#------------------------------------------------------------------------------ # File: Font.pm # # Description: Read meta information from font files # # Revisions: 2010/01/15 - P. Harvey Created # # References: 1) http://developer.apple.com/textfonts/TTRefMan/RM06/Chap6.html # 2) http://www.microsoft.com/typography/otspec/otff.htm # 3) http://partners.adobe.com/public/developer/opentype/index_font_file.html # 4) http://partners.adobe.com/public/developer/en/font/5178.PFM.pdf # 5) http://opensource.adobe.com/svn/opensource/flex/sdk/trunk/modules/compiler/src/java/flex2/compiler/util/MimeMappings.java # 6) http://www.adobe.com/devnet/font/pdfs/5004.AFM_Spec.pdf # 7) https://www.w3.org/TR/WOFF/ # 8) https://www.w3.org/TR/WOFF2/ #------------------------------------------------------------------------------ package Image::ExifTool::Font; use strict; use vars qw($VERSION %ttLang); use Image::ExifTool qw(:DataAccess :Utils); use Image::ExifTool::XMP; $VERSION = '1.13'; sub ProcessOTF($$); # OTF tags to process (skip all others) my %processTag = ( name => 1, C2PA => 1 ); # TrueType 'name' platform codes my %ttPlatform = ( 0 => 'Unicode', 1 => 'Macintosh', 2 => 'ISO', 3 => 'Windows', 4 => 'Custom', ); # convert TrueType 'name' character encoding to ExifTool Charset (ref 1/2) my %ttCharset = ( Macintosh => { 0 => 'MacRoman', 17 => 'MacMalayalam', 1 => 'MacJapanese', 18 => 'MacSinhalese', 2 => 'MacChineseTW', 19 => 'MacBurmese', 3 => 'MacKorean', 20 => 'MacKhmer', 4 => 'MacArabic', 21 => 'MacThai', 5 => 'MacHebrew', 22 => 'MacLaotian', 6 => 'MacGreek', 23 => 'MacGeorgian', 7 => 'MacCyrillic', 24 => 'MacArmenian', # 7=Russian 8 => 'MacRSymbol', 25 => 'MacChineseCN', 9 => 'MacDevanagari', 26 => 'MacTibetan', 10 => 'MacGurmukhi', 27 => 'MacMongolian', 11 => 'MacGujarati', 28 => 'MacGeez', 12 => 'MacOriya', 29 => 'MacCyrillic', # 29=Slavic 13 => 'MacBengali', 30 => 'MacVietnam', 14 => 'MacTamil', 31 => 'MacSindhi', 15 => 'MacTelugu', 32 => '', # 32=uninterpreted 16 => 'MacKannada', }, Windows => { 0 => 'Symbol', 4 => 'Big5', 1 => 'UCS2', 5 => 'Wansung', 2 => 'ShiftJIS', 6 => 'Johab', 3 => 'PRC', 10 => 'UCS4', }, Unicode => { # (we don't currently handle the various Unicode flavours) 0 => 'UCS2', # Unicode 1.0 semantics 1 => 'UCS2', # Unicode 1.1 semantics 2 => 'UCS2', # ISO 10646 semantics 3 => 'UCS2', # Unicode 2.0 and onwards semantics, Unicode BMP only. 4 => 'UTF16',# Unicode 2.0 and onwards semantics, Unicode full repertoire. # 5 => Unicode Variation Sequences (not used in Naming table) }, ISO => { # (deprecated) 0 => 'UTF8', # (7-bit ASCII) 1 => 'UCS2', # ISO 10646 2 => 'Latin', # ISO 8859-1 }, Custom => { }, ); # convert TrueType 'name' language code to ExifTool language code %ttLang = ( # Macintosh language codes (also used by QuickTime.pm) # oddities: # 49 - Cyrillic version 83 - Roman # 50 - Arabic version 84 - Arabic # 146 - with dot above Macintosh => { 0 => 'en', 24 => 'lt', 48 => 'kk', 72 => 'ml', 129 => 'eu', 1 => 'fr', 25 => 'pl', 49 => 'az', 73 => 'kn', 130 => 'ca', 2 => 'de', 26 => 'hu', 50 => 'az', 74 => 'ta', 131 => 'la', 3 => 'it', 27 => 'et', 51 => 'hy', 75 => 'te', 132 => 'qu', 4 => 'nl-NL', 28 => 'lv', 52 => 'ka', 76 => 'si', 133 => 'gn', 5 => 'sv', 29 => 'smi', 53 => 'ro', 77 => 'my', 134 => 'ay', 6 => 'es', 30 => 'fo', 54 => 'ky', 78 => 'km', 135 => 'tt', 7 => 'da', 31 => 'fa', 55 => 'tg', 79 => 'lo', 136 => 'ug', 8 => 'pt', 32 => 'ru', 56 => 'tk', 80 => 'vi', 137 => 'dz', 9 => 'no', 33 => 'zh-CN', 57 => 'mn-MN', 81 => 'id', 138 => 'jv', 10 => 'he', 34 => 'nl-BE', 58 => 'mn-CN', 82 => 'tl', 139 => 'su', 11 => 'ja', 35 => 'ga', 59 => 'ps', 83 => 'ms-MY', 140 => 'gl', 12 => 'ar', 36 => 'sq', 60 => 'ku', 84 => 'ms-BN', 141 => 'af', 13 => 'fi', 37 => 'ro', 61 => 'ks', 85 => 'am', 142 => 'br', 14 => 'el', 38 => 'cs', 62 => 'sd', 86 => 'ti', 144 => 'gd', 15 => 'is', 39 => 'sk', 63 => 'bo', 87 => 'om', 145 => 'gv', 16 => 'mt', 40 => 'sl', 64 => 'ne', 88 => 'so', 146 => 'ga', 17 => 