Otl.php

<?php
 
namespace Mpdf;
 
use Mpdf\Css\TextVars;
use Mpdf\Fonts\FontCache;
use Mpdf\Shaper\Indic;
use Mpdf\Shaper\Myanmar;
use Mpdf\Shaper\Sea;
 
define("_OTL_OLD_SPEC_COMPAT_1", true);
 
define("_DICT_NODE_TYPE_SPLIT", 0x01);
define("_DICT_NODE_TYPE_LINEAR", 0x02);
define("_DICT_INTERMEDIATE_MATCH", 0x03);
define("_DICT_FINAL_MATCH", 0x04);
 
class Otl
{
 
const _OTL_OLD_SPEC_COMPAT_1 = true;
const _DICT_NODE_TYPE_SPLIT = 0x01;
const _DICT_NODE_TYPE_LINEAR = 0x02;
const _DICT_INTERMEDIATE_MATCH = 0x03;
const _DICT_FINAL_MATCH = 0x04;
 
private $mpdf;
 
private $fontCache;
 
var $arabLeftJoining;
 
var $arabRightJoining;
 
var $arabTransparentJoin;
 
var $arabTransparent;
 
var $GSUBdata;
 
var $GPOSdata;
 
var $GSUBfont;
 
var $fontkey;
 
var $ttfOTLdata;
 
var $glyphIDtoUni;
 
var $_pos;
 
var $GSUB_offset;
 
var $GPOS_offset;
 
var $MarkAttachmentType;
 
var $MarkGlyphSets;
 
var $GlyphClassMarks;
 
var $GlyphClassLigatures;
 
var $GlyphClassBases;
 
var $GlyphClassComponents;
 
var $Ignores;
 
var $LuCoverage;
 
var $OTLdata;
 
var $assocLigs;
 
var $assocMarks;
 
var $shaper;
 
var $restrictToSyllable;
 
var $lbdicts; // Line-breaking dictionaries
 
var $LuDataCache;
 
var $debugOTL = false;
 
public function __construct(Mpdf $mpdf, FontCache $fontCache)
{
$this->mpdf = $mpdf;
$this->fontCache = $fontCache;
 
$this->current_fh = '';
 
$this->lbdicts = [];
$this->LuDataCache = [];
}
 
function applyOTL($str, $useOTL)
{
if (!$this->arabLeftJoining) {
$this->arabic_initialise();
}
 
$this->OTLdata = [];
if (trim($str) == '') {
return $str;
}
if (!$useOTL) {
return $str;
}
 
// 1. Load GDEF data
//==============================
$this->fontkey = $this->mpdf->CurrentFont['fontkey'];
$this->glyphIDtoUni = $this->mpdf->CurrentFont['glyphIDtoUni'];
if (!isset($this->GDEFdata[$this->fontkey])) {
include $this->fontCache->tempFilename($this->fontkey . '.GDEFdata.php');
$this->GSUB_offset = $this->GDEFdata[$this->fontkey]['GSUB_offset'] = $GSUB_offset;
$this->GPOS_offset = $this->GDEFdata[$this->fontkey]['GPOS_offset'] = $GPOS_offset;
$this->GSUB_length = $this->GDEFdata[$this->fontkey]['GSUB_length'] = $GSUB_length;
$this->MarkAttachmentType = $this->GDEFdata[$this->fontkey]['MarkAttachmentType'] = $MarkAttachmentType;
$this->MarkGlyphSets = $this->GDEFdata[$this->fontkey]['MarkGlyphSets'] = $MarkGlyphSets;
$this->GlyphClassMarks = $this->GDEFdata[$this->fontkey]['GlyphClassMarks'] = $GlyphClassMarks;
$this->GlyphClassLigatures = $this->GDEFdata[$this->fontkey]['GlyphClassLigatures'] = $GlyphClassLigatures;
$this->GlyphClassComponents = $this->GDEFdata[$this->fontkey]['GlyphClassComponents'] = $GlyphClassComponents;
$this->GlyphClassBases = $this->GDEFdata[$this->fontkey]['GlyphClassBases'] = $GlyphClassBases;
} else {
$this->GSUB_offset = $this->GDEFdata[$this->fontkey]['GSUB_offset'];
$this->GPOS_offset = $this->GDEFdata[$this->fontkey]['GPOS_offset'];
$this->GSUB_length = $this->GDEFdata[$this->fontkey]['GSUB_length'];
$this->MarkAttachmentType = $this->GDEFdata[$this->fontkey]['MarkAttachmentType'];
$this->MarkGlyphSets = $this->GDEFdata[$this->fontkey]['MarkGlyphSets'];
$this->GlyphClassMarks = $this->GDEFdata[$this->fontkey]['GlyphClassMarks'];
$this->GlyphClassLigatures = $this->GDEFdata[$this->fontkey]['GlyphClassLigatures'];
$this->GlyphClassComponents = $this->GDEFdata[$this->fontkey]['GlyphClassComponents'];
$this->GlyphClassBases = $this->GDEFdata[$this->fontkey]['GlyphClassBases'];
}
 
// 2. Prepare string as HEX string and Analyse character properties
//=================================================================
$earr = $this->mpdf->UTF8StringToArray($str, false);
 
$scriptblock = 0;
$scriptblocks = [];
$scriptblocks[0] = 0;
$vstr = '';
$OTLdata = [];
$subchunk = 0;
$charctr = 0;
foreach ($earr as $char) {
$ucd_record = Ucdn::get_ucd_record($char);
$sbl = $ucd_record[6];
 
// Special case - Arabic End of Ayah
if ($char == 1757) {
$sbl = Ucdn::SCRIPT_ARABIC;
}
 
if ($sbl && $sbl != 40 && $sbl != 102) {
if ($scriptblock == 0) {
$scriptblock = $sbl;
$scriptblocks[$subchunk] = $scriptblock;
} elseif ($scriptblock > 0 && $scriptblock != $sbl) {
// *************************************************
// NEW (non-common) Script encountered in this chunk. Start a new subchunk
$subchunk++;
$scriptblock = $sbl;
$charctr = 0;
$scriptblocks[$subchunk] = $scriptblock;
}
}
 
$OTLdata[$subchunk][$charctr]['general_category'] = $ucd_record[0];
$OTLdata[$subchunk][$charctr]['bidi_type'] = $ucd_record[2];
 
