Edit file File name : Silf.pm Content :package Font::TTF::Silf; =head1 NAME Font::TTF::Silf - The main Graphite table =head1 DESCRIPTION The Silf table holds the core of the Graphite rules for a font. A Silf table has potentially multiple silf subtables, although there is usually only one. Within a silf subtable, there are a number of passes which contain the actual finite state machines to match rules and the constraint and action code to be executed when a rule matches. =head1 INSTANCE VARIABLES =over 4 =item Version Silf table format version =item Compiler Lowest compiler version necessary to fully support the semantics expressed in this Graphite description =item SILF An array of Silf subtables =over 4 =item maxGlyphID The maximum glyph id referenced including pseudo and non glyphs =item Ascent Extra ascent to be added to the font ascent. =item Descent Extra descent to be added to the font descent. Both values are assumed to be positive for a descender below the base line. =item substPass Pass index into PASS of the first substitution pass. =item posPass Pass index into PASS of the first positioning pass. =item justPass Pass index into PASS of the first justification pass. =item bidiPass Pass index of the pass before which the bidirectional processing pass will be executed. 0xFF indicates that there is no bidi pass to be executed. =item Flags A bitfield of flags: 0 - Indicates there are line end contextual rules in one of the passes =item maxPreContext Maximum length of a context preceding a cross line boundary contextualisation. =item maxPostContext Maximum length of a context following a cross line boundary contextualsation. =item attrPseudo Glyph attribute for the actual glyph id associated with a pseudo glyph. =item attrBreakWeight Glyph attribute number of the attribute holding the default breakweight associated with a glyph. =item attrDirectionality Glyph attribute number of the attribute holding the default directionality value associated with a glyph. =item JUST The may be a number of justification levels each with their own property values. This points to an array of hashes, one for each justification level. =over 4 =item attrStretch Glyph attribute number for the amount of stretch allowed before this glyph. =item attrShrink Glyph attribute number for the amount of shrink allowed before this glyph. =item attrStep Glyph attribute number specifying the minimum granularity of actual spacing associated with this glyph at this level. =item attrWeight Glyph attribute number giving the weight associated with spreading space across a run of glyphs. =item runto Which level starts the next stage. =back =item numLigComp Number of initial glyph attributes that represent ligature components =item numUserAttr Number of user defined slot attributes referenced. Tells the engine how much space to allocate to a slot for user attributes. =item maxCompPerLig Maximum number of components per ligature. =item direction Supported directions for this writing system =item CRIT_FEATURE Array of critical features. =item