, {'text' => "\n"}) { $content->{'parent'} = $first_preformatted; push @master_menu_title_contents, $content; } unshift @{$first_preformatted->{'contents'}}, @master_menu_title_contents; return Texinfo::Structuring::new_block_command(\@master_menu_contents, undef, 'detailmenu'); } else { return undef; } } # TODO document sub new_complete_menu_master_menu($$$) { my $self = shift; my $labels = shift; my $node = shift; my $menu_node = new_complete_node_menu($node); if ($node->{'extra'}->{'normalized'} eq 'Top' and $node->{'extra'}->{'associated_section'} and $node->{'extra'}->{'associated_section'}->{'cmdname'} eq 'top' and $menu_node) { my $detailmenu = new_master_menu($self, $labels, [$menu_node]); if ($detailmenu) { # add a blank line before the detailed node listing my $menu_comment = {'type' => 'menu_comment', 'parent' => $menu_node}; push @{$menu_node->{'contents'}}, $menu_comment; my $preformatted = {'type' => 'preformatted', 'parent' => $menu_comment}; push @{$menu_comment->{'contents'}}, $preformatted; my $empty_line = {'type' => 'after_menu_description_line', 'text' => "\n", 'parent' => $preformatted}; push @{$preformatted->{'contents'}}, $empty_line; $detailmenu->{'parent'} = $menu_node; push @{$menu_node->{'contents'}}, $detailmenu; } } return $menu_node; } sub _print_down_menus($$); sub _print_down_menus($$) { my $node = shift; my $labels = shift; my @menus; if ($node->{'extra'}->{'menus'} and scalar(@{$node->{'extra'}->{'menus'}})) { @menus = @{$node->{'extra'}->{'menus'}}; } else { my $section = $node->{'extra'}->{'associated_section'}; my $current_menu = Texinfo::Structuring::new_complete_node_menu($node, undef); if (defined($current_menu)) { @menus = ( $current_menu ); } } my @master_menu_contents; if (@menus) { my @node_children; foreach my $menu (@menus) { foreach my $entry (@{$menu->{'contents'}}) { if ($entry->{'type'} and $entry->{'type'} eq 'menu_entry') { push @master_menu_contents, Texinfo::Common::copy_tree($entry); # gather node children to recursively print their menus my ($normalized_entry_node, $node) = _normalized_entry_associated_internal_node($entry, $labels); if (defined($node) and $node->{'extra'}) { push @node_children, $node; } } } } if (scalar(@master_menu_contents)) { # Prepend node title my $node_title_contents; if ($node->{'extra'}->{'associated_section'} and $node->{'extra'}->{'associated_section'}->{'args'} and $node->{'extra'}->{'associated_section'}->{'args'}->[0] and $node->{'extra'}->{'associated_section'}->{'args'}->[0]->{'contents'}) { $node_title_contents = Texinfo::Common::copy_contents( $node->{'extra'}->{'associated_section'}->{'args'}->[0]->{'contents'}); } else { $node_title_contents = Texinfo::Common::copy_contents($node->{'args'}->[0]->{'contents'}); } my $menu_comment = {'type' => 'menu_comment', 'contents' => []}; $menu_comment->{'contents'}->[0] = {'type' => 'preformatted', 'parent' => $menu_comment}; $menu_comment->{'contents'}->[0]->{'contents'} = [{'text' => "\n", 'type' => 'empty_line'}, @$node_title_contents, {'text' => "\n", 'type' => 'empty_line'}, {'text' => "\n", 'type' => 'empty_line'}]; foreach my $content (@{$menu_comment->{'contents'}->[0]->{'contents'}}) { $content->{'parent'} = $menu_comment->{'contents'}->[0]; } unshift @master_menu_contents, $menu_comment; # now recurse in the children foreach my $child (@node_children) { push @master_menu_contents, _print_down_menus($child, $labels); } } } return @master_menu_contents; } if (0) { # it is needed to mark the translation as gdt is called like # Texinfo::Translations::gdt($self, ' --- The Detailed Node Listing ---') # and not like gdt(' --- The Detailed Node Listing ---') gdt(' --- The Detailed Node Listing ---'); } sub _sort_index_entries($$) { my $key1 = shift; my $key2 = shift; my $key_index = 0; # the keys array corresponds to th emain entry and subentries foreach my $key1_str (@{$key1->{'keys'}}) { my $res = _sort_string($key1_str, $key2->{'keys'}->[$key_index]); if ($res != 0) { return $res; } $key_index ++; if (scalar(@{$key2->{'keys'}}) <= $key_index) { last; } } my $res = (scalar(@{$key1->{'keys'}}) <=> scalar(@{$key2->{'keys'}})); if ($res == 0) { $res = ($key1->{'number'} <=> $key2->{'number'}); } # This may happen if 2 indices are merged as the number is per # index name. if ($res == 0) { $res = ($key1->{'index_name'} cmp $key2->{'index_name'}); } return $res; } # This is a duplicate of the functions above, for efficiency sub _collator_sort_string($$$) { my $a = shift; my $b = shift; my $collator = shift; return (($a =~ /^[[:alpha:]]/ and $b =~ /^[[:alpha:]]/) or ($a !