//$OTLdata[$subchunk][$charctr]['combining_class'] = $ucd_record[1];
//$OTLdata[$subchunk][$charctr]['bidi_type'] = $ucd_record[2];
//$OTLdata[$subchunk][$charctr]['mirrored'] = $ucd_record[3];
//$OTLdata[$subchunk][$charctr]['east_asian_width'] = $ucd_record[4];
//$OTLdata[$subchunk][$charctr]['normalization_check'] = $ucd_record[5];
//$OTLdata[$subchunk][$charctr]['script'] = $ucd_record[6];
 
$charasstr = $this->unicode_hex($char);
 
if (strpos($this->GlyphClassMarks, $charasstr) !== false) {
$OTLdata[$subchunk][$charctr]['group'] = 'M';
} elseif ($char == 32 || $char == 12288) {
$OTLdata[$subchunk][$charctr]['group'] = 'S';
} // 12288 = 0x3000 = CJK space
else {
$OTLdata[$subchunk][$charctr]['group'] = 'C';
}
 
$OTLdata[$subchunk][$charctr]['uni'] = $char;
$OTLdata[$subchunk][$charctr]['hex'] = $charasstr;
$charctr++;
}
 
/* PROCESS EACH SUBCHUNK WITH DIFFERENT SCRIPTS */
for ($sch = 0; $sch <= $subchunk; $sch++) {
$this->OTLdata = $OTLdata[$sch];
$scriptblock = $scriptblocks[$sch];
 
// 3. Get Appropriate Scripts, and Shaper engine from analysing text and list of available scripts/langsys in font
//==============================
// Based on actual script block of text, select shaper (and line-breaking dictionaries)
if (Ucdn::SCRIPT_DEVANAGARI <= $scriptblock && $scriptblock <= Ucdn::SCRIPT_MALAYALAM) {
$this->shaper = "I";
} // INDIC shaper
elseif ($scriptblock == Ucdn::SCRIPT_ARABIC || $scriptblock == Ucdn::SCRIPT_SYRIAC) {
$this->shaper = "A";
} // ARABIC shaper
elseif ($scriptblock == Ucdn::SCRIPT_NKO || $scriptblock == Ucdn::SCRIPT_MANDAIC) {
$this->shaper = "A";
} // ARABIC shaper
elseif ($scriptblock == Ucdn::SCRIPT_KHMER) {
$this->shaper = "K";
} // KHMER shaper
elseif ($scriptblock == Ucdn::SCRIPT_THAI) {
$this->shaper = "T";
} // THAI shaper
elseif ($scriptblock == Ucdn::SCRIPT_LAO) {
$this->shaper = "L";
} // LAO shaper
elseif ($scriptblock == Ucdn::SCRIPT_SINHALA) {
$this->shaper = "S";
} // SINHALA shaper
elseif ($scriptblock == Ucdn::SCRIPT_MYANMAR) {
$this->shaper = "M";
} // MYANMAR shaper
elseif ($scriptblock == Ucdn::SCRIPT_NEW_TAI_LUE) {
$this->shaper = "E";
} // SEA South East Asian shaper
elseif ($scriptblock == Ucdn::SCRIPT_CHAM) {
$this->shaper = "E";
} // SEA South East Asian shaper
elseif ($scriptblock == Ucdn::SCRIPT_TAI_THAM) {
$this->shaper = "E";
} // SEA South East Asian shaper
else {
$this->shaper = "";
}
// Get scripttag based on actual text script
$scripttag = Ucdn::$uni_scriptblock[$scriptblock];
 
$GSUBscriptTag = '';
$GSUBlangsys = '';
$GPOSscriptTag = '';
$GPOSlangsys = '';
$is_old_spec = false;
 
$ScriptLang = $this->mpdf->CurrentFont['GSUBScriptLang'];
if (count($ScriptLang)) {
list($GSUBscriptTag, $is_old_spec) = $this->_getOTLscriptTag($ScriptLang, $scripttag, $scriptblock, $this->shaper, $useOTL, 'GSUB');
if ($this->mpdf->fontLanguageOverride && strpos($ScriptLang[$GSUBscriptTag], $this->mpdf->fontLanguageOverride) !== false) {
$GSUBlangsys = str_pad($this->mpdf->fontLanguageOverride, 4);
} elseif ($GSUBscriptTag && isset($ScriptLang[$GSUBscriptTag]) && $ScriptLang[$GSUBscriptTag] != '') {
$GSUBlangsys = $this->_getOTLLangTag($this->mpdf->currentLang, $ScriptLang[$GSUBscriptTag]);
}
}
$ScriptLang = $this->mpdf->CurrentFont['GPOSScriptLang'];
 
// NB If after GSUB, the same script/lang exist for GPOS, just use these...
if ($GSUBscriptTag && $GSUBlangsys && isset($ScriptLang[$GSUBscriptTag]) && strpos($ScriptLang[$GSUBscriptTag], $GSUBlangsys) !== false) {
$GPOSlangsys = $GSUBlangsys;
$GPOSscriptTag = $GSUBscriptTag;
} // else repeat for GPOS
// [Font XBRiyaz has GSUB tables for latn, but not GPOS for latn]
elseif (count($ScriptLang)) {
list($GPOSscriptTag, $dummy) = $this->_getOTLscriptTag($ScriptLang, $scripttag, $scriptblock, $this->shaper, $useOTL, 'GPOS');
if ($GPOSscriptTag && $this->mpdf->fontLanguageOverride && strpos($ScriptLang[$GPOSscriptTag], $this->mpdf->fontLanguageOverride) !== false) {
$GPOSlangsys = str_pad($this->mpdf->fontLanguageOverride, 4);
} elseif ($GPOSscriptTag && isset($ScriptLang[$GPOSscriptTag]) && $ScriptLang[$GPOSscriptTag] != '') {
$GPOSlangsys = $this->_getOTLLangTag($this->mpdf->currentLang, $ScriptLang[$GPOSscriptTag]);
}
}
 