scripts Array of script tags that indicate which set of GDL rules to execute if there is more than one in a font. =item lbGID Glyph ID of the linebreak pseudo glyph. =item pseudos Hash of Unicode values to pseduo glyph ids. =item classes This is an array of classes, each of which is an array of glyph ids in class order. =item PASS The details of rules and actions are stored in passes. This value is an array of pass subobjects one for each pass. =over 4 =item flags This is a bitfield: 0 - If true, this pass makes no change to the slot stream considered as a sequence of glyph ids. Only slot attributes are expected to change (for example during positioning). =item maxRuleLoop How many times the engine will allow rules to be tested and run without the engine advancing through the input slot stream. =item maxRuleContext Number of slots of input needed to run this pass. =item maxBackup Number of slots by which the following pass needs to trail this pass (i.e. the maximum this pass is allowed to back up). =item numRules Number of action code blocks, and so uncompressed rules, in this pass. =item numRows Number of rows in the finite state machine. =item numTransitional Number of rows in the finite state machine that are not final states. This specifies the number of rows in the fsm element. =item numSuccess Number of success states. A success state may also be a transitional state. =item numColumns Number of columns in the finite state machine. =item colmap A hash, indexed by glyphid, that gives the fsm column number associated with that glyphid. If not present, then the glyphid is not part of the fsm and will finish fsm processing if it occurs. =item rulemap An array of arrays, one for each success state. Each array holds a list of rule numbers associated with that state. =item minRulePreContext Minimum number of items in a rule's precontext. =item maxRulePreContext The maximum number of items in any rule's precontext. =item startStates Array of starting state numbers dependeing on the length of actual precontext. There are maxRulePreContext - minRulePreContext + 1 of these. =item ruleSortKeys An array of sort keys one for each rule giving the length of the rule including its precontext. =item rulePreContexts An array of precontext lengths for each rule. =item fsm A two dimensional array such that $p->{'fsm'}[$row][$col] gives the row of the next node to try in the fsm. =item passConstraintLen Length in bytes of the passConstraint code. =item passConstraintCode A byte string holding the pass constraint code. =item constraintCode An array of byte strings holding the constraint code for each rule. =item actionCode An array of byte strings holding the action code for each rule. =back =back =back =cut use Font::TTF::Table; use Font::TTF::Utils; use strict; use vars qw(@ISA); @ISA = qw(Font::TTF::Table); =head2 @opcodes Each array holds the name of the opcode, the number of operand bytes and a string describing the operands. The characters in the string have the following meaning: c - lsb of class id C - msb