PageRenderTime 61ms CodeModel.GetById 21ms RepoModel.GetById 0ms app.codeStats 1ms

/htdocs/gposgsub.html

https://bitbucket.org/sortsmill/sortsmill-tools
HTML | 1789 lines | 1319 code | 467 blank | 3 comment | 0 complexity | d03705ac1e038ce583a27e753ea02586 MD5 | raw file
Possible License(s): GPL-3.0, LGPL-2.0

Large files files are truncated, but you can click here to view the full file

    

  1. <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
    2. 

  2.     "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
    3. 

  3. 

  4. <html xmlns="http://www.w3.org/1999/xhtml">
    5. 

  5. <head>
    6. 

  6.   <meta name="generator" content=
    7. 

  7.   "HTML Tidy for Linux (vers 25 March 2009), see www.w3.org" />
    8. 

  8.   <!-- Created with AOLpress/2.0 -->
    9. 

  9.   <!-- AP: Created on: 29-Dec-2002 -->
    10. 

  10.   <!-- AP: Last modified: 20-Mar-2008 -->
    11. 

  11. 

  12.   <title>Advanced Typography tables</title>
    13. 

  13.   <link rel="icon" href="fftype16.png" />
    14. 

  14.   <link rel="stylesheet" type="text/css" href="FontForge.css" />
    15. 

  15.   <style type="text/css">
    16. 

  16. /*<![CDATA[*/
    17. 

  17.   h4.c4 {color: red}
    18. 

  18.   p.c3 {text-align: center}
    19. 

  19.   p.c2 {text-align: left}
    20. 

  20.   h1.c1 {text-align: center}
    21. 

  21.   /*]]>*/
    22. 

  22.   </style>
    23. 

  23. </head>
    24. 

  24. 

  25. <body>
    26. 

  26.   <div id="in">
    27. 

  27.     <h1 class="c1">Advanced Typography Tables</h1>
    28. 

  28. 

  29.     <p>These differ between OpenType (originally called TrueType
    30. 

  30.     Open) and Apple (GX or Apple Advanced Typography). My support
    31. 

  31.     for both OpenType and Apple is incomplete.</p>
    32. 

  32. 

  33.     <ul>
    34. 

  34.       <li><a href="#opentype">The <b><code>GPOS</code></b>,
    35. 

  35.       <b><code>GSUB</code></b> and <b><code>GDEF</code></b>
    36. 

  36.       opentype tables</a></li>
    37. 

  37. 

  38.       <li><a href="#AAT">Apple Advanced Typography</a></li>
    39. 

  39. 

  40.       <li><a href="#Conversion">What features can be converted
    41. 

  41.       between OpenType and AAT?</a></li>
    42. 

  42. 

  43.       <li><a href="TrueOpenTables.html">other true type and open
    44. 

  44.       type tables</a></li>
    45. 

  45. 

  46.       <li><a href="#Unsupported">What is unsupported</a></li>
    47. 

  47. 

  48.       <li><a href="non-standard.html">FontForge's non-standard
    49. 

  49.       extensions</a></li>
    50. 

  50.     </ul>
    51. 

  51. 

  52.     <h2>The <b><code>GPOS,</code></b> <b><code>GSUB,</code></b>
    53. 

  53.     <b><code>GDEF</code></b> and <code>BASE</code> <a name=
    54. 

  54.     "opentype" id="opentype">opentype</a> tables</h2>
    55. 

  55. 

  56.     <p>The first two tables are used for positioning and
    57. 

  57.     substituting glyphs. The GPOS table can control things like:
    58. 

  58.     kerning, accent positioning, cursive joining, etc. The GSUB
    59. 

  59.     table can control things like: ligatures, arabic forms,
    60. 

  60.     vertical rotation, conversion to small caps, indic glyph
    61. 

  61.     rearrangement, etc. GDEF contains some rather esoteric glyph
    62. 

  62.     information, ligature carets, etc. BASE contains information on
    63. 

  63.     baseline placement and line heights.</p>
    64. 

  64. 

  65.     <p>This page assumes basic familiarity with the abilities of
    66. 

  66.     the tables, for more information on them read, study and
    67. 

  67.     inwardly digest the opentype docs on:</p>
    68. 

  68. 

  69.     <ul>
    70. 

  70.       <li><a href=
    71. 

  71.       "http://partners.adobe.com/public/developer/opentype/index_table_formats.html">
    72. 

  72.       The header for both GPOS and GSUB</a></li>
    73. 

  73. 

  74.       <li><a href=
    75. 

  75.       "http://partners.adobe.com/public/developer/opentype/index_table_formats2.html">
    76. 

  76.       The GPOS table</a>, for positioning glyphs</li>
    77. 

  77. 

  78.       <li><a href=
    79. 

  79.       "http://partners.adobe.com/public/developer/opentype/index_table_formats1.html">
    80. 

  80.       The GSUB table</a>, for substituting glyphs</li>
    81. 

  81. 

  82.       <li><a href=
    83. 

  83.       "http://partners.adobe.com/public/developer/opentype/index_table_formats5.html">
    84. 

  84.       The GDEF table</a>, for classifying glyphs and for providing
    85. 

  85.       a ligature caret table</li>
    86. 

  86. 

  87.       <li><a href=
    88. 

  88.       "http://partners.adobe.com/public/developer/opentype/index_base.html">
    89. 

  89.       The BASE table</a>, for baseline placement</li>
    90. 

  90. 

  91.       <li><a href=
    92. 

  92.       "http://partners.adobe.com/public/developer/opentype/index_tag3.html">
    93. 

  93.       The list of feature tags supported by opentype</a></li>
    94. 

  94. 

  95.       <li><a href=
    96. 