////////////////////////////////////////////////////////////////
// This is just for the font_dump_OTL utility to set script and langsys override
if (isset($this->mpdf->overrideOTLsettings) && isset($this->mpdf->overrideOTLsettings[$this->fontkey])) {
$GSUBscriptTag = $GPOSscriptTag = $this->mpdf->overrideOTLsettings[$this->fontkey]['script'];
$GSUBlangsys = $GPOSlangsys = $this->mpdf->overrideOTLsettings[$this->fontkey]['lang'];
}
////////////////////////////////////////////////////////////////
 
if (!$GSUBscriptTag && !$GSUBlangsys && !$GPOSscriptTag && !$GPOSlangsys) {
// Remove ZWJ and ZWNJ
for ($i = 0; $i < count($this->OTLdata); $i++) {
if ($this->OTLdata[$i]['uni'] == 8204 || $this->OTLdata[$i]['uni'] == 8205) {
array_splice($this->OTLdata, $i, 1);
}
}
$this->schOTLdata[$sch] = $this->OTLdata;
$this->OTLdata = [];
continue;
}
 
// Don't use MYANMAR shaper unless using v2 scripttag
if ($this->shaper == 'M' && $GSUBscriptTag != 'mym2') {
$this->shaper = '';
}
 
$GSUBFeatures = (isset($this->mpdf->CurrentFont['GSUBFeatures'][$GSUBscriptTag][$GSUBlangsys]) ? $this->mpdf->CurrentFont['GSUBFeatures'][$GSUBscriptTag][$GSUBlangsys] : false);
$GPOSFeatures = (isset($this->mpdf->CurrentFont['GPOSFeatures'][$GPOSscriptTag][$GPOSlangsys]) ? $this->mpdf->CurrentFont['GPOSFeatures'][$GPOSscriptTag][$GPOSlangsys] : false);
 
$this->assocLigs = []; // Ligatures[$posarr lpos] => nc
$this->assocMarks = []; // assocMarks[$posarr mpos] => array(compID, ligPos)
 
if (!isset($this->GDEFdata[$this->fontkey]['GSUBGPOStables'])) {
$this->ttfOTLdata = $this->GDEFdata[$this->fontkey]['GSUBGPOStables'] = $this->fontCache->load($this->fontkey . '.GSUBGPOStables.dat', 'rb');
if (!$this->ttfOTLdata) {
throw new MpdfException('Can\'t open file ' . $this->fontCache->tempFilename($this->fontkey . '.GSUBGPOStables.dat'));
}
} else {
$this->ttfOTLdata = $this->GDEFdata[$this->fontkey]['GSUBGPOStables'];
}
 
if ($this->debugOTL) {
$this->_dumpproc('BEGIN', '-', '-', '-', '-', -1, '-', 0);
}
 
////////////////////////////////////////////////////////////////
///////// LINE BREAKING FOR KHMER, THAI + LAO /////////////////
////////////////////////////////////////////////////////////////
// Insert U+200B at word boundaries using dictionaries
if ($this->mpdf->useDictionaryLBR && ($this->shaper == "K" || $this->shaper == "T" || $this->shaper == "L")) {
// Sets $this->OTLdata[$i]['wordend']=true at possible end of word boundaries
$this->seaLineBreaking();
} // Insert U+200B at word boundaries for Tibetan
elseif ($this->mpdf->useTibetanLBR && $scriptblock == Ucdn::SCRIPT_TIBETAN) {
// Sets $this->OTLdata[$i]['wordend']=true at possible end of word boundaries
$this->tibetanLineBreaking();
}
 
 
////////////////////////////////////////////////////////////////
////////// GSUB /////////////////////////////////
////////////////////////////////////////////////////////////////
if (($useOTL & 0xFF) && $GSUBscriptTag && $GSUBlangsys && $GSUBFeatures) {
// 4. Load GSUB data, Coverage & Lookups
//=================================================================
 
$this->GSUBfont = $this->fontkey . '.GSUB.' . $GSUBscriptTag . '.' . $GSUBlangsys;
 
if (!isset($this->GSUBdata[$this->GSUBfont])) {
if ($this->fontCache->has($this->mpdf->CurrentFont['fontkey'] . '.GSUB.' . $GSUBscriptTag . '.' . $GSUBlangsys . '.php')) {
include $this->fontCache->tempFilename($this->mpdf->CurrentFont['fontkey'] . '.GSUB.' . $GSUBscriptTag . '.' . $GSUBlangsys . '.php');
$this->GSUBdata[$this->GSUBfont]['rtlSUB'] = $rtlSUB;
$this->GSUBdata[$this->GSUBfont]['finals'] = $finals;
if ($this->shaper == 'I') {
$this->GSUBdata[$this->GSUBfont]['rphf'] = $rphf;
$this->GSUBdata[$this->GSUBfont]['half'] = $half;
$this->GSUBdata[$this->GSUBfont]['pref'] = $pref;
$this->GSUBdata[$this->GSUBfont]['blwf'] = $blwf;
$this->GSUBdata[$this->GSUBfont]['pstf'] = $pstf;
}
} else {
$this->GSUBdata[$this->GSUBfont] = ['rtlSUB' => [], 'rphf' => [], 'rphf' => [],
'pref' => [], 'blwf' => [], 'pstf' => [], 'finals' => ''
];
}
}
 
if (!isset($this->GSUBdata[$this->fontkey])) {
include $this->fontCache->tempFilename($this->fontkey . '.GSUBdata.php');
$this->GSLuCoverage = $this->GSUBdata[$this->fontkey]['GSLuCoverage'] = $GSLuCoverage;
} else {
$this->GSLuCoverage = $this->GSUBdata[$this->fontkey]['GSLuCoverage'];
}
 
$this->GSUBLookups = $this->mpdf->CurrentFont['GSUBLookups'];
 
 
// 5(A). GSUB - Shaper - ARABIC
//==============================
if ($this->shaper == 'A') {
//-----------------------------------------------------------------------------------
// a. Apply initial GSUB Lookups (in order specified in lookup list but only selecting from certain tags)
//-----------------------------------------------------------------------------------
$tags = 'locl ccmp';
$omittags = '';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, true);
}
$this->_applyGSUBrules($usetags, $GSUBscriptTag, $GSUBlangsys);
 
//-----------------------------------------------------------------------------------
// b. Apply context-specific forms GSUB Lookups (initial, isolated, medial, final)
//-----------------------------------------------------------------------------------
// Arab and Syriac are the only scripts requiring the special joining - which takes the place of
// isol fina medi init rules in GSUB (+ fin2 fin3 med2 in Syriac syrc)
$tags = 'isol fina fin2 fin3 medi med2 init';
$omittags = '';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, true);
}
 