~ /^[[:alpha:]]/ and $b !~ /^[[:alpha:]]/)) ? ($collator->cmp ($a, $b)) : (($a =~ /^[[:alpha:]]/ && 1) || -1); } sub _collator_sort_index_entries($$$) { my $key1 = shift; my $key2 = shift; my $collator = shift; my $key_index = 0; # the keys array corresponds to th emain entry and subentries foreach my $key1_str (@{$key1->{'keys'}}) { my $res = _collator_sort_string($key1_str, $key2->{'keys'}->[$key_index], $collator); if ($res != 0) { return $res; } $key_index ++; if (scalar(@{$key2->{'keys'}}) <= $key_index) { last; } } my $res = (scalar(@{$key1->{'keys'}}) <=> scalar(@{$key2->{'keys'}})); if ($res == 0) { $res = ($key1->{'number'} <=> $key2->{'number'}); } # This may happen if 2 indices are merged as the number is per # index name. if ($res == 0) { $res = ($key1->{'index_name'} cmp $key2->{'index_name'}); } return $res; } sub setup_index_entry_keys_formatting($) { my $customization_info = shift; my $options = {'ascii_punctuation' => 1, Texinfo::Convert::Text::copy_options_for_convert_text( $customization_info)}; if (not $customization_info->get_conf('ENABLE_ENCODING') or lc($customization_info->get_conf('OUTPUT_ENCODING_NAME')) ne 'utf-8') { $options->{'sort_string'} = 1; } return $options; } # can be used for subentries sub index_entry_sort_string($$$$;$) { my $main_entry = shift; my $entry_tree_element = shift; my $sortas = shift; my $options = shift; my $collator = shift; my $entry_key; if (defined($sortas)) { $entry_key = $sortas; } else { $entry_key = Texinfo::Convert::Text::convert_to_text( $entry_tree_element, $options); # FIXME do that for sortas too? if (defined($main_entry->{'entry_element'} ->{'extra'}->{'index_ignore_chars'})) { my $ignore_chars = quotemeta($main_entry->{'entry_element'} ->{'extra'}->{'index_ignore_chars'}); if ($ignore_chars ne '') { $entry_key =~ s/[$ignore_chars]//g; } } } # This avoids varying results depending on whether the string is # represented internally in UTF-8. See "the Unicode bug" in the # "perlunicode" man page. utf8::upgrade($entry_key); my $sort_entry_key; if ($collator) { $sort_entry_key = $collator->getSortKey(uc($entry_key)); } else { $sort_entry_key = uc($entry_key); } return ($entry_key, $sort_entry_key); } # This is a stub for the Unicode::Collate module. Although this module is # a core Perl module, some distributions may install a stripped-down Perl # that doesn't include it, so providing this fall-back allows texi2any # to run in such cases. Using this fall-back will change index sorting, # especially of punctuation characters and in non-English manuals. # # This fall-back also allows checking the performance impact of # Unicode::Collate (last checked as about a 5% increase in runtime for # typical Info output). package Texinfo::CollateStub; sub new($%) { my $class = shift; my %options = @_; my $self = {}; bless $self, $class; return $self; } sub getSortKey($$) { my $self = shift; my $string = shift; return $string; } sub cmp($$$) { my ($self, $a, $b) = @_; return ($a cmp $b); } package Texinfo::Structuring; # if true pre-set collating keys #my $default_preset_keys = 0; my $default_preset_keys = 1; # the structure returned depends on $SORT_BY_LETTER being set # or not. It is described in the pod documentation. sub sort_indices($$$$;$$) { my $registrar = shift; my $customization_information = shift; my $index_entries = shift; my $indices_information = shift; my $sort_by_letter = shift; my $preset_keys = shift; $preset_keys = $default_preset_keys if (!defined($preset_keys)); # The 'Non-Ignorable' for variable collation elements means that they are # treated as normal characters. This allows to have spaces and punctuation # marks sort before letters. # http://www.unicode.org/reports/tr10/#Variable_Weighting my %collate_options = ( 'variable' => 'Non-Ignorable' ); # TODO