of class id f - feature index g - lsb of glyph attribute id G - msb of glyph attribute id l - lsb of a 32-bit extension to a 16-bit number L - msb of a 32-bit number m - glyph metric id n - lsb of a number N - msb of a 16-bit number o - offset (jump) s - slot reference S - slot attribute id v - variable number of following arguments =cut our @opcodes = ( ["nop", 0, ""], ["push_byte", 1, "n"], ["push_byte_u", 1, "n"], ["push_short", 2, "Nn"], ["push_short_u", 2, "Nn"], ["push_long", 4, "LlNn"], ["add", 0, ""], ["sub", 0, ""], ["mul", 0, ""], ["div", 0, ""], ["min", 0, ""], ["max", 0, ""], ["neg", 0, ""], ["trunc8", 0, ""], ["trunc16", 0, ""], ["cond", 0, ""], ["and", 0, ""], ["or", 0, ""], ["not", 0, ""], ["equal", 0, ""], # 16 ["not_eq", 0, ""], ["less", 0, ""], ["gtr", 0, ""], ["less_eq", 0, ""], ["gtr_eq", 0, ""], ["next", 0, ""], ["next_n", 1, "n"], ["copy_next", 0, ""], ["put_glyph_8bit_obs", 1, "c"], ["put_subs_8bit_obs", 3, "scc"], ["put_copy", 1, "s"], ["insert", 0, ""], ["delete", 0, ""], ["assoc", -1, "v"], ["cntxt_item", 2, "so"], ["attr_set", 1, "S"], # 32 ["attr_add", 1, "S"], ["attr_sub", 1, "S"], ["attr_set_slot", 1, "S"], ["iattr_set_slot", 2, "Sn"], ["push_slot_attr", 2, "Ss"], ["push_glyph_attr_obs", 2, "gs"], ["push_glyph_metric", 3, "msn"], ["push_feat", 2, "fs"], ["push_att_to_gattr_obs", 2, "gs"], ["push_att_to_glyph_metric", 3, "msn"], ["push_islot_attr", 3, "Ssn"], ["push_iglyph_attr", 3, "gsn"], ["pop_ret", 0, ""], ["ret_zero", 0, ""], ["ret_true", 0, ""], ["iattr_set", 2, "Sn"], # 48 ["iattr_add", 2, "Sn"], ["iattr_sub", 2, "Sn"], ["push_proc_state", 1, "n"], ["push_version", 0, ""], ["put_subs", 5, "sCcCc"], ["put_subs2", 4, "cscc"], ["put_subs3", 7, "scscscc"], ["put_glyph", 2, "Cc"], ["push_glyph_attr", 3, "Ggs"], ["push_att_to_glyph_attr", 3, "Ggs"], ["bitand", 0, ""], ["bitor", 0, ""], ["bitnot", 0, ""], ["setbits", 4, "NnNn"], ["setfeat", 2, "fs"] ); # 64 my ($i) = 0; our %opnames = map {$_->[0] => $i++} @opcodes; =head2 read Reads the Silf table into the internal data structure =cut sub read { my ($self) = @_; $self->SUPER::read or return $self; my ($dat, $d); my ($fh) = $self->{' INFILE'}; my ($moff) = $self->{' OFFSET'}; my ($numsilf, @silfo); $fh->read($dat, 4); ($self->{'Version'}) = TTF_Unpack("v", $dat); if ($self->{'Version'} >= 3) { $fh->read($dat, 4); ($self->{'Compiler'}) = TTF_Unpack("v", $dat); } $fh->read($dat, 4); ($numsilf) = TTF_Unpack("S", $dat); $fh->read($dat, $numsilf * 4); foreach my $i (0 .. $numsilf - 1) { push (@silfo, TTF_Unpack("L", substr($dat, $i * 4, 4))); } foreach my $sili (0 .. $numsilf - 1) { my ($silf) = {}; my (@passo, @classo, $classbase, $numJust, $numCritFeatures, $numScript, $numPasses, $numPseudo, $i); push (@{$self->{'SILF'}}, $silf); $fh->seek($moff + $silfo[$sili], 0); if ($self->{'Version'} >= 3) { $fh->read($dat, 8); ($silf->{'Version'}) = TTF_Unpack("v", $dat); } $fh->read($dat, 