$this->arabGlyphs = $this->GSUBdata[$this->GSUBfont]['rtlSUB'];
 
$gcms = explode("| ", $this->GlyphClassMarks);
$gcm = [];
foreach ($gcms as $g) {
$gcm[hexdec($g)] = 1;
}
$this->arabTransparentJoin = $this->arabTransparent + $gcm;
$this->arabic_shaper($usetags, $GSUBscriptTag);
 
//-----------------------------------------------------------------------------------
// c. Set Kashida points (after joining occurred - medi, fina, init) but before other substitutions
//-----------------------------------------------------------------------------------
//if ($scriptblock == Ucdn::SCRIPT_ARABIC ) {
for ($i = 0; $i < count($this->OTLdata); $i++) {
// Put the kashida marker on the character BEFORE which is inserted the kashida
// Kashida marker is inverse of priority i.e. Priority 1 => 7, Priority 7 => 1.
// Priority 1 User-inserted Kashida 0640 = Tatweel
// The user entered a Kashida in a position
// Position: Before the user-inserted kashida
if ($this->OTLdata[$i]['uni'] == 0x0640) {
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 8; // Put before the next character
} // Priority 2 Seen (0633) FEB3, FEB4; Sad (0635) FEBB, FEBC
// Initial or medial form
// Connecting to the next character
// Position: After the character
elseif ($this->OTLdata[$i]['uni'] == 0xFEB3 || $this->OTLdata[$i]['uni'] == 0xFEB4 || $this->OTLdata[$i]['uni'] == 0xFEBB || $this->OTLdata[$i]['uni'] == 0xFEBC) {
$checkpos = $i + 1;
while (isset($this->OTLdata[$checkpos]) && strpos($this->GlyphClassMarks, $this->OTLdata[$checkpos]['hex']) !== false) {
$checkpos++;
}
if (isset($this->OTLdata[$checkpos])) {
$this->OTLdata[$checkpos]['GPOSinfo']['kashida'] = 7; // Put after marks on next character
}
} // Priority 3 Taa Marbutah (0629) FE94; Haa (062D) FEA2; Dal (062F) FEAA
// Final form
// Connecting to previous character
// Position: Before the character
elseif ($this->OTLdata[$i]['uni'] == 0xFE94 || $this->OTLdata[$i]['uni'] == 0xFEA2 || $this->OTLdata[$i]['uni'] == 0xFEAA) {
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 6;
} // Priority 4 Alef (0627) FE8E; Tah (0637) FEC2; Lam (0644) FEDE; Kaf (0643) FEDA; Gaf (06AF) FB93
// Final form
// Connecting to previous character
// Position: Before the character
elseif ($this->OTLdata[$i]['uni'] == 0xFE8E || $this->OTLdata[$i]['uni'] == 0xFEC2 || $this->OTLdata[$i]['uni'] == 0xFEDE || $this->OTLdata[$i]['uni'] == 0xFEDA || $this->OTLdata[$i]['uni'] == 0xFB93) {
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 5;
} // Priority 5 RA (0631) FEAE; Ya (064A) FEF2 FEF4; Alef Maqsurah (0649) FEF0 FBE9
// Final or Medial form
// Connected to preceding medial BAA (0628) = FE92
// Position: Before preceding medial Baa
// Although not mentioned in spec, added Farsi Yeh (06CC) FBFD FBFF; equivalent to 064A or 0649
elseif ($this->OTLdata[$i]['uni'] == 0xFEAE || $this->OTLdata[$i]['uni'] == 0xFEF2 || $this->OTLdata[$i]['uni'] == 0xFEF0 || $this->OTLdata[$i]['uni'] == 0xFEF4 || $this->OTLdata[$i]['uni'] == 0xFBE9 || $this->OTLdata[$i]['uni'] == 0xFBFD || $this->OTLdata[$i]['uni'] == 0xFBFF
) {
$checkpos = $i - 1;
while (isset($this->OTLdata[$checkpos]) && strpos($this->GlyphClassMarks, $this->OTLdata[$checkpos]['hex']) !== false) {
$checkpos--;
}
if (isset($this->OTLdata[$checkpos]) && $this->OTLdata[$checkpos]['uni'] == 0xFE92) {
$this->OTLdata[$checkpos]['GPOSinfo']['kashida'] = 4; // ******* Before preceding BAA
}
} // Priority 6 WAW (0648) FEEE; Ain (0639) FECA; Qaf (0642) FED6; Fa (0641) FED2
// Final form
// Connecting to previous character
// Position: Before the character
elseif ($this->OTLdata[$i]['uni'] == 0xFEEE || $this->OTLdata[$i]['uni'] == 0xFECA || $this->OTLdata[$i]['uni'] == 0xFED6 || $this->OTLdata[$i]['uni'] == 0xFED2) {
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 3;
}
 
// Priority 7 Other connecting characters
// Final form
// Connecting to previous character
// Position: Before the character
/* This isn't in the spec, but using MS WORD as a basis, give a lower priority to the 3 characters already checked
 in (5) above. Test case:
 &#x62e;&#x652;&#x631;&#x64e;&#x649;&#x670;
 &#x641;&#x64e;&#x62a;&#x64f;&#x630;&#x64e;&#x643;&#x651;&#x650;&#x631;
 */
 
if (!isset($this->OTLdata[$i]['GPOSinfo']['kashida'])) {
if (strpos($this->GSUBdata[$this->GSUBfont]['finals'], $this->OTLdata[$i]['hex']) !== false) { // ANY OTHER FINAL FORM
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 2;
} elseif (strpos('0FEAE 0FEF0 0FEF2', $this->OTLdata[$i]['hex']) !== false) { // not already included in 5 above
$this->OTLdata[$i]['GPOSinfo']['kashida'] = 1;
}
}
}
 
//-----------------------------------------------------------------------------------
// d. Apply Presentation Forms GSUB Lookups (+ any discretionary) - Apply one at a time in Feature order
//-----------------------------------------------------------------------------------
$tags = 'rlig calt liga clig mset';
 
$omittags = 'locl ccmp nukt akhn rphf rkrf pref blwf abvf half pstf cfar vatu cjct init medi fina isol med2 fin2 fin3 ljmo vjmo tjmo';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, false);
}
 