Unicode::Collate has been in perl core long enough, but # Unicode::Collate::Locale is present since perl major version 5.14 only, # released in 2011. So probably better to use Unicode::Collate until 2031 # (and if documentlanguage is not set) and switch to Unicode::Collate::Locale # at this date. #my $collator = Unicode::Collate::Locale->new('locale' => $documentlanguage, # %collate_options); # The Unicode::Collate sorting changes often, based on the UCA version. # To test the result with a specific version, the UCA_Version should be set, # and, more importantly the table should correspond to that version. # To test a specific table, in tp, do # wget -N http://www.unicode.org/Public/UCA/6.2.0/allkeys.txt # mkdir -p Unicode/Collate/ # mv allkeys.txt Unicode/Collate/allkeys-6.2.0.txt # The table argument leads to a very important slowdown, so the argument # should only be used for checks. # The test results seem to be consistent with 6.2.0, corresponding # to the perl 5.18.0 Unicode::Collate # to test for 6.2.0 #%collate_options = (%collate_options, # 'UCA_Version' => 24, # 'table' => 'allkeys-6.2.0.txt'); # To test files affected for UCA corresponding to perl 5.8.1 # wget -N http://www.unicode.org/Public/UCA/3.1.1/allkeys-3.1.1.txt #%collate_options = (%collate_options, # 'UCA_Version' => 9, # 'table' => 'allkeys-3.1.1.txt'); # Fall back to stub if Unicode::Collate not available. my $collator; eval { require Unicode::Collate; Unicode::Collate->import; }; my $unicode_collate_loading_error = $@; if ($unicode_collate_loading_error eq '') { $collator = Unicode::Collate->new(%collate_options); } else { $collator = Texinfo::CollateStub->new(); } my $entries_collator; $entries_collator = $collator if $preset_keys; my $sorted_index_entries; my $index_entries_sort_strings = {}; return $sorted_index_entries, $index_entries_sort_strings unless ($index_entries); my $options = setup_index_entry_keys_formatting($customization_information); $sorted_index_entries = {}; foreach my $index_name (keys(%$index_entries)) { # used if not $sort_by_letter my $sortable_index_entries = []; # used if $sort_by_letter my $index_letter_hash = {}; 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 $main_entry_sortas; my $convert_to_text_options = {%$options, 'code' => $indices_information->{$entry_index_name}->{'in_code'}}; $main_entry_sortas = $main_entry_element->{'extra'}->{'sortas'} if ($main_entry_element->{'extra'}); my ($entry_key, $sort_entry_key) = index_entry_sort_string($index_entry, Texinfo::Common::index_content_element($main_entry_element), $main_entry_sortas, $convert_to_text_options, $entries_collator); my @entry_keys; my @sort_entry_keys; my $letter = ''; if ($entry_key !~ /\S/) { my $entry_cmdname = $main_entry_element->{'cmdname'}; $entry_cmdname = $main_entry_element->{'extra'}->{'original_def_cmdname'} if (!defined($entry_cmdname)); $registrar->line_warn($customization_information, sprintf(__("empty index key in \@%s"), $entry_cmdname), $main_entry_element->{'source_info'}); push @entry_keys, ''; push @sort_entry_keys, ''; } else { push @entry_keys, $entry_key; push @sort_entry_keys, $sort_entry_key; if ($sort_by_letter) { # the following line leads to each accented letter being separate # $letter = uc(substr($entry_key, 0, 1)); my $letter_string = uc(substr($entry_key, 0, 1)); # determine main letter by decomposing and removing diacritics $letter = Unicode::Normalize::NFKD($letter_string); $letter =~ s/\p{NonspacingMark}//g; # following code is less good, as the upper-casing may lead to # two letters in case of the german Eszett that becomes SS. So # it is better to upper-case first and remove diacritics after. #my $normalized_string = Unicode::Normalize::NFKD(uc($entry_key)); #$normalized_string =~ s/\p{NonspacingMark}//g; #$letter = substr($normalized_string, 0, 1); } } my $subentry_nr = 0; my $subentry = $main_entry_element; while ($subentry->{'extra'} and $subentry->{'extra'}->{'subentry'}) { $subentry_nr ++; $subentry = $subentry->{'extra'}->{'subentry'}; my ($subentry_key, $sort_subentry_key) = index_entry_sort_string($index_entry, {'contents' => $subentry->{'args'}->[0]->{'contents'}}, $subentry->{'extra'}->{'sortas'}, $convert_to_text_options, $entries_collator); if ($subentry_key !