20); ($silf->{'maxGlyphID'}, $silf->{'Ascent'}, $silf->{'Descent'}, $numPasses, $silf->{'substPass'}, $silf->{'posPass'}, $silf->{'justPass'}, $silf->{'bidiPass'}, $silf->{'Flags'}, $silf->{'maxPreContext'}, $silf->{'maxPostContext'}, $silf->{'attrPseudo'}, $silf->{'attrBreakWeight'}, $silf->{'attrDirectionality'}, $silf->{'attrMirror'}, $silf->{'passBits'}, $numJust) = TTF_Unpack("SssCCCCCCCCCCCCCC", $dat); if ($numJust) { foreach my $j (0 .. $silf->{'numJust'} - 1) { my ($just) = {}; push (@{$silf->{'JUST'}}, $just); $fh->read($dat, 8); ($just->{'attrStretch'}, $just->{'attrShrink'}, $just->{'attrStep'}, $just->{'attrWeight'}, $just->{'runto'}) = TTF_Unpack("CCCCC", $dat); } } $fh->read($dat, 10); ($silf->{'numLigComp'}, $silf->{'numUserAttr'}, $silf->{'maxCompPerLig'}, $silf->{'direction'}, $silf->{'attCollisions'}, $d, $d, $d, $numCritFeatures) = TTF_Unpack("SCCCCCCCC", $dat); if ($numCritFeatures) { $fh->read($dat, $numCritFeatures * 2); $silf->{'CRIT_FEATURE'} = [TTF_Unpack("S$numCritFeatures", $dat)]; } $fh->read($dat, 2); ($d, $numScript) = TTF_Unpack("CC", $dat); if ($numScript) { $fh->read($dat, $numScript * 4); foreach (0 .. $numScript - 1) { push (@{$silf->{'scripts'}}, unpack('a4', substr($dat, $_ * 4, 4))); } } $fh->read($dat, 2); ($silf->{'lbGID'}) = TTF_Unpack("S", $dat); $fh->read($dat, $numPasses * 4 + 4); @passo = unpack("N*", $dat); $fh->read($dat, 8); ($numPseudo) = TTF_Unpack("S", $dat); if ($numPseudo) { $fh->read($dat, $numPseudo * 6); foreach (0 .. $numPseudo - 1) { my ($uni, $gid) = TTF_Unpack("LS", substr($dat, $_ * 6, 6)); $silf->{'pseudos'}{$uni} = $gid; } } $classbase = $fh->tell(); $fh->read($dat, 4); my ($numClasses, $numLinearClasses) = TTF_Unpack("SS", $dat); $silf->{'numLinearClasses'} = $numLinearClasses; $fh->read($dat, ($numClasses + 1) * ($self->{'Version'} >= 4 ? 4 : 2)); @classo = unpack($self->{'Version'} >= 4 ? "N*" : "n*", $dat); $fh->read($dat, $classo[-1] - $classo[0]); for ($i = 0; $i < $numLinearClasses; $i++) { push (@{$silf->{'classes'}}, [unpack("n*", substr($dat, $classo[$i] - $classo[0], $classo[$i+1] - $classo[$i]))]) } for ($i = $numLinearClasses; $i < $numClasses; $i++) { my (@res); my (@c) = unpack("n*", substr($dat, $classo[$i] - $classo[0] + 8, $classo[$i+1] - $classo[$i] - 8)); for (my $j = 0; $j < @c; $j += 2) { $res[$c[$j+1]] = $c[$j]; } push (@{$silf->{'classes'}}, \@res); } foreach (0 .. $numPasses - 1) { $self->read_pass($fh, $passo[$_], $moff + $silfo[$sili], $silf, $_); } } return $self; } sub chopcode { my ($dest, $dat, $offsets, $isconstraint) = @_; my ($last) = $offsets->[-1]; my ($i); for ($i = $#{$offsets} - 1; $i >= 0; $i--) { if ((!$isconstraint || $offsets->[$i]) && $offsets->[$i] != $last) { unshift(@{$dest}, substr($dat, $offsets->[$i], $last - $offsets->[$i])); $last = $offsets->[$i]; } else { unshift(@{$dest}, ""); } } } sub read_pass { my ($self, $fh, $offset, $base, $silf, $id) = @_; my ($pass) = {'id' => $id}; my ($d, $dat, $i, @orulemap, @oconstraints, @oactions, $numRanges); $fh->seek($offset + $base, 0); # printf "pass base = %04X\n", $offset; push (@{$silf->{'PASS'}}, $pass); $fh->read($dat, 