$ts = explode(' ', $usetags);
foreach ($ts as $ut) { // - Apply one at a time in Feature order
$this->_applyGSUBrules($ut, $GSUBscriptTag, $GSUBlangsys);
}
//-----------------------------------------------------------------------------------
// e. NOT IN SPEC
// If space precedes a mark -> substitute a &nbsp; before the Mark, to prevent line breaking Test:
//-----------------------------------------------------------------------------------
for ($ptr = 1; $ptr < count($this->OTLdata); $ptr++) {
if ($this->OTLdata[$ptr]['general_category'] == Ucdn::UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK && $this->OTLdata[$ptr - 1]['uni'] == 32) {
$this->OTLdata[$ptr - 1]['uni'] = 0xa0;
$this->OTLdata[$ptr - 1]['hex'] = '000A0';
}
}
} // 5(I). GSUB - Shaper - INDIC and SINHALA and KHMER
//===================================
elseif ($this->shaper == 'I' || $this->shaper == 'K' || $this->shaper == 'S') {
$this->restrictToSyllable = true;
//-----------------------------------------------------------------------------------
// a. First decompose/compose split mattras
// (normalize) ??????? Nukta/Halant order etc ??????????????????????????????????????????????????????????????????????????
//-----------------------------------------------------------------------------------
for ($ptr = 0; $ptr < count($this->OTLdata); $ptr++) {
$char = $this->OTLdata[$ptr]['uni'];
$sub = Indic::decompose_indic($char);
if ($sub) {
$newinfo = [];
for ($i = 0; $i < count($sub); $i++) {
$newinfo[$i] = [];
$ucd_record = Ucdn::get_ucd_record($sub[$i]);
$newinfo[$i]['general_category'] = $ucd_record[0];
$newinfo[$i]['bidi_type'] = $ucd_record[2];
$charasstr = $this->unicode_hex($sub[$i]);
if (strpos($this->GlyphClassMarks, $charasstr) !== false) {
$newinfo[$i]['group'] = 'M';
} else {
$newinfo[$i]['group'] = 'C';
}
$newinfo[$i]['uni'] = $sub[$i];
$newinfo[$i]['hex'] = $charasstr;
}
array_splice($this->OTLdata, $ptr, 1, $newinfo);
$ptr += count($sub) - 1;
}
/* Only Composition-exclusion exceptions that we want to recompose. */
if ($this->shaper == 'I') {
if ($char == 0x09AF && isset($this->OTLdata[$ptr + 1]) && $this->OTLdata[$ptr + 1]['uni'] == 0x09BC) {
$sub = 0x09DF;
$newinfo = [];
$newinfo[0] = [];
$ucd_record = Ucdn::get_ucd_record($sub);
$newinfo[0]['general_category'] = $ucd_record[0];
$newinfo[0]['bidi_type'] = $ucd_record[2];
$newinfo[0]['group'] = 'C';
$newinfo[0]['uni'] = $sub;
$newinfo[0]['hex'] = $this->unicode_hex($sub);
array_splice($this->OTLdata, $ptr, 2, $newinfo);
}
}
}
//-----------------------------------------------------------------------------------
// b. Analyse characters - group as syllables/clusters (Indic); invalid diacritics; add dotted circle
//-----------------------------------------------------------------------------------
$indic_category_string = '';
foreach ($this->OTLdata as $eid => $c) {
Indic::set_indic_properties($this->OTLdata[$eid], $scriptblock); // sets ['indic_category'] and ['indic_position']
//$c['general_category']
//$c['combining_class']
//$c['uni'] = $char;
 
$indic_category_string .= Indic::$indic_category_char[$this->OTLdata[$eid]['indic_category']];
}
 
$broken_syllables = false;
if ($this->shaper == 'I') {
Indic::set_syllables($this->OTLdata, $indic_category_string, $broken_syllables);
} elseif ($this->shaper == 'S') {
Indic::set_syllables_sinhala($this->OTLdata, $indic_category_string, $broken_syllables);
} elseif ($this->shaper == 'K') {
Indic::set_syllables_khmer($this->OTLdata, $indic_category_string, $broken_syllables);
}
$indic_category_string = '';
 
//-----------------------------------------------------------------------------------
// c. Initial Re-ordering (Indic / Khmer / Sinhala)
//-----------------------------------------------------------------------------------
// Find base consonant
// Decompose/compose and reorder Matras
// Reorder marks to canonical order
 
$indic_config = Indic::$indic_configs[$scriptblock];
$dottedcircle = false;
if ($broken_syllables) {
if ($this->mpdf->_charDefined($this->mpdf->fonts[$this->fontkey]['cw'], 0x25CC)) {
$dottedcircle = [];
$ucd_record = Ucdn::get_ucd_record(0x25CC);
$dottedcircle[0]['general_category'] = $ucd_record[0];
$dottedcircle[0]['bidi_type'] = $ucd_record[2];
$dottedcircle[0]['group'] = 'C';
$dottedcircle[0]['uni'] = 0x25CC;
$dottedcircle[0]['indic_category'] = Indic::OT_DOTTEDCIRCLE;
$dottedcircle[0]['indic_position'] = Indic::POS_BASE_C;
 
$dottedcircle[0]['hex'] = '025CC'; // TEMPORARY *****
}
}
Indic::initial_reordering($this->OTLdata, $this->GSUBdata[$this->GSUBfont], $broken_syllables, $indic_config, $scriptblock, $is_old_spec, $dottedcircle);
 
//-----------------------------------------------------------------------------------
// d. Apply initial and basic shaping forms GSUB Lookups (one at a time)
//-----------------------------------------------------------------------------------
if ($this->shaper == 'I' || $this->shaper == 'S') {
$tags = 'locl ccmp nukt akhn rphf rkrf pref blwf half pstf vatu cjct';
} elseif ($this->shaper == 'K') {
$tags = 'locl ccmp pref blwf abvf pstf cfar';
}
$this->_applyGSUBrulesIndic($tags, $GSUBscriptTag, $GSUBlangsys, $is_old_spec);
 
//-----------------------------------------------------------------------------------
// e. Final Re-ordering (Indic / Khmer / Sinhala)
//-----------------------------------------------------------------------------------
// Reorder matras
// Reorder reph
// Reorder pre-base reordering consonants:
 