'tr', 41 => 'yi', 65 => 'sa', 89 => 'sw', 147 => 'to', 18 => 'hr', 42 => 'sr', 66 => 'mr', 90 => 'rw', 148 => 'el', 19 => 'zh-TW', 43 => 'mk', 67 => 'bn', 91 => 'rn', 149 => 'kl', 20 => 'ur', 44 => 'bg', 68 => 'as', 92 => 'ny', 150 => 'az', 21 => 'hi', 45 => 'uk', 69 => 'gu', 93 => 'mg', 22 => 'th', 46 => 'be', 70 => 'pa', 94 => 'eo', 23 => 'ko', 47 => 'uz', 71 => 'or', 128 => 'cy', }, # Windows language codes (http://msdn.microsoft.com/en-us/library/0h88fahh(VS.85).aspx) # Notes: This isn't an exact science. The reference above gives language codes # which are different from some ISO 639-1 numbers. Also, some Windows language # codes don't appear to have ISO 639-1 equivalents. # 0x0428 - fa by ref above # 0x048c - no ISO equivalent # 0x081a/0x83c - sr-SP # 0x0c0a - modern? # 0x2409 - Caribbean country code not found in ISO 3166-1 Windows => { 0x0401 => 'ar-SA', 0x0438 => 'fo', 0x0481 => 'mi', 0x1409 => 'en-NZ', 0x0402 => 'bg', 0x0439 => 'hi', 0x0482 => 'oc', 0x140a => 'es-CR', 0x0403 => 'ca', 0x043a => 'mt', 0x0483 => 'co', 0x140c => 'fr-LU', 0x0404 => 'zh-TW', 0x043b => 'se-NO', 0x0484 => 'gsw', 0x141a => 'bs-BA', 0x0405 => 'cs', 0x043c => 'gd', 0x0485 => 'sah', 0x143b => 'smj-SE', 0x0406 => 'da', 0x043d => 'yi', 0x0486 => 'ny', 0x1801 => 'ar-MA', 0x0407 => 'de-DE', 0x043e => 'ms-MY', 0x0487 => 'rw', 0x1809 => 'en-IE', 0x0408 => 'el', 0x043f => 'kk', 0x048c => 'Dari', 0x180a => 'es-PA', 0x0409 => 'en-US', 0x0440 => 'ky', 0x0801 => 'ar-IQ', 0x180c => 'fr-MC', 0x040a => 'es-ES', 0x0441 => 'sw', 0x0804 => 'zh-CN', 0x181a => 'sr-BA', 0x040b => 'fi', 0x0442 => 'tk', 0x0807 => 'de-CH', 0x183b => 'sma-NO', 0x040c => 'fr-FR', 0x0443 => 'uz-UZ', 0x0809 => 'en-GB', 0x1c01 => 'ar-TN', 0x040d => 'he', 0x0444 => 'tt', 0x080a => 'es-MX', 0x1c09 => 'en-ZA', 0x040e => 'hu', 0x0445 => 'bn-IN', 0x080c => 'fr-BE', 0x1c0a => 'es-DO', 0x040f => 'is', 0x0446 => 'pa', 0x0810 => 'it-CH', 0x1c1a => 'sr-BA', 0x0410 => 'it-IT', 0x0447 => 'gu', 0x0813 => 'nl-BE', 0x1c3b => 'sma-SE', 0x0411 => 'ja', 0x0448 => 'wo', 0x0814 => 'nn', 0x2001 => 'ar-OM', 0x0412 => 'ko', 0x0449 => 'ta', 0x0816 => 'pt-PT', 0x2009 => 'en-JM', 0x0413 => 'nl-NL', 0x044a => 'te', 0x0818 => 'ro-MO', 0x200a => 'es-VE', 0x0414 => 'no-NO', 0x044b => 'kn', 0x0819 => 'ru-MO', 0x201a => 'bs-BA', 0x0415 => 'pl', 0x044c => 'ml', 0x081a => 'sr-RS', 0x203b => 'sms', 0x0416 => 'pt-BR', 0x044d => 'as', 0x081d => 'sv-FI', 0x2401 => 'ar-YE', 0x0417 => 'rm', 0x044e => 'mr', 0x082c => 'az-AZ', 0x2409 => 'en-CB', 0x0418 => 'ro', 0x044f => 'sa', 0x082e => 'dsb', 0x240a => 'es-CO', 0x0419 => 'ru', 0x0450 => 'mn-MN', 0x083b => 'se-SE', 0x243b => 'smn', 0x041a => 'hr', 0x0451 => 'bo', 0x083c => 'ga', 0x2801 => 'ar-SY', 0x041b => 'sk', 0x0452 => 'cy', 0x083e => 'ms-BN', 0x2809 => 'en-BZ', 0x041c => 'sq', 0x0453 => 'km', 0x0843 => 'uz-UZ', 0x280a => 'es-PE', 0x041d => 'sv-SE', 0x0454 => 'lo', 0x0845 => 'bn-BD', 0x2c01 => 'ar-JO', 0x041e => 'th', 0x0456 => 'gl', 0x0850 => 'mn-CN', 0x2c09 => 'en-TT', 0x041f => 'tr', 0x0457 => 'kok', 0x085d => 'iu-CA', 0x2c0a => 'es-AR', 0x0420 => 'ur', 0x045a => 'syr', 0x085f => 'tmh', 0x3001 => 'ar-LB', 0x0421 => 'id', 0x045b => 'si', 0x086b => 'qu-EC', 0x3009 => 'en-ZW', 0x0422 => 'uk', 0x045d => 'iu-CA', 0x0c01 => 'ar-EG', 0x300a => 'es-EC', 0x0423 => 'be', 0x045e => 'am', 0x0c04 => 'zh-HK', 0x3401 => 'ar-KW', 0x0424 => 'sl', 0x0461 => 'ne', 0x0c07 => 'de-AT', 0x3409 => 'en-PH', 0x0425 => 'et', 0x0462 => 'fy', 0x0c09 => 'en-AU', 0x340a => 'es-CL', 0x0426 => 'lv', 0x0463 => 'ps', 0x0c0a => 'es-ES', 0x3801 => 'ar-AE', 0x0427 => 'lt', 0x0464 => 'fil', 0x0c0c => 'fr-CA', 0x380a => 'es-UY', 0x0428 => 'tg', 0x0465 => 'dv', 0x0c1a => 'sr-RS', 0x3c01 => 'ar-BH', 0x042a => 'vi', 0x0468 => 'ha', 0x0c3b => 'se-FI', 0x3c0a => 'es-PY', 0x042b => 'hy', 0x046a => 'yo', 0x0c6b => 'qu-PE', 0x4001 => 'ar-QA', 0x042c => 'az-AZ', 0x046b => 'qu-BO', 0x1001 => 'ar-LY', 0x4009 => 'en-IN', 0x042d => 'eu', 0x046c => 'st', 0x1004 => 'zh-SG', 0x400a => 'es-BO', 