40); ($pass->{'flags'}, $pass->{'maxRuleLoop'}, $pass->{'maxRuleContext'}, $pass->{'maxBackup'}, $pass->{'numRules'}, $d, $d, $d, $d, $d, $pass->{'numRows'}, $pass->{'numTransitional'}, $pass->{'numSuccess'}, $pass->{'numColumns'}, $numRanges) = TTF_Unpack("CCCCSSLLLLSSSSS", $dat); $fh->read($dat, $numRanges * 6); foreach $i (0 .. $numRanges - 1) { my ($first, $last, $col) = TTF_Unpack('SSS', substr($dat, $i * 6, 6)); foreach ($first .. $last) { $pass->{'colmap'}{$_} = $col; } } $fh->read($dat, $pass->{'numSuccess'} * 2 + 2); @orulemap = unpack("n*", $dat); $fh->read($dat, $orulemap[-1] * 2); foreach (0 .. $pass->{'numSuccess'} - 1) { push (@{$pass->{'rulemap'}}, [unpack("n*", substr($dat, $orulemap[$_] * 2, ($orulemap[$_+1] - $orulemap[$_]) * 2))]); } $fh->read($dat, 2); ($pass->{'minRulePreContext'}, $pass->{'maxRulePreContext'}) = TTF_Unpack("CC", $dat); $fh->read($dat, ($pass->{'maxRulePreContext'} - $pass->{'minRulePreContext'} + 1) * 2); $pass->{'startStates'} = [unpack('n*', $dat)]; $fh->read($dat, $pass->{'numRules'} * 2); $pass->{'ruleSortKeys'} = [unpack('n*', $dat)]; $fh->read($dat, $pass->{'numRules'}); $pass->{'rulePreContexts'} = [unpack('C*', $dat)]; $fh->read($dat, 3); ($pass->{'collisionThreshold'}, $pass->{'passConstraintLen'}) = TTF_Unpack("CS", $dat); $fh->read($dat, ($pass->{'numRules'} + 1) * 2); @oconstraints = unpack('n*', $dat); $fh->read($dat, ($pass->{'numRules'} + 1) * 2); @oactions = unpack('n*', $dat); foreach (0 .. $pass->{'numTransitional'} - 1) { $fh->read($dat, $pass->{'numColumns'} * 2); push (@{$pass->{'fsm'}}, [unpack('n*', $dat)]); } $fh->read($dat, 1); if ($pass->{'passConstraintLen'}) { $fh->read($pass->{'passConstraintCode'}, $pass->{'passConstraintLen'}); } $fh->read($dat, $oconstraints[-1]); $pass->{'constraintCode'} = []; chopcode($pass->{'constraintCode'}, $dat, \@oconstraints, 1); $fh->read($dat, $oactions[-1]); $pass->{'actionCode'} = []; chopcode($pass->{'actionCode'}, $dat, \@oactions, 0); return $pass; } sub chopranges { my ($map, $numg) = @_; my ($dat, $numRanges); my (@keys) = sort {$a <=> $b} keys %{$map}; my ($first, $last, $col, $g); $first = -1; $last = -1; $col = -1; foreach $g (@keys) { next unless ($g > 0 or $g eq '0'); if ($g != $last + 1 || $map->{$g} != $col) { if ($col != -1) { $dat .= pack("nnn", $first, $last, $col); $numRanges++; } $first = $last = $g; $col = $map->{$g}; } else { $last++; } } if ($col != -1) { $dat .= pack("nnn", $first, $last, $col); $numRanges++; } return ($numRanges, $dat); } sub unpack_code { my ($self, $str) = @_; my (@res, $i, $j); my ($l) = length($str); for ($i = 0; $i < $l; ) { my ($a) = unpack('C', substr($str, $i, 1)); my ($o) = $opcodes[$a]; my (@args); my (@types) = split('', $o->[2]); ++$i; for ($j = 0; $j < @types; ++$j) { my ($t) = $types[$j]; if ($t eq 'v') { my ($n) = unpack('C', substr($str, $i, 1)); push (@args, unpack('C*', substr($str, $i + 1, $n))); $i += $n + 1; } elsif ($t eq 'L' or $t eq 'N' or $t eq 'G' or $t eq 'C') { push (@args, unpack('n', substr($str, $i, 2))); $i += 2; $j++; } else { push (@args, unpack($t eq 's' ? 