Indic::final_reordering($this->OTLdata, $this->GSUBdata[$this->GSUBfont], $indic_config, $scriptblock, $is_old_spec);
 
//-----------------------------------------------------------------------------------
// f. Apply 'init' feature to first syllable in word (indicated by ['mask']) Indic::FLAG(Indic::INIT);
//-----------------------------------------------------------------------------------
if ($this->shaper == 'I' || $this->shaper == 'S') {
$tags = 'init';
$this->_applyGSUBrulesIndic($tags, $GSUBscriptTag, $GSUBlangsys, $is_old_spec);
}
 
//-----------------------------------------------------------------------------------
// g. Apply Presentation Forms GSUB Lookups (+ any discretionary)
//-----------------------------------------------------------------------------------
$tags = 'pres abvs blws psts haln rlig calt liga clig mset';
 
$omittags = 'locl ccmp nukt akhn rphf rkrf pref blwf abvf half pstf cfar vatu cjct init medi fina isol med2 fin2 fin3 ljmo vjmo tjmo';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, false);
}
if ($this->shaper == 'K') { // Features are applied one at a time, working through each codepoint
$this->_applyGSUBrulesSingly($usetags, $GSUBscriptTag, $GSUBlangsys);
} else {
$this->_applyGSUBrules($usetags, $GSUBscriptTag, $GSUBlangsys);
}
$this->restrictToSyllable = false;
} // 5(M). GSUB - Shaper - MYANMAR (ONLY mym2)
//==============================
// NB Old style 'mymr' is left to go through the default shaper
elseif ($this->shaper == 'M') {
$this->restrictToSyllable = true;
//-----------------------------------------------------------------------------------
// a. Analyse characters - group as syllables/clusters (Myanmar); invalid diacritics; add dotted circle
//-----------------------------------------------------------------------------------
$myanmar_category_string = '';
foreach ($this->OTLdata as $eid => $c) {
Myanmar::set_myanmar_properties($this->OTLdata[$eid]); // sets ['myanmar_category'] and ['myanmar_position']
$myanmar_category_string .= Myanmar::$myanmar_category_char[$this->OTLdata[$eid]['myanmar_category']];
}
$broken_syllables = false;
Myanmar::set_syllables($this->OTLdata, $myanmar_category_string, $broken_syllables);
$myanmar_category_string = '';
 
//-----------------------------------------------------------------------------------
// b. Re-ordering (Myanmar mym2)
//-----------------------------------------------------------------------------------
$dottedcircle = false;
if ($broken_syllables) {
if ($this->mpdf->_charDefined($this->mpdf->fonts[$this->fontkey]['cw'], 0x25CC)) {
$dottedcircle = [];
$ucd_record = Ucdn::get_ucd_record(0x25CC);
$dottedcircle[0]['general_category'] = $ucd_record[0];
$dottedcircle[0]['bidi_type'] = $ucd_record[2];
$dottedcircle[0]['group'] = 'C';
$dottedcircle[0]['uni'] = 0x25CC;
$dottedcircle[0]['myanmar_category'] = Myanmar::OT_DOTTEDCIRCLE;
$dottedcircle[0]['myanmar_position'] = Myanmar::POS_BASE_C;
$dottedcircle[0]['hex'] = '025CC';
}
}
Myanmar::reordering($this->OTLdata, $this->GSUBdata[$this->GSUBfont], $broken_syllables, $dottedcircle);
 
//-----------------------------------------------------------------------------------
// c. Apply initial and basic shaping forms GSUB Lookups (one at a time)
//-----------------------------------------------------------------------------------
 
$tags = 'locl ccmp rphf pref blwf pstf';
$this->_applyGSUBrulesMyanmar($tags, $GSUBscriptTag, $GSUBlangsys);
 
//-----------------------------------------------------------------------------------
// d. Apply Presentation Forms GSUB Lookups (+ any discretionary)
//-----------------------------------------------------------------------------------
$tags = 'pres abvs blws psts haln rlig calt liga clig mset';
$omittags = 'locl ccmp nukt akhn rphf rkrf pref blwf abvf half pstf cfar vatu cjct init medi fina isol med2 fin2 fin3 ljmo vjmo tjmo';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, false);
}
$this->_applyGSUBrules($usetags, $GSUBscriptTag, $GSUBlangsys);
$this->restrictToSyllable = false;
} // 5(E). GSUB - Shaper - SEA South East Asian (New Tai Lue, Cham, Tai Tam)
//==============================
elseif ($this->shaper == 'E') {
/* HarfBuzz says: If the designer designed the font for the 'DFLT' script,
 * use the default shaper. Otherwise, use the SEA shaper.
 * Note that for some simple scripts, there may not be *any*
 * GSUB/GPOS needed, so there may be no scripts found! */
 
$this->restrictToSyllable = true;
//-----------------------------------------------------------------------------------
// a. Analyse characters - group as syllables/clusters (Indic); invalid diacritics; add dotted circle
//-----------------------------------------------------------------------------------
$sea_category_string = '';
foreach ($this->OTLdata as $eid => $c) {
Sea::set_sea_properties($this->OTLdata[$eid], $scriptblock); // sets ['sea_category'] and ['sea_position']
//$c['general_category']
//$c['combining_class']
//$c['uni'] = $char;
 
$sea_category_string .= Sea::$sea_category_char[$this->OTLdata[$eid]['sea_category']];
}
 
$broken_syllables = false;
Sea::set_syllables($this->OTLdata, $sea_category_string, $broken_syllables);
$sea_category_string = '';
 
//-----------------------------------------------------------------------------------
// b. Apply locl and ccmp shaping forms - before initial re-ordering; GSUB Lookups (one at a time)
//-----------------------------------------------------------------------------------
$tags = 'locl ccmp';
$this->_applyGSUBrulesSingly($tags, $GSUBscriptTag, $GSUBlangsys);
 
//-----------------------------------------------------------------------------------
// c. Initial Re-ordering
//-----------------------------------------------------------------------------------
// Find base consonant
// Decompose/compose and reorder Matras
// Reorder marks to canonical order
 