0x042e => 'hsb', 0x046d => 'ba', 0x1007 => 'de-LU', 0x4409 => 'en-MY', 0x042f => 'mk', 0x046e => 'lb', 0x1009 => 'en-CA', 0x440a => 'es-SV', 0x0430 => 'st', 0x046f => 'kl', 0x100a => 'es-GT', 0x4809 => 'en-SG', 0x0431 => 'ts', 0x0470 => 'ig', 0x100c => 'fr-CH', 0x480a => 'es-HN', 0x0432 => 'tn', 0x0478 => 'yi', 0x101a => 'hr-BA', 0x4c0a => 'es-NI', 0x0434 => 'xh', 0x047a => 'arn', 0x103b => 'smj-NO',0x500a => 'es-PR', 0x0435 => 'zu', 0x047c => 'moh', 0x1401 => 'ar-DZ', 0x540a => 'es-US', 0x0436 => 'af', 0x047e => 'br', 0x1404 => 'zh-MO', 0x0437 => 'ka', 0x0480 => 'ug', 0x1407 => 'de-LI', }, Unicode => { }, ISO => { }, Custom => { }, ); # the 63 known WOFF2 tags my @knownTags = ( 'cmap', 'head', 'hhea', 'hmtx', 'maxp', 'name', 'OS/2', 'post', 'cvt', 'fpgm', 'glyf', 'loca', 'prep', 'CFF', 'VORG', 'EBDT', 'EBLC', 'gasp', 'hdmx', 'kern', 'LTSH', 'PCLT', 'VDMX', 'vhea', 'vmtx', 'BASE', 'GDEF', 'GPOS', 'GSUB', 'EBSC', 'JSTF', 'MATH', 'CBDT', 'CBLC', 'COLR', 'CPAL', 'SVG', 'sbix', 'acnt', 'avar', 'bdat', 'bloc', 'bsln', 'cvar', 'fdsc', 'feat', 'fmtx', 'fvar', 'gvar', 'hsty', 'just', 'lcar', 'mort', 'morx', 'opbd', 'prop', 'trak', 'Zapf', 'Silf', 'Glat', 'Gloc', 'Feat', 'Sill', ); # eclectic table of tags for various format font files %Image::ExifTool::Font::Main = ( GROUPS => { 2 => 'Document' }, NOTES => q{ This table contains a collection of tags found in font files of various formats. ExifTool current recognizes OTF, TTF, TTC, DFONT, PFA, PFB, PFM, AFM, ACFM, AMFM, WOFF and WOFF2 font files. }, name => { SubDirectory => { TagTable => 'Image::ExifTool::Font::Name' }, }, C2PA => { SubDirectory => { TagTable => 'Image::ExifTool::Jpeg2000::Main', Start => 20 }, }, PFM => { Name => 'PFMHeader', SubDirectory => { TagTable => 'Image::ExifTool::Font::PFM' }, }, PSInfo => { Name => 'PSFontInfo', SubDirectory => { TagTable => 'Image::ExifTool::Font::PSInfo' }, }, AFM => { Name => 'AFM', SubDirectory => { TagTable => 'Image::ExifTool::Font::AFM' }, }, numfonts => 'NumFonts', fontname => 'FontName', postfont => { Name => 'PostScriptFontName', Description => 'PostScript Font Name', }, # for WOFF files WOFFVersion => { }, XML => { SubDirectory => { TagTable => 'Image::ExifTool::Font::XML', IgnoreProp => { metadata => 1 }, }, }, ); # TrueType name tags (ref 1/2) %Image::ExifTool::Font::Name = ( GROUPS => { 2 => 'Document' }, NOTES => q{ The following tags are extracted from the TrueType font "name" table found in OTF, TTF, TTC, DFONT, WOFF and WOFF2 files. These tags support localized languages by adding a hyphen followed by a language code to the end of the tag name (eg. "Copyright-fr" or "License-en-US"). Tags with no language code use the default language of "en". }, 0 => { Name => 'Copyright', Groups => { 2 => 'Author' } }, 1 => 'FontFamily', 2 => 'FontSubfamily', 3 => 'FontSubfamilyID', 4 => 'FontName', # full name 5 => 'NameTableVersion', 6 => { Name => 'PostScriptFontName', Description => 'PostScript Font Name' }, 7 => 'Trademark', 8 => 'Manufacturer', 9 => 'Designer', 10 => 'Description', 11 => 'VendorURL', 12 => 'DesignerURL', 13 => 'License', 14 => 'LicenseInfoURL', 16 => 'PreferredFamily', 17 => 'PreferredSubfamily', 18 => 'CompatibleFontName', 19 => 'SampleText', 20 => { Name => 'PostScriptFontName', Description => 'PostScript Font Name', }, 21 => 'WWSFamilyName', 22 => 'WWSSubfamilyName', ); # PostScript Font Metric file header (ref 4) %Image::ExifTool::Font::PFM = ( GROUPS => { 2 => 'Document' }, PROCESS_PROC => \&Image::ExifTool::ProcessBinaryData, NOTES => 'Tags extracted from the PFM file header.', 0 => { Name => 'PFMVersion', Format => 'int16u', PrintConv => 'sprintf("%x.