~ /\S/) { my $entry_cmdname = $main_entry_element->{'cmdname'}; $entry_cmdname = $main_entry_element->{'extra'}->{'original_def_cmdname'} if (!defined($entry_cmdname)); $registrar->line_warn($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 { push @entry_keys, $subentry_key; push @sort_entry_keys, $sort_subentry_key; } } foreach my $sub_entry_key (@sort_entry_keys) { if ($sub_entry_key ne '') { my $sortable_entry = {'entry' => $index_entry, 'keys' => \@sort_entry_keys, 'number' => $index_entry->{'entry_number'}, 'index_name' => $entry_index_name}; if ($sort_by_letter) { push @{$index_letter_hash->{$letter}}, $sortable_entry; } else { push @{$sortable_index_entries}, $sortable_entry; } last; } } $index_entries_sort_strings->{$index_entry} = join(', ', @entry_keys); } if ($sort_by_letter) { # need to use directly the collator here as there is no # separate sort keys. my @sorted_letters = sort {_collator_sort_string($a, $b, $collator)} (keys %$index_letter_hash); foreach my $letter (@sorted_letters) { my @sorted_letter_entries; if ($preset_keys) { @sorted_letter_entries = map {$_->{'entry'}} sort {_sort_index_entries($a, $b)} @{$index_letter_hash->{$letter}}; } else { @sorted_letter_entries = map {$_->{'entry'}} sort {_collator_sort_index_entries($a, $b, $collator)} @{$index_letter_hash->{$letter}}; } push @{$sorted_index_entries->{$index_name}}, { 'letter' => $letter, 'entries' => \@sorted_letter_entries }; } } else { if ($preset_keys) { $sorted_index_entries->{$index_name} = [ map {$_->{'entry'}} sort {_sort_index_entries($a, $b)} @{$sortable_index_entries} ]; } else { $sorted_index_entries->{$index_name} = [ map {$_->{'entry'}} sort {_collator_sort_index_entries($a, $b, $collator)} @{$sortable_index_entries} ]; } } } return $sorted_index_entries, $index_entries_sort_strings; } sub merge_indices($) { my $indices_information = shift; my $merged_index_entries; foreach my $index_name (keys(%$indices_information)) { my $index_info = $indices_information->{$index_name}; next if ($index_info->{'merged_in'}); foreach my $contained_index (keys (%{$index_info->{'contained_indices'}})) { if ($indices_information->{$contained_index}->{'index_entries'}) { $merged_index_entries = {} if (! $merged_index_entries); push @{$merged_index_entries->{$index_name}}, @{$indices_information->{$contained_index}->{'index_entries'}}; } } } return $merged_index_entries; } 1; __END__ =head1 NAME Texinfo::Structuring - information on Texinfo::Parser tree =head1 SYNOPSIS use Texinfo::Structuring qw(sectioning_structure nodes_tree number_floats associate_internal_references split_by_node split_by_section split_pages merge_indices sort_indices elements_directions elements_file_directions); # $tree is a Texinfo document tree. $parser is a Texinfo::Parser object. # $config is an object implementing the get_conf() method. my $registrar = $parser->registered_errors(); my $sections_root = sectioning_structure ($registrar, $config, $tree); my ($labels, $targets_list, $nodes_list) = $parser->labels_information(); my $parser_information = $parser->global_information(); my $global_commands = $parser->global_commands_information(); set_menus_node_directions($registrar, $config, $parser_information, $global_commands, $nodes_list, $labels); my $top_node = nodes_tree($registrar, $config, $parser_information, $nodes_list, $labels); complete_node_tree_with_menus($registrar, $config, $nodes_list, $top_node); my $refs = $parser->internal_references_information(); check_nodes_are_referenced($registrar, $config, $nodes_list, $top_node, $labels, $refs); associate_internal_references($registrar, $parser, $parser_information, $labels, $refs); number_floats($parser->floats_information()); my $tree_units; if ($split_at_nodes) { $tree_units = split_by_node($tree); } else { $tree_units = split_by_section($tree); } split_pages($tree_units, $split); elements_directions($config, $labels, $tree_units); elements_file_directions($tree_units); my $indices_information = $parser->indices_information(); my $merged_index_entries = merge_indices($indices_information); my $index_entries_sorted; if ($sort_by_letter) { $index_entries_sorted = sort_indices($registrar, $config, $merged_index_entries, $indices_information, 'by_letter'); } else { $index_entries_sorted = sort_indices($registrar, $config, $merged_index_entries, $indices_information); } =head1 NOTES The Texinfo Perl module main purpose is to be used in C to convert Texinfo to other formats. There is no promise of API stability. =head1 DESCRIPTION Texinfo::Structuring first allows to collect information on a Texinfo tree. In most case, it also requires information from a parser object to do that job. Thanks to C the hierarchy of sectioning commands is determined. The directions implied by menus are determined with C. The node tree is analysed with C. Nodes directions are completed with menu directions with C. Floats get their standard numbering with C and internal references are matched up with nodes, floats or anchors with C. The following methods depend on the output format, so are usually called from converters. It is also possible to associate top-level contents of the tree, which consist in nodes and sectioning commands with tree unit elements that group together a node and the next sectioning element. With C nodes are considered to be the main sectioning elements, while with C the sectioning command elements are the main elements. The first mode is typical of Info format, while the second corresponds to a traditional book. The elements may be further split in I, which are not pages as in book pages, but more like web pages, and hold series of tree unit elements. The elements may have directions to other elements prepared by C. C should also set direction related to files, provided files are associated with elements by the user. C may be used to merge indices, which may be sorted with C. =head1 METHODS No method is exported in the default case. Most methods takes a L C<$registrar> as argument for error reporting. Most also require Texinfo customization variables information, which means an object implementing the C method, in practice the main program configuration or a converter (L). Other common input arguments such as parser information, labels or refs are obtained from a parser, see L. =over =item associate_internal_references($registrar, $customization_information, $parser_information, $labels, $refs) X> Verify that internal references (C<@ref> and similar without fourth of fifth argument and menu entries) have an associated node, anchor or float. Set the C key in the C hash C hash for menu entries and in the first argument C hash for internal references C<@ref> and similar @-commands. Register errors in I<$registrar>. =item check_nodes_are_referenced($registrar, $customization_information, $nodes_list, $top_node, $labels, $refs) X> Check that all the nodes are referenced (in menu, @*ref or node direction). Register errors in I<$registrar>. Should be called after C in order to have the autogenerated menus available. =item complete_node_tree_with_menus($registrar, $customization_information, $nodes_list, $top_node) X> Complete nodes directions with menu directions. Check consistency of menus, sectionning and nodes direction structures. Register errors in I<$registrar>. =item elements_directions($customization_information, $labels, $tree_units) X> Directions are set up for the tree unit elements in the array reference I<$tree_units> given in argument. The corresponding hash is in C<< {'structure'}->{'directions'} >> and keys correspond to directions while values are elements. The following directions are set up: =over =item This The element itself. =item Forward Element next. =item Back Previous element. =item NodeForward Following node element in reading order. It is the next node, or the first in menu or the next of the up node. =item NodeBack Preceding node element. =item NodeUp =item NodeNext =item NodePrev The up, next and previous node elements. =item Up =item Next =item Prev The up, next and previous section elements. =item FastBack For top level elements, the previous top level element. For other elements the up top level element. For example, for a chapter element it is the previous chapter, for a subsection element it is the chapter