$dottedcircle = false;
if ($broken_syllables) {
if ($this->mpdf->_charDefined($this->mpdf->fonts[$this->fontkey]['cw'], 0x25CC)) {
$dottedcircle = [];
$ucd_record = Ucdn::get_ucd_record(0x25CC);
$dottedcircle[0]['general_category'] = $ucd_record[0];
$dottedcircle[0]['bidi_type'] = $ucd_record[2];
$dottedcircle[0]['group'] = 'C';
$dottedcircle[0]['uni'] = 0x25CC;
$dottedcircle[0]['sea_category'] = Sea::OT_GB;
$dottedcircle[0]['sea_position'] = Sea::POS_BASE_C;
 
$dottedcircle[0]['hex'] = '025CC'; // TEMPORARY *****
}
}
Sea::initial_reordering($this->OTLdata, $this->GSUBdata[$this->GSUBfont], $broken_syllables, $scriptblock, $dottedcircle);
 
//-----------------------------------------------------------------------------------
// d. Apply basic shaping forms GSUB Lookups (one at a time)
//-----------------------------------------------------------------------------------
$tags = 'pref abvf blwf pstf';
$this->_applyGSUBrulesSingly($tags, $GSUBscriptTag, $GSUBlangsys);
 
//-----------------------------------------------------------------------------------
// e. Final Re-ordering
//-----------------------------------------------------------------------------------
 
Sea::final_reordering($this->OTLdata, $this->GSUBdata[$this->GSUBfont], $scriptblock);
 
//-----------------------------------------------------------------------------------
// f. Apply Presentation Forms GSUB Lookups (+ any discretionary)
//-----------------------------------------------------------------------------------
$tags = 'pres abvs blws psts';
 
$omittags = 'locl ccmp nukt akhn rphf rkrf pref blwf abvf half pstf cfar vatu cjct init medi fina isol med2 fin2 fin3 ljmo vjmo tjmo';
$usetags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$usetags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, false);
}
$this->_applyGSUBrules($usetags, $GSUBscriptTag, $GSUBlangsys);
$this->restrictToSyllable = false;
} // 5(D). GSUB - Shaper - DEFAULT (including THAI and LAO and MYANMAR v1 [mymr] and TIBETAN)
//==============================
else { // DEFAULT
//-----------------------------------------------------------------------------------
// a. First decompose/compose in Thai / Lao - Tibetan
//-----------------------------------------------------------------------------------
// Decomposition for THAI or LAO
/* This function implements the shaping logic documented here:
 *
 * http://linux.thai.net/~thep/th-otf/shaping.html
 *
 * The first shaping rule listed there is needed even if the font has Thai
 * OpenType tables.
 *
 *
 * The following is NOT specified in the MS OT Thai spec, however, it seems
 * to be what Uniscribe and other engines implement. According to Eric Muller:
 *
 * When you have a SARA AM, decompose it in NIKHAHIT + SARA AA, *and* move the
 * NIKHAHIT backwards over any tone mark (0E48-0E4B).
 *
 * <0E14, 0E4B, 0E33> -> <0E14, 0E4D, 0E4B, 0E32>
 *
 * This reordering is legit only when the NIKHAHIT comes from a SARA AM, not
 * when it's there to start with. The string <0E14, 0E4B, 0E4D> is probably
 * not what a user wanted, but the rendering is nevertheless nikhahit above
 * chattawa.
 *
 * Same for Lao.
 *
 * Thai Lao
 * SARA AM: U+0E33 U+0EB3
 * SARA AA: U+0E32 U+0EB2
 * Nikhahit: U+0E4D U+0ECD
 *
 * Testing shows that Uniscribe reorder the following marks:
 * Thai: <0E31,0E34..0E37,0E47..0E4E>
 * Lao: <0EB1,0EB4..0EB7,0EC7..0ECE>
 *
 * Lao versions are the same as Thai + 0x80.
 */
if ($this->shaper == 'T' || $this->shaper == 'L') {
for ($ptr = 0; $ptr < count($this->OTLdata); $ptr++) {
$char = $this->OTLdata[$ptr]['uni'];
if (($char & ~0x0080) == 0x0E33) { // if SARA_AM (U+0E33 or U+0EB3)
$NIKHAHIT = $char + 0x1A;
$SARA_AA = $char - 1;
$sub = [$SARA_AA, $NIKHAHIT];
 
$newinfo = [];
$ucd_record = Ucdn::get_ucd_record($sub[0]);
$newinfo[0]['general_category'] = $ucd_record[0];
$newinfo[0]['bidi_type'] = $ucd_record[2];
$charasstr = $this->unicode_hex($sub[0]);
if (strpos($this->GlyphClassMarks, $charasstr) !== false) {
$newinfo[0]['group'] = 'M';
} else {
$newinfo[0]['group'] = 'C';
}
$newinfo[0]['uni'] = $sub[0];
$newinfo[0]['hex'] = $charasstr;
$this->OTLdata[$ptr] = $newinfo[0]; // Substitute SARA_AM => SARA_AA
 
$ntones = 0; // number of (preceding) tone marks
// IS_TONE_MARK ((x) & ~0x0080, 0x0E34 - 0x0E37, 0x0E47 - 0x0E4E, 0x0E31)
while (isset($this->OTLdata[$ptr - 1 - $ntones]) && (
($this->OTLdata[$ptr - 1 - $ntones]['uni'] & ~0x0080) == 0x0E31 ||
(($this->OTLdata[$ptr - 1 - $ntones]['uni'] & ~0x0080) >= 0x0E34 &&
($this->OTLdata[$ptr - 1 - $ntones]['uni'] & ~0x0080) <= 0x0E37) ||
(($this->OTLdata[$ptr - 1 - $ntones]['uni'] & ~0x0080) >= 0x0E47 &&
($this->OTLdata[$ptr - 1 - $ntones]['uni'] & ~0x0080) <= 0x0E4E)
)
) {
$ntones++;
}
 
$newinfo = [];
$ucd_record = Ucdn::get_ucd_record($sub[1]);
$newinfo[0]['general_category'] = $ucd_record[0];
$newinfo[0]['bidi_type'] = $ucd_record[2];
$charasstr = $this->unicode_hex($sub[1]);
if (strpos($this->GlyphClassMarks, $charasstr) !== false) {
$newinfo[0]['group'] = 'M';
} else {
$newinfo[0]['group'] = 'C';
}
$newinfo[0]['uni'] = $sub[1];
$newinfo[0]['hex'] = $charasstr;
// Insert NIKAHIT
array_splice($this->OTLdata, $ptr - $ntones, 0, $newinfo);
 