%.2x",$val>>8,$val&0xff)', }, 6 => { Name => 'Copyright', Format => 'string[60]', Groups => { 2 => 'Author' } }, 66 => { Name => 'FontType', Format => 'int16u' }, 68 => { Name => 'PointSize', Format => 'int16u' }, 70 => { Name => 'YResolution', Format => 'int16u' }, 72 => { Name => 'XResolution', Format => 'int16u' }, 74 => { Name => 'Ascent', Format => 'int16u' }, 76 => { Name => 'InternalLeading', Format => 'int16u' }, 78 => { Name => 'ExternalLeading', Format => 'int16u' }, 80 => { Name => 'Italic' }, 81 => { Name => 'Underline' }, 82 => { Name => 'Strikeout' }, 83 => { Name => 'Weight', Format => 'int16u' }, 85 => { Name => 'CharacterSet' }, 86 => { Name => 'PixWidth', Format => 'int16u' }, 88 => { Name => 'PixHeight', Format => 'int16u' }, 90 => { Name => 'PitchAndFamily' }, 91 => { Name => 'AvgWidth', Format => 'int16u' }, 93 => { Name => 'MaxWidth', Format => 'int16u' }, 95 => { Name => 'FirstChar' }, 96 => { Name => 'LastChar' }, 97 => { Name => 'DefaultChar' }, 98 => { Name => 'BreakChar' }, 99 => { Name => 'WidthBytes', Format => 'int16u' }, # 101 => { Name => 'DeviceTypeOffset', Format => 'int32u' }, # 105 => { Name => 'FontNameOffset', Format => 'int32u' }, # 109 => { Name => 'BitsPointer', Format => 'int32u' }, # 113 => { Name => 'BitsOffset', Format => 'int32u' }, ); # PostScript FontInfo attributes (PFA, PFB) (ref PH) %Image::ExifTool::Font::PSInfo = ( GROUPS => { 2 => 'Document' }, NOTES => 'Tags extracted from PostScript font files (PFA and PFB).', FullName => { }, FamilyName => { Name => 'FontFamily' }, Weight => { }, ItalicAngle => { }, isFixedPitch=> { }, UnderlinePosition => { }, UnderlineThickness => { }, Copyright => { Groups => { 2 => 'Author' } }, Notice => { Groups => { 2 => 'Author' } }, version => { }, FontName => { }, FontType => { }, FSType => { }, ); # Adobe Font Metrics tags (AFM) (ref 6) %Image::ExifTool::Font::AFM = ( GROUPS => { 2 => 'Document' }, NOTES => 'Tags extracted from Adobe Font Metrics files (AFM, ACFM and AMFM).', 'Creation Date' => { Name => 'CreateDate', Groups => { 2 => 'Time' } }, FontName => { }, FullName => { }, FamilyName => { Name => 'FontFamily' }, Weight => { }, Version => { }, Notice => { Groups => { 2 => 'Author' } }, EncodingScheme => { }, MappingScheme => { }, EscChar => { }, CharacterSet=> { }, Characters => { }, IsBaseFont => { }, # VVector => { }, IsFixedV => { }, CapHeight => { }, XHeight => { }, Ascender => { }, Descender => { }, ); # WOFF XML %Image::ExifTool::Font::XML = ( GROUPS => { 1 => 'XML', 2 => 'Document' }, PROCESS_PROC => \&Image::ExifTool::XMP::ProcessXMP, NOTES => 'Tags found in WOFF and WOFF2 XML metadata.', version => { }, uniqueidId => { Name => 'UniqueID' }, vendorName => { }, vendorUrl => { Name => 'VendorURL' }, vendorDir => { }, vendorClass => { }, creditsCreditName => { Name => 'CreditName' }, creditsCreditUrl => { Name => 'CreditURL' }, creditsCreditRole => { Name => 'CreditRole' }, creditsCreditDir => { Name => 'CreditDir' }, creditsCreditClass => { Name => 'CreditClass' }, descriptionUrl => { }, descriptionText => { Name => 'Description' }, licenseUrl => { Name => 'LicenseURL' }, licenseId => { Name => 'LicenseID' }, licenseText => { Name => 'License' }, copyrightText => { Name => 'Copyright', Groups => { 2 => 'Author' } }, trademarkText => { Name => 'Trademark' }, licenseeDir => { }, licenseeName => { }, licenseeClass => { }, extensionId => { Name => 'ExtensionID' }, extensionName => { }, extensionItemId => { Name => 'ExtensionItemID' }, extensionItemName => { }, extensionItemValue => { }, ); #------------------------------------------------------------------------------ # Read information from a TrueType font collection (TTC) (refs 2,3) # Inputs: 0) ExifTool ref, 1) dirInfo ref # Returns: 1 on success, 0 if this wasn't a valid TrueType font collection sub ProcessTTC($$) { my ($et, $dirInfo) = @_; my $raf = $$dirInfo{RAF}; my ($buff, $i); return 0 unless $raf->Read($buff, 12) == 12; return 0 unless $buff =~ /^ttcf\0[\x01\x02]\0\0/; SetByteOrder('MM'); my $num = Get32u(\$buff, 8); # might as well put a limit on the number of fonts we will parse (< 256) return 0 unless $num < 0x100 and $raf->Read($buff, $num * 4) == $num * 4; $et->SetFileType('TTC'); return 1 if $$et{OPTIONS}{FastScan} and $$et{OPTIONS}{FastScan} == 3; my $tagTablePtr = GetTagTable('Image::ExifTool::Font::Main'); $et->HandleTag($tagTablePtr, 'numfonts', $num); # loop through all fonts in the collection for ($i=0; $i<$num; ++$i) { my $n = $i + 1; $et->VPrint(0, "Font $n:\n"); $$et{SET_GROUP1} = "+$n"; my $offset = Get32u(\$buff, $i * 4); $raf->Seek($offset, 0) or last; ProcessOTF($et, $dirInfo) or last; } delete $$et{SET_GROUP1}; return 1; } #------------------------------------------------------------------------------ # Process an OTF tag table entry (refs 1,2) # Inputs: 0) ExifTool ref, 1) entry index, 2) tag, 3) data ref, # 4) offset if uncompressed, 5) true to skip because data is transformed # Returns: undef on success, 0 to stop processing table, or error string otherwise sub ProcessTableEntry($$$$;$$) { my ($et, $idx, $tag, $dataPt, $offset, $transformed) = @_; my $verbose = $et->Options('Verbose'); my $size = length $$dataPt; $offset or $offset = 0; if ($verbose) { $tag =~ s/([\0-\x1f\x7f-\xff])/sprintf('\x%.2x',ord $1)/ge; my $str = sprintf("%s%d) Tag '%s' (offset 0x%.4x, %d bytes)\n", $$et{INDENT}, $idx, $tag, $offset, $size); $et->VPrint(0, $str); $et->VerboseDump($dataPt, Addr => $offset) if $verbose > 2; return undef unless $processTag{$tag}; } return undef if $transformed or $size < 8; unless ($tag eq 'name') { my $tagTablePtr = GetTagTable('Image::ExifTool::Font::Main'); $et->HandleTag($tagTablePtr, $tag, undef, DataPt => $dataPt, Size => length($$dataPt)); return undef; } # process the 'name' tag my $entries = Get16u($dataPt, 2); my $recEnd = 6 + $entries * 12; if ($recEnd > $size) { $et->Warn('Truncated name record'); return 0; } my $strStart = Get16u($dataPt, 4); if ($strStart < $recEnd or $strStart > $size) { $et->Warn('Invalid string offset'); return 0; } # parse language-tag record (in format 1 Naming table only) (ref 2) my ($i, %langTag); if (Get16u($dataPt, 0) == 1 and $recEnd + 2 <= $size) { my $langTags = Get16u($dataPt, $recEnd); if ($langTags and $recEnd + 2 + $langTags * 4 < $size) { for ($i=0; $i<$langTags; ++$i) { my $pt = $recEnd + 2 + $i * 4; my $langLen = Get16u($dataPt, $pt); # make sure the language string length is reasonable (UTF-16BE) last if $langLen == 0 or $langLen & 0x01 or $langLen > 40; my $langPt = Get16u($dataPt, $pt + 2) + $strStart; last if $langPt + $langLen > $size; my $lang = substr($$dataPt, $langPt, $langLen); $lang = $et->Decode($lang,'UCS2','MM','UTF8'); $lang =~ tr/-_a-zA-Z0-9//dc; # remove naughty characters $langTag{$i + 0x8000} = $lang; } } } my $tagTablePtr = GetTagTable('Image::ExifTool::Font::Name'); my $oldIndent = $$et{INDENT}; $$et{INDENT} .= '| '; $et->VerboseDir('Name', $entries) if $verbose; for ($i=0; $i<$entries; ++$i) { my $pt = 6 + $i * 12; my $platform = Get16u($dataPt, $pt); my $encoding = Get16u($dataPt, $pt + 2); my $langID = Get16u($dataPt, $pt + 4); my $nameID = Get16u($dataPt, $pt + 6); my $strLen = Get16u($dataPt, $pt + 8); my $strPt = Get16u($dataPt, $pt + 10) + $strStart; if ($strPt + $strLen <= $size) { my $val = substr($$dataPt, $strPt, $strLen); my ($lang, $charset, $extra); my $sys = $ttPlatform{$platform}; # translate from specified encoding if ($sys) { $lang = $ttLang{$sys}{$langID} || $langTag{$langID}; $charset = $ttCharset{$sys}{$encoding}; if (not $charset) { if (not defined $charset and not $$et{FontWarn}) { $et->Warn("Unknown $sys character set ($encoding)"); $$et{FontWarn} = 1; } } else { # translate to ExifTool character set $val = $et->Decode($val, $charset); } } else { $et->Warn("Unknown platform ($platform) for name $nameID"); } # get the tagInfo for our specific language (use 'en' for default) my $tagInfo = $et->GetTagInfo($tagTablePtr, $nameID); if ($tagInfo and $lang and $lang ne 'en') { my $langInfo = Image::ExifTool::GetLangInfo($tagInfo, $lang); $tagInfo = $langInfo if $langInfo; } if ($verbose) { $langID > 0x400 and $langID = sprintf('0x%x', $langID); $extra = ", Plat=$platform/" . ($sys || 'Unknown') . ', ' . "Enc=$encoding/" . ($charset || 'Unknown') . ', ' . "Lang=$langID/" . ($lang || 'Unknown'); } $et->HandleTag($tagTablePtr, $nameID, $val, TagInfo => $tagInfo, DataPt => $dataPt, DataPos => $offset, Start => $strPt, Size => $strLen, Index => $i, Extra => $extra, ); } } $$et{INDENT} = $oldIndent; return $verbose ? undef : 0; } #------------------------------------------------------------------------------ # Read information from a TrueType font file (OTF or TTF) (refs 1,2) # Inputs: 0) ExifTool ref, 1) dirInfo ref # Returns: 1 on success, 0 if this wasn't a valid TrueType font file sub ProcessOTF($$) { my ($et, $dirInfo) = @_; my $raf = $$dirInfo{RAF}; my ($tbl, $buff, $pos, $i); my $base = $$dirInfo{Base} || 0; return 0 unless $raf->Read($buff, 12) == 12; return 0 unless $buff =~ /^(\0\x01\0\0|OTTO|true|typ1|\xa5(kbd|lst))[\0\x01]/; $et->SetFileType($1 eq 'OTTO' ? 'OTF' : 'TTF'); return 1 if $$et{OPTIONS}{FastScan} and $$et{OPTIONS}{FastScan} == 3; SetByteOrder('MM'); my $numTables = Get16u(\$buff, 4); return 0 unless $numTables > 0 and $numTables < 0x200; my $len = $numTables * 16; return 0 unless $raf->Read($tbl, $len) == $len; my $verbose = $et->Options('Verbose'); for ($pos=0, $i=0; $pos<$len; $pos+=16, ++$i) { # look for tags to process my $tag = substr($tbl, $pos, 4); next unless $processTag{$tag} or $verbose; my $offset = Get32u(\$tbl, $pos + 8); my $size = Get32u(\$tbl, $pos + 12); unless ($raf->Seek($offset+$base, 0) and $raf->Read($buff, $size) == $size) { $et->Warn("Error reading '${tag}' data"); next; } my $err = ProcessTableEntry($et, $i, $tag, \$buff, $offset); $err and $et->Warn($err), last; last if defined $err; } return 1; } #------------------------------------------------------------------------------ # Read information from an Adobe Font Metrics file (AFM, ACFM, AMFM) (ref 6) # Inputs: 0) ExifTool ref, 1) dirInfo ref # Returns: 1 on success, 0 if this wasn't a recognized AFM-type file sub ProcessAFM($$) { my ($et, $dirInfo) = @_; my $raf = $$dirInfo{RAF}; my ($buff, $comment); require Image::ExifTool::PostScript; local $/ = Image::ExifTool::PostScript::GetInputRecordSeparator($raf); $raf->ReadLine($buff); return 0 unless $buff =~ /^Start(Comp|Master)?FontMetrics\s+\d+/; my $ftyp = $1 ? ($1 eq 'Comp' ? 'ACFM' : 'AMFM') : 'AFM'; $et->SetFileType($ftyp, 'application/x-font-afm'); return 1 if $$et{OPTIONS}{FastScan} and $$et{OPTIONS}{FastScan} == 3; my $tagTablePtr = GetTagTable('Image::ExifTool::Font::AFM'); for (;;) { $raf->ReadLine($buff) or last; if (defined $comment and $buff !~ /^Comment\s/) { $et->FoundTag('Comment', $comment); undef $comment; } $buff =~ /^(\w+)\s+(.*?)[\x0d\x0a]/ or next; my ($tag, $val) = ($1, $2); if ($tag eq 'Comment' and $val =~ /^(Creation Date):\s+(.*)/) { ($tag, $val) = ($1, $2); } $val =~ s/^\((.*)\)$/$1/; # (some values may be in brackets) if ($tag eq 'Comment') { # concatinate all comments into a single value $comment = defined($comment) ? "$comment\n$val" : $val; next; } unless ($et->HandleTag($tagTablePtr, $tag, $val)) { # end parsing if we start any subsection last if $tag =~ /^Start/ and $tag ne 'StartDirection'; } } return 1; } #------------------------------------------------------------------------------ # Read WOFF2 255UInt16 integer (ref 8) # Inputs: 0) raf ref # Returns: value, or undef on error sub Read255UInt16($) { my $raf = shift; my $buff; return undef unless $raf->Read($buff, 1); my $val = unpack('C', $buff); if ($val == 253) { return undef unless $raf->Read($buff, 2) == 2; $val = unpack('n', $buff); } elsif ($val == 254) { return undef unless $raf->Read($buff, 1); $val = unpack('C', $buff) + 253 * 2; } elsif ($val == 255) { return undef unless $raf->Read($buff, 1); $val = unpack('C', $buff) + 253; } return $val; } #------------------------------------------------------------------------------ # Read WOFF2 UIntBase128 integer (ref 8) # Inputs: 0) raf ref # Returns: value, or undef on error sub ReadUIntBase128($) { my $raf = shift; my $buff; my $val = 