element that contains the subsection. =item FastForward The next top level section element. =back =item elements_file_directions($tree_units) X> In the directions reference described above for C, sets the I and I directions to the elements in previous and following files. It also sets I directions for all the elements by using the directions of the first element in file. So, for example, I is the next node of the first element in the file of each element. The API for association of pages/elements to files is not defined yet. =item @nodes_list = get_node_node_childs_from_sectioning($node) X> I<$node> is a node tree element. Find the node I<$node> children based on the sectioning structure. For the node associated with C<@top> sectioning command, the sections associated with parts are considered. =item $entry_key = index_entry_sort_string($main_entry, $entry_tree_element, $sortas, $options) X> Return a string suitable as a sort string, for index entries. The index entry processed is I<$entry_tree_element>, and can be a C<@subentry>. I<$main_entry> is the main index entry tree element that can be used to gather information. I<$sortas> can be given to override the sort string (typically obtained from C<@sortas>). The I<$options> are options used for Texinfo to text conversion for the generation of the sort string, typically obtained from L. =item $merged_entries = merge_indices($indices_information) X> Using information returned by L<< C|Texinfo::Parser/$indices_information = $parser->indices_information() >>, a structure holding all the index entries by index name is returned, with all the entries of merged indices merged with those of the indice merged into. The I<$merged_entries> returned is a hash reference whose keys are the index names and values arrays of index entry structures described in details in L. =item $new_block = new_block_command($content, $parent, $command_name) X> Returns the texinfo tree corresponding to a block command named I<$command_name> with contents I<$content> and parent in tree I<$parent>. =item $new_menu = new_complete_node_menu($node, $use_sections) X> Returns a texinfo tree menu for node I<$node>, pointing to the children of the node obtained with the sectioning structure. If I<$use_sections> is set, use section names for the menu entry names. =item $detailmenu = new_master_menu($translations, $labels, $menus) X> Returns a detailmenu tree element formatted as a master node. I<$translations>, if defined, should be a L object and should also hold customization information. I<$menus> is an array reference containing the regular menus of the Top node. =item $entry = new_node_menu_entry($node, $use_sections) X> Returns the texinfo tree corresponding to a single menu entry pointing to I<$node>. If I<$use_sections> is set, use the section name for the menu entry name. Returns C if the node argument is missing. =item $top_node = nodes_tree($registrar, $customization_information, $parser_information, $nodes_list, $labels) X> Goes through nodes and set directions. Returns the top node. Register errors in I<$registrar>. This functions sets, in the C node element hash: =over =item node_up =item node_prev =item node_next Up, next and previous directions for the node. =back =item number_floats($float_information) X> Number the floats as described in the Texinfo manual. Sets the I key in the C hash of the float tree elements. =item $command_name = section_level_adjusted_command_name($element) X> Return the sectioning command name corresponding to the sectioning element I<$element>, adjusted in order to take into account raised and lowered sections, when needed. =item $sections_root, $sections_list = sectioning_structure($registrar, $customization_information, $tree) X> This function goes through the tree and gather information on the document structure for sectioning commands. It returns I<$sections_root> the root of the sectioning commands tree and a reference on the sections elements list. Errors are registered in I<$registrar>. It sets section elements C hash values: =over =item section_level The level in the sectioning tree hierarchy. 