$ptr++;
}
}
}
 
if ($scriptblock == Ucdn::SCRIPT_TIBETAN) {
// =========================
// Reordering TIBETAN
// =========================
// Tibetan does not need to need a shaper generally, as long as characters are presented in the correct order
// so we will do one minor change here:
// From ICU: If the present character is a number, and the next character is a pre-number combining mark
// then the two characters are reordered
// From MS OTL spec the following are Digit modifiers (Md): 0F18–0F19, 0F3E–0F3F
// Digits: 0F20–0F33
// On testing only 0x0F3F (pre-based mark) seems to need re-ordering
for ($ptr = 0; $ptr < count($this->OTLdata) - 1; $ptr++) {
if (Indic::in_range($this->OTLdata[$ptr]['uni'], 0x0F20, 0x0F33) && $this->OTLdata[$ptr + 1]['uni'] == 0x0F3F) {
$tmp = $this->OTLdata[$ptr + 1];
$this->OTLdata[$ptr + 1] = $this->OTLdata[$ptr];
$this->OTLdata[$ptr] = $tmp;
}
}
 
 
// =========================
// Decomposition for TIBETAN
// =========================
/* Recommended, but does not seem to change anything...
 for($ptr=0; $ptr<count($this->OTLdata); $ptr++) {
 $char = $this->OTLdata[$ptr]['uni'];
 $sub = Indic::decompose_indic($char);
 if ($sub) {
 $newinfo = array();
 for($i=0;$i<count($sub);$i++) {
 $newinfo[$i] = array();
 $ucd_record = Ucdn::get_ucd_record($sub[$i]);
 $newinfo[$i]['general_category'] = $ucd_record[0];
 $newinfo[$i]['bidi_type'] = $ucd_record[2];
 $charasstr = $this->unicode_hex($sub[$i]);
 if (strpos($this->GlyphClassMarks, $charasstr)!==false) { $newinfo[$i]['group'] = 'M'; }
 else { $newinfo[$i]['group'] = 'C'; }
 $newinfo[$i]['uni'] = $sub[$i];
 $newinfo[$i]['hex'] = $charasstr;
 }
 array_splice($this->OTLdata, $ptr, 1, $newinfo);
 $ptr += count($sub)-1;
 }
 }
 */
}
 
 
//-----------------------------------------------------------------------------------
// b. Apply all GSUB Lookups (in order specified in lookup list)
//-----------------------------------------------------------------------------------
$tags = 'locl ccmp pref blwf abvf pstf pres abvs blws psts haln rlig calt liga clig mset RQD';
// pref blwf abvf pstf required for Tibetan
// " RQD" is a non-standard tag in Garuda font - presumably intended to be used by default ? "ReQuireD"
// Being a 3 letter tag is non-standard, and does not allow it to be set by font-feature-settings
 
 
/* ?Add these until shapers witten?
 Hangul: ljmo vjmo tjmo
 */
 
$omittags = '';
$useGSUBtags = $tags;
if (!empty($this->mpdf->OTLtags)) {
$useGSUBtags = $this->_applyTagSettings($tags, $GSUBFeatures, $omittags, false);
}
// APPLY GSUB rules (as long as not Latin + SmallCaps - but not OTL smcp)
if (!(($this->mpdf->textvar & TextVars::FC_SMALLCAPS) && $scriptblock == Ucdn::SCRIPT_LATIN && strpos($useGSUBtags, 'smcp') === false)) {
$this->_applyGSUBrules($useGSUBtags, $GSUBscriptTag, $GSUBlangsys);
}
}
}
 
// Shapers - KHMER & THAI & LAO - Replace Word boundary marker with U+200B
// Also TIBETAN (no shaper)
//=======================================================
if (($this->shaper == "K" || $this->shaper == "T" || $this->shaper == "L") || $scriptblock == Ucdn::SCRIPT_TIBETAN) {
// Set up properties to insert a U+200B character
$newinfo = [];
//$newinfo[0] = array('general_category' => 1, 'bidi_type' => 14, 'group' => 'S', 'uni' => 0x200B, 'hex' => '0200B');
$newinfo[0] = [
'general_category' => Ucdn::UNICODE_GENERAL_CATEGORY_FORMAT,
'bidi_type' => Ucdn::BIDI_CLASS_BN,
'group' => 'S', 'uni' => 0x200B, 'hex' => '0200B'];
// Then insert U+200B at (after) all word end boundaries
for ($i = count($this->OTLdata) - 1; $i > 0; $i--) {
// Make sure after GSUB that wordend has not been moved - check next char is not in the same syllable
if (isset($this->OTLdata[$i]['wordend']) && $this->OTLdata[$i]['wordend'] &&
isset($this->OTLdata[$i + 1]['uni']) && (!isset($this->OTLdata[$i + 1]['syllable']) || !isset($this->OTLdata[$i + 1]['syllable']) || $this->OTLdata[$i + 1]['syllable'] != $this->OTLdata[$i]['syllable'])) {
array_splice($this->OTLdata, $i + 1, 0, $newinfo);
$this->_updateLigatureMarks($i, 1);
} elseif ($this->OTLdata[$i]['uni'] == 0x2e) { // Word end if Full-stop.
array_splice($this->OTLdata, $i + 1, 0, $newinfo);
$this->_updateLigatureMarks($i, 1);
}
}
}
 
 
// Shapers - INDIC & ARABIC & KHMER & SINHALA & MYANMAR - Remove ZWJ and ZWNJ
//=======================================================
if ($this->shaper == 'I' || $this->shaper == 'S' || $this->shaper == 'A' || $this->shaper == 'K' || $this->shaper == 'M') {
// Remove ZWJ and ZWNJ
for ($i = 0; $i < count($this->OTLdata); $i++) {
if ($this->OTLdata[$i]['uni'] == 8204 || $this->OTLdata[$i]['uni'] == 8205) {
array_splice($this->OTLdata, $i, 1);
$this->_updateLigatureMarks($i, -1);
}
}
}
 
 
////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////
////////// GPOS /////////////////////////////////
////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////
if (($useOTL & 0xFF) && $GPOSscriptTag && $GPOSlangsys && $GPOSFeatures) {
$this->Entry = [];
$this->Exit = [];