0; foreach (0..4) { return undef unless $raf->Read($buff, 1); my $byte = unpack('C', $buff); return undef if not $_ and $byte == 0x80; return undef if $val & 0xfe000000; $val = ($val << 7) | ($byte & 0x7f); return $val unless $byte & 0x80; } return undef; # can't be longer than 5 bytes } #------------------------------------------------------------------------------ # Uncompress data # Inputs: 0) ExifTool ref, 1) data ref # Returns: true on success sub Uncompress($$) { my ($et, $dataPt) = @_; my $stat; unless (eval { require Compress::Zlib }) { $et->Warn('Install Compress::Zlib to read compressed metadata'); return 0; } my $inflate = Compress::Zlib::inflateInit(); $inflate and ($$dataPt, $stat) = $inflate->inflate($$dataPt); unless ($inflate and $stat == Compress::Zlib::Z_STREAM_END()) { $et->Warn('Error uncompressing metadata'); return 0; } return 1; } #------------------------------------------------------------------------------ # Brotli uncompress data # Inputs: 0) ExifTool ref, 1) data ref # Returns: true on success sub Unbrotli($$) { my ($et, $dataPt) = @_; unless (eval { require IO::Uncompress::Brotli }) { $et->Warn('Install IO::Compress::Brotli to decode Brotli-compressed metadata'); return 0; } eval { $$dataPt = IO::Uncompress::Brotli::unbro($$dataPt, 100000000) }; if ($@) { $et->Warn('Error decoding metadata'); $et->Warn('Try updating to IO::Uncompress::Brotli 0.004 or later'); return 0; } return 1; } #------------------------------------------------------------------------------ # Read information from WOFF/WOFF2 font files # Inputs: 0) ExifTool ref, 1) dirInfo ref # Returns: 1 on success, 0 if this wasn't a recognized WOFF file sub ProcessWOFF($$) { my ($et, $dirInfo) = @_; my $raf = $$dirInfo{RAF}; my ($buff, $tbl, $i); $raf->Seek(0,0) and $raf->Read($buff,48) == 48 or return 0; $buff =~ /^(wOF[F2])/ or return 0; my ($type, $off) = $1 eq 'wOFF' ? ('woff' , 20) : ('woff2', 24); $et->SetFileType(uc($type), "font/$type"); SetByteOrder('MM'); my $flavor = substr($buff, 4, 4); my $numTables = Get16u(\$buff, 12); my ($vh, $vl, $metaPos, $metaLen) = unpack("x${off}nnNN", $buff); my $verbose = $et->Options('Verbose'); my $tagTablePtr = GetTagTable('Image::ExifTool::Font::Main'); $et->HandleTag($tagTablePtr, WOFFVersion => "$vh.$vl"); # # read font table # if ($type eq 'woff') { unless ($raf->Seek($off+24,0) and $raf->Read($tbl,$numTables*20)==$numTables*20) { $et->Warn('Error reading font table'); return 1; } for ($i=0; $i<$numTables; ++$i) { my $pt = $i * 20; my ($tag, $pos, $compLen, $len) = unpack("x${pt}a4N3", $tbl); next unless $processTag{$tag} or $verbose; $raf->Seek($pos,0) and $raf->Read($buff,$compLen)==$compLen or $et->Warn('Bad font table entry'), return 1; my $dataPos; if ($compLen eq $len) { $dataPos = $pos; } else { next unless Uncompress($et, \$buff); } my $err = ProcessTableEntry($et, $i, $tag, \$buff, $dataPos); $err and $et->Warn($err), return 1; last if defined $err; } } else { # WOFF2 my $compSize = Get32u(\$buff, 20); my ($err, @entry, $entry); $raf->Seek($off+24,0) or $et->Warn('Error seeking to font table'), return 1; for ($i=0; $i<$numTables; ++$i) { $raf->Read($buff, 1) or $err = 1, last; my $flags = unpack('C', $buff); my $tag = $knownTags[$flags & 0x3f]; $tag or $raf->Read($tag, 4) == 4 or $err = 1, last; my $len = ReadUIntBase128($raf); defined $len or $err = 1, last; my $transformed; if (($tag eq 'glyf' or $tag eq 'loca') xor $flags & 0xc0) { # a non-null transform was used $len = ReadUIntBase128($raf); $transformed = 1; } # save information about this entry for later push @entry, [ $i, $tag, $len, $transformed ]; } $err and $et->Warn('Error reading font table'), return 1; # skip the collection table if necessary if ($flavor eq 'ttcf') { $raf->Seek(4, 1) or $et->Warn('Seek error'), return 1; my $n = Read255UInt16($raf); defined $n and $raf->Seek(4,1) or $et->Warn('Error reading collection table'), return 1; $raf->Seek(4, 1) or $err = 1, last; for ($i=0; $i<$n; ++$i) { defined Read255UInt16($raf) or $err = 1, last; } $err and $et->Warn('Error reading collection directory'), return 1; } $raf->Read($buff,$compSize) == $compSize or $et->Warn('Error reading font data'), return 1; return 1 unless Unbrotli($et, \$buff); # after all that exhausting and frankly unnecessary work (poor file design), # we finally have the uncompressed font data so we can process the table entries my $pos = 0; foreach $entry (@entry) { my ($i, $tag, $len, $transformed) = @$entry; if ($processTag{$tag} or $verbose) { my $dat = substr($buff, $pos, $len); my $err = ProcessTableEntry($et, $i, $tag, \$dat, undef, $transformed); $err and $et->Warn($err), return 1; last if defined $err; } $pos += $len; } } # # read compressed XML-format metadata (NC) # if ($metaLen) { unless ($raf->Seek($metaPos,0) and $raf->Read($buff,$metaLen)==$metaLen) { $et->Warn('Error reading metadata'); return 1; } if ($type eq 'woff') { return 1 unless Uncompress($et, \$buff); } else { # WOFF2 return 1 unless Unbrotli($et, \$buff); } # (we don't properly support XML structures) my $oldStruct = $et->Options('Struct'); $et->Options(Struct => 0); $et->HandleTag($tagTablePtr, 'XML', $buff); $et->Options(Struct => $oldStruct); } return 1; } #------------------------------------------------------------------------------ # Read information from various format font files # Inputs: 0) ExifTool ref, 1) dirInfo ref # Returns: 1 on success, 0 if this wasn't a recognized Font file sub ProcessFont($$) { my ($et, $dirInfo) = @_; my $raf = $$dirInfo{RAF}; my ($buff, $buf2, $rtnVal); return 0 unless $raf->Read($buff, 24) and $raf->Seek(0,0); if ($buff =~ /^(\0\x01\0\0|OTTO|true|typ1)[\0\x01]/) { # OTF, TTF $rtnVal = ProcessOTF($et, $dirInfo); } elsif ($buff =~ /^ttcf\0[\x01\x02]\0\0/) { # TTC $rtnVal = ProcessTTC($et, $dirInfo); } elsif ($buff =~ /^Start(Comp|Master)?FontMetrics\s+\d+/s) { # AFM $rtnVal = ProcessAFM($et, $dirInfo); } elsif ($buff =~ /^(.{6})?%!(PS-(AdobeFont-|Bitstream )|FontType1-)/s) {# PFA, PFB $raf->Seek(6,0) and $et->SetFileType('PFB') if $1; require Image::ExifTool::PostScript; $rtnVal = Image::ExifTool::PostScript::ProcessPS($et, $dirInfo); } elsif ($buff =~ /^\0[\x01\x02]/ and $raf->Seek(0, 2) and # PFM # validate file size $raf->Tell() > 117 and $raf->Tell() == unpack('x2V',$buff) and # read PFM header $raf->Seek(0,0) and $raf->Read($buff,117) == 117 and # validate "DeviceType" string (must be "PostScript\0") SetByteOrder('II') and $raf->Seek(Get32u(\$buff, 101), 0) and # the DeviceType should be "PostScript\0", but FontForge # incorrectly writes "Postscript\0", so ignore case $raf->Read($buf2, 11) == 11 and lc($buf2) eq "postscript\0") { $et->SetFileType('PFM'); return 1 if $$et{OPTIONS}{FastScan} and $$et{OPTIONS}{FastScan} == 3; SetByteOrder('II'); my $tagTablePtr = GetTagTable('Image::ExifTool::Font::Main'); # process the PFM header $et->HandleTag($tagTablePtr, 'PFM', $buff); # extract the font names my $nameOff = Get32u(\$buff, 105); if ($raf->Seek($nameOff, 0) and $raf->Read($buff, 256) and $buff =~ /^([\x20-\xff]+)\0([\x20-\xff]+)\0/) { $et->HandleTag($tagTablePtr, 'fontname', $1); $et->HandleTag($tagTablePtr, 'postfont', $2); } $rtnVal = 1; } elsif ($buff =~ /^(wOF[F2])/) { $rtnVal = ProcessWOFF($et, $dirInfo); } else { $rtnVal = 0; } return $rtnVal; } 1; # end __END__ =head1 NAME Image::ExifTool::Font - Read meta information from font files =head1 SYNOPSIS This module is used by Image::ExifTool =head1 DESCRIPTION This module contains the routines required by Image::ExifTool to read meta information from various format font files. Currently recognized font file types are OTF, TTF, TTC, DFONT, PFA, PFB, PFM, AFM, ACFM, AMFM, WOFF and WOFF2. =head1 AUTHOR Copyright 2003-2026, Phil Harvey (philharvey66 at gmail.com) This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =head1 REFERENCES =over 4 =item L =item L =item L =item L =item L =item L =item L =item L =back =head1 SEE ALSO L, L =cut