0 is for C<@top> or C<@part>, 1 for C<@chapter>, C<@appendix>... This level is corrected by C<@raisesections> and C<@lowersections>. =item section_number The sectioning element number. =item section_childs An array holding sectioning elements children of the element. =item section_up =item section_prev =item section_next The up, previous and next sectioning elements. =item toplevel_next =item toplevel_prev =item toplevel_up The next and previous and up sectioning elements of toplevel sectioning elements (like C<@top>, C<@chapter>, C<@appendix>), not taking into account C<@part> elements. =back =item set_menus_node_directions($registrar, $customization_information, $parser_information, $global_commands, $nodes_list, $labels); X> Goes through menu and set directions. Register errors in I<$registrar>. This functions sets, in the C node element hash reference: =over =item menu_child The first child in the menu of the node. =item menu_up =item menu_next =item menu_prev Up, next and previous directions as set in menus. =back =item $option = setup_index_entry_keys_formatting($customization_information) X> Return options for conversion of Texinfo to text relevant for index keys sorting. =item ($index_entries_sorted, $index_entries_sort_strings) = sort_indices($registrar, $customization_information, $merged_index_entries, $indices_information, $sort_by_letter) X> If I<$sort_by_letter> is set, sort by letter, otherwise sort all entries together. In both cases, a hash reference with index names as keys I<$index_entries_sorted> is returned. When sorting by letter, an array reference of letter hash references is associated with each index name. Each letter hash reference has two keys, a I key with the letter, and an I key with an array reference of sorted index entries beginning with the letter. When simply sorting, the array of the sorted index entries is associated with the index name. I<$index_entries_sort_strings> is a hash reference associating the index entries with the strings that were used to sort them. Register errors in I<$registrar>. =item $tree_units = split_by_node($tree) X> Returns a reference array of tree units where a node is associated to the following sectioning commands. Sectioning commands without nodes are also with the previous node, while nodes without sectioning commands are alone in their tree units. Tree units are regular tree elements with type I, the associated nodes and sectioning tree elements are in the array associated with the C key. The associated elements have a I key set in the C hash that points to the associated tree unit. Tree units have directions in the C hash reference, namely I and I pointing to the previous and the next tree unit. In the C hash reference, tree units have: =over =item unit_command The node command associated with the element. =back =item $tree_units = split_by_section($tree) X> Similarly with C, returns an array of tree units. This time, lone nodes are associated with the previous sections and lone sections makes up a tree unit. The C and C hash keys set are the same, except that I is the sectioning command associated with the element. =item $pages = split_pages($tree_units, $split) X> The tree units from the array reference argument have an extra I value set in the C hash reference to the first tree unit in the group, and based on the value of I<$split>. The possible values for I<$split> are =over =item chapter The tree units are split at chapter or other toplevel sectioning tree units. =item node Each element has its own page. =item section The tree units are split at sectioning commands below chapter. =item value evaluating to false No splitting, only one page is returned, holding all the tree units. =back =item warn_non_empty_parts($registrar, $customization_information, $global_commands) X> Register a warning in I<$registrar> for each C<@part> that is not empty in I<$global_commands> information (typically obtained by calling C on a parser). =back =head1 SEE ALSO L, L. =head1 AUTHOR Patrice Dumas, Epertusus@free.frE =head1 COPYRIGHT AND LICENSE Copyright 2010- Free Software Foundation, Inc. See the source file for all copyright years. This library is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version. =cut