'c' : 'C', substr($str, $i, 1))); $i++; } } push (@res, [$o->[0], @args]); } return @res; } sub pack_code { my ($self, $cmds) = @_; my ($res); foreach my $c (@{$cmds}) { my ($ind) = $opnames{$c->[0]}; my ($i) = 1; $res .= pack('C', $ind); # my (@types) = unpack('C*', $opcodes[$ind][2]); my (@types) = split('', $opcodes[$ind][2]); for (my $j = 0; $j < @types; $j++) { my ($t) = $types[$j]; if ($t eq 'v') { my ($n) = scalar @{$c} - 1; $res .= pack('C*', $n, @{$c}[1..$#{$c}]); $i += $n; } elsif ($t eq 'C' or $t eq 'G' or $t eq 'L' or $t eq 'N') { $res .= pack('n', $c->[$i]); $j++; } else { $res .= pack($t eq 's' ? 'c' : 'C', $c->[$i]); } $i++; } } return $res; } sub packcode { my ($code, $isconstraint) = @_; my ($dat, $c, $res); $c = 1; $dat = "\000"; foreach (@{$code}) { if ($_) { push(@{$res}, $c); $dat .= $_; $c += length($_); } else { push(@{$res}, $isconstraint ? 0 : $c); } } push(@{$res}, $c); return ($res, $dat); } sub out_pass { my ($self, $fh, $pass, $silf, $subbase) = @_; my (@orulemap, $dat, $actiondat, $numRanges, $c); my (@offsets, $res, $pbase); $pbase = $fh->tell(); # printf "pass base = %04X, ", $pbase - $subbase; $fh->print(TTF_Pack("CCCCSSLLLLSSSS", $pass->{'flags'}, $pass->{'maxRuleLoop'}, $pass->{'maxRuleContext'}, $pass->{'maxBackup'}, $pass->{'numRules'}, 24, 0, 0, 0, 0, $pass->{'numRows'}, $pass->{'numTransitional'}, $pass->{'numSuccess'}, $pass->{'numColumns'})); ($numRanges, $dat) = chopranges($pass->{'colmap'}); # print "numranges = $numRanges\n"; $fh->print(TTF_Pack("SSSS", TTF_bininfo($numRanges, 6))); $fh->print($dat); $dat = ""; $c = 0; # print "transitions = $pass->{'numTransitional'}, success = $pass->{'numSuccess'}, rows = $pass->{'numRows'}\n"; my ($sucbase) = $pass->{'numRows'} - $pass->{'numSuccess'}; foreach (0 .. ($pass->{'numSuccess'} - 1)) { push(@orulemap, $c); if (defined $pass->{'rulemap'}[$_]) { $dat .= pack("n*", @{$pass->{'rulemap'}[$_]}); $c += @{$pass->{'rulemap'}[$_]}; } else { print "No rules for " . ($sucbase + $_); if ($sucbase + $_ < $pass->{'numTransitional'}) { print ": (" . join(",", @{$pass->{'fsm'}[$sucbase + $_]}) . ")"; } print "\n"; } } push (@orulemap, $c); $fh->print(pack("n*", @orulemap)); $fh->print($dat); $fh->print(TTF_Pack("CC", $pass->{'minRulePreContext'}, $pass->{'maxRulePreContext'})); $fh->print(pack("n*", @{$pass->{'startStates'}})); $fh->print(pack("n*", @{$pass->{'ruleSortKeys'}})); $fh->print(pack("C*", @{$pass->{'rulePreContexts'}})); $fh->print(TTF_Pack("CS", 0, $pass->{'passConstraintLen'})); my ($oconstraints, $oactions); ($oconstraints, $dat) = packcode($pass->{'constraintCode'}, 1); ($oactions, $actiondat) = packcode($pass->{'actionCode'}, 0); # printf "constraint offsets @ %X\n", $fh->tell(); $fh->print(pack("n*", @{$oconstraints})); # printf "action offsets @ %X\n", $fh->tell(); $fh->print(pack("n*", @{$oactions})); # printf "fsm @ %X\n", $fh->tell(); foreach (@{$pass->{'fsm'}}) { $fh->print(pack("n*", @{$_})); } # printf "end of fsm @ %X\n", $fh->tell(); $fh->print(pack("C", $pass->{'collisionThreshold'})); push(@offsets, $fh->tell() - $subbase); $fh->print($pass->{'passConstraintCode'}); push(@offsets, $fh->tell() - $subbase); $fh->print($dat); push(@offsets, $fh->tell() - $subbase); $fh->print($actiondat); push(@offsets, 0); print join(", ", @offsets) . "\n"; $res = $fh->tell(); $fh->seek($pbase + 8, 0); $fh->print(pack("N*", @offsets)); $fh->seek($res, 0); # printf "end = %04X\n", $res - $subbase; return $res; } =head2 out Outputs a Silf data structure to a font file in binary format =cut sub out { my ($self, $fh) = @_; my ($silf, $base, $subbase, $silfc, $end); return $self->SUPER::out($fh) unless ($self->{' read'}); $base = $fh->tell(); if ($self->{'Version'} >= 3) { $fh->print(TTF_Pack("vvSS", $self->{'Version'}, $self->{'Compiler'}, $#{$self->{'SILF'}} + 1, 0)); } else { $fh->print(TTF_Pack("vSS", $self->{'Version'}, $#{$self->{'SILF'}} + 1, 0)); } $fh->print(pack('N*', (0) x (@{$self->{'SILF'}}))); foreach $silf (@{$self->{'SILF'}}) { my ($subbase) = $fh->tell(); my ($numlin, $i, @opasses, $oPasses, $oPseudo, $ooPasses); if ($self->{'Version'} >= 3) { $fh->seek($base + 12 + $silfc * 4, 0); $fh->print(pack('N', $subbase - $base)); $fh->seek($subbase, 0); $fh->print(TTF_Pack("vSS", $silf->{'Version'}, $ooPasses, $oPseudo)); } else { $fh->seek($base + 8 + $silfc * 4, 0); $fh->print(pack('N', $subbase - $base)); $fh->seek($subbase, 0); } $fh->print(TTF_Pack("SssCCCCCCCCCCCCCC", $silf->{'maxGlyphID'}, $silf->{'Ascent'}, $silf->{'Descent'}, scalar @{$silf->{'PASS'}}, $silf->{'substPass'}, $silf->{'posPass'}, $silf->{'justPass'}, $silf->{'bidiPass'}, $silf->{'Flags'}, $silf->{'maxPreContext'}, $silf->{'maxPostContext'}, $silf->{'attrPseudo'}, $silf->{'attrBreakWeight'}, $silf->{'attrDirectionality'}, $silf->{'attrMirror'}, $silf->{'passBits'}, $#{$silf->{'JUST'}} + 1)); foreach (@{$silf->{'JUST'}}) { $fh->print(TTF_Pack("CCCCCCCC", $_->{'attrStretch'}, $_->{'attrShrink'}, $_->{'attrStep'}, $_->{'attrWeight'}, $_->{'runto'}, 0, 0, 0)); } $fh->print(TTF_Pack("SCCCCCCCC", $silf->{'numLigComp'}, $silf->{'numUserAttr'}, $silf->{'maxCompPerLig'}, $silf->{'direction'}, $silf->{'attCollisions'}, 0, 0, 0, $#{$silf->{'CRIT_FEATURE'}} + 1)); $fh->print(pack("n*", @{$silf->{'CRIT_FEATURE'}})); $fh->print(TTF_Pack("CC", 0, $#{$silf->{'scripts'}} + 1)); foreach (@{$self->{'scripts'}}) { $fh->print(pack("a4", $_)); } $fh->print(TTF_Pack("S", $silf->{'lbGID'})); $ooPasses = $fh->tell(); if ($silf->{'PASS'}) { $fh->print(pack("N*", (0) x (@{$silf->{'PASS'}} + 1)));} $oPseudo = $fh->tell() - $subbase; my (@pskeys) = keys %{$silf->{'pseudos'}}; $fh->print(TTF_Pack("SSSS", TTF_bininfo(scalar @pskeys, 6))); foreach my $k (sort {$a <=> $b} @pskeys) { $fh->print(TTF_Pack("Ls", $k, $silf->{'pseudos'}{$k})); } $numlin = $silf->{'numLinearClasses'}; $fh->print(TTF_Pack("SS", scalar @{$silf->{'classes'}}, $numlin)); my (@coffsets); # printf "%X, ", $fh->tell() - $base; my ($cbase) = (scalar @{$silf->{'classes'}} + 1) * ($self->{'Version'} >= 4 ? 4 : 2) + 4; for ($i = 0; $i < $numlin; $i++) { push (@coffsets, $cbase); $cbase += 2 * scalar @{$silf->{'classes'}[$i]}; } my (@nonlinclasses); for ($i = $numlin; $i < @{$silf->{'classes'}}; $i++) { my (@c, $d, @d); my $c = $silf->{'classes'}[$i]; push (@coffsets, $cbase); @c = sort {$c->[$a] <=> $c->[$b]} (0 .. $#{$c}); foreach $d (@c) { push (@d, $c->[$d], $d); } push (@nonlinclasses, [@d]); my ($len) = scalar @d; $cbase += 8 + 2 * $len; } push (@coffsets, $cbase); $fh->print(pack(($self->{'Version'} >= 4 ? 'N*' : 'n*'), @coffsets)); for ($i = 0; $i < $numlin; $i++) { $fh->print(pack("n*", @{$silf->{'classes'}[$i]})); } # printf "%X, ", $fh->tell() - $base; for ($i = $numlin; $i < @{$silf->{'classes'}}; $i++) { my ($num) = scalar @{$nonlinclasses[$i-$numlin]}; my (@bin) = TTF_bininfo($num/2, 1); $fh->print(TTF_Pack("SSSS", @bin)); $fh->print(pack("n*", @{$nonlinclasses[$i-$numlin]})); } $oPasses = $fh->tell() - $subbase; # printf "original pass = %04X\n", $oPasses; push (@opasses, $oPasses); foreach (@{$silf->{'PASS'}}) { push(@opasses, $self->out_pass($fh, $_, $silf, $subbase) - $subbase); } $end = $fh->tell(); $fh->seek($ooPasses, 0); $fh->print(pack("N*", @opasses)); if ($self->{'Version'} >= 3) { $fh->seek($subbase + 4, 0); $fh->print(TTF_Pack("SS", $ooPasses - $subbase, $oPseudo)); } $fh->seek($end, 0); $silfc++; } } sub XML_element { my ($self, $context, $depth, $k, $val, $ind) = @_; my ($fh) = $context->{'fh'}; my ($i); return $self if ($k eq 'LOC'); if ($k eq 'classes') { $fh->print("$depth<classes>\n"); foreach $i (0 .. $#{$val}) { $fh->printf("$depth <class num='%d'>\n", $i); $fh->printf("$depth " . join(" ", map{sprintf("%d", $_)} @{$val->[$i]})); $fh->print("\n$depth </class>\n"); } $fh->print("$depth</classes>\n"); } elsif ($k eq 'fsm') { $fh->print("$depth<fsm>\n"); my ($i) = 0; foreach (@{$val}) { $fh->print("$depth <row index='$i'>" . join(" ", @{$_}) . "</row>\n"); $i++; } $fh->print("$depth</fsm>\n"); } elsif ($k eq 'colmap') { my ($i); $fh->print("$depth<colmap>"); foreach my $k (sort {$a <=> $b} keys %{$val}) { if ($i++ % 8 == 0) { $fh->print("\n$depth "); } $fh->printf(" %d=%d", $k, $val->{$k}); } $fh->print("\n$depth</colmap>\n"); } elsif ($k eq 'constraintCode' or $k eq 'actionCode') { $fh->print("$depth<$k>\n"); foreach my $i (0 .. $#{$val}) { my (@rules) = $self->unpack_code($val->[$i]); next unless (@rules); $fh->print("$depth <elem index='$i' code='" . join(" ", unpack('C*', $val->[$i])) . "'>\n"); foreach my $r (@rules) { $fh->print("$depth $r->[0]: ". join(", ", @{$r}[1..$#{$r}]) . "\n"); } $fh->print("$depth </elem>\n"); } $fh->print("$depth</$k>\n"); } else { return $self->SUPER::XML_element($context, $depth, $k, $val, $ind); } $self; } =head2 $t->minsize() Returns the minimum size this table can be. If it is smaller than this, then the table must be bad and should be deleted or whatever. =cut sub minsize { return 4; } 1; =head1 AUTHOR Martin Hosken L<http://scripts.sil.org/FontUtils>. =head1 LICENSING Copyright (c) 1998-2016, SIL International (http://www.sil.org) This module is released under the terms of the Artistic License 2.0. For details, see the full text of the license in the file LICENSE. =cut Save