  96.       "http://partners.adobe.com/public/developer/opentype/index_tag1.html">
    97. 

  97.       The list of script tags supported by opentype</a></li>
    98. 

  98. 

  99.       <li><a href=
    100. 

  100.       "http://partners.adobe.com/public/developer/opentype/index_tag2.html">
    101. 

  101.       The list of language tags supported by opentype</a></li>
    102. 

  102. 

  103.       <li><a href=
    104. 

  104.       "http://partners.adobe.com/public/developer/opentype/index_tag4.html">
    105. 

  105.       The list of baseline tags supported by opentype</a></li>
    106. 

  106.     </ul>
    107. 

  107. 

  108.     <p>The basic idea of the GPOS and GSUB tables is that each
    109. 

  109.     script (or language within a script) has a set of "features"
    110. 

  110.     that are available to it. A feature in turn is associated with
    111. 

  111.     a lookup which contains data for the feature. An example of a
    112. 

  112.     script is 'latn' for latin, 'arab' for arabic, 'hani' for
    113. 

  113.     chinese ideographs. Two examples of features are 'kern' which
    114. 

  114.     provides kerning information between pairs of glyphs and 'init'
    115. 

  115.     which will transform one set of glyphs to another when those
    116. 

  116.     glyphs are at the beginning of a word.</p>
    117. 

  117. 

  118.     <p>FontForge <a href="gposgsub.html#Unsupported">does not
    119. 

  119.     support</a> the full range of possibilities inherent in these
    120. 

  120.     tables.</p>
    121. 

  121. 

  122.     <h3>The <b><code><a name="GPOS" id="GPOS">GPOS</a></code></b>
    123. 

  123.     table</h3>
    124. 

  124. 

  125.     <p>FontForge will read the following sub tables of the GPOS
    126. 

  126.     table:</p>
    127. 

  127. 

  128.     <table border="1">
    129. 

  129.       <tr>
    130. 

  130.         <th></th>
    131. 

  131. 

  132.         <th>name</th>
    133. 

  133. 

  134.         <th>Reading support</th>
    135. 

  135. 

  136.         <th>Writing support</th>
    137. 

  137.       </tr>
    138. 

  138. 

  139.       <tr>
    140. 

  140.         <td>1</td>
    141. 

  141. 

  142.         <td>single adjustment</td>
    143. 

  143. 

  144.         <td>This sub-table allows the font designer to change the
    145. 

  145.         metrics of a specific glyph. The feature tag will provide a
    146. 

  146.         context in which the change should occur. For instance the
    147. 

  147.         'tnum' (tabular numbers) feature could change a
    148. 

  148.         proportionally spaced digit by adjusting its advance width
    149. 

  149.         to be some set value and then centering the digit (by
    150. 

  150.         adjusting the left side bearing) within the new width.</td>
    151. 

  151. 

  152.         <td>These can be created with the <a href=
    153. 

  153.         "charinfo.html">Element-&gt;Char Info</a>-&gt;Position
    154. 

  154.         command.</td>
    155. 

  155.       </tr>
    156. 

  156. 

  157.       <tr>
    158. 

  158.         <td>2</td>
    159. 

  159. 

  160.         <td>pair adjustment</td>
    161. 

  161. 

  162.         <td>This sub-table allows the font designer to change the
    163. 

  163.         metrics of a specific pair of glyph. The most common use of
    164. 

  164.         this is for kerning where the advance width of the first
    165. 

  165.         glyph is altered depending on which glyph follows it. But
    166. 

  166.         the table is more general than that and could support mark
    167. 

  167.         (accent, vowel) positioning over a base glyph (though that
    168. 

  168.         is more efficiently done with the mark to base
    169. 

  169.         subtable).</td>
    170. 

  170. 

  171.         <td>'kern' feature s may be created from the <a href=
    172. 

  172.         "metricsview.html">Metrics View</a>. 'vkrn' with <a href=
    173. 

  173.         "metricsmenu.html#VKernFromHKern">Metrics-&gt;VKern From
    174. 

  174.         HKern</a>.</td>
    175. 

  175.       </tr>
    176. 

  176. 

  177.       <tr>
    178. 

  178.         <td>3</td>
    179. 

  179. 

  180.         <td>cursive attachment</td>
    181. 

  181. 

  182.         <td>This sub-table allows the font designer to force
    183. 

  183.         adjacent glyphs to join at specific points. It can be used
    184. 

  184.         to generate the slanted script style needed for Urdu.</td>
    185. 

  185. 

  186.         <td>Only the 'curs' feature is supported for this
    187. 

  187.         sub-table. These may be created with the <a href=
    188. 

  188.         "pointmenu.html#AddAnchor">Points-&gt;Add Anchor</a>
    189. 

  189.         command</td>
    190. 

  190.       </tr>
    191. 

  191. 

  192.       <tr>
    193. 

  193.         <td>4</td>
    194. 

  194. 

  195.         <td>mark to base</td>
    196. 

  196. 

  197.         <td>This sub-table allows the font designer to specify how
    198. 

  198.         mark glyphs (accents, vowel signs, etc.) are positioned
    199. 

  199.         over base glyphs. Every glyph can have an attachment point
    200. 

  200.         and the mark's attachment point will be placed on the
    201. 

  201.         base's attachment point so the two join properly. See my
    202. 

  202.         <a href="overview.html#Anchors">example</a> in the
    203. 

  203.         overview.</td>
    204. 

  204. 

  205.         <td>These may be created with the <a href=
    206. 

  206.         "pointmenu.html#AddAnchor">Points-&gt;Add Anchor</a>
    207. 

  207.         command</td>
    208. 

  208.       </tr>
    209. 

  209. 

  210.       <tr>
    211. 

  211.         <td>5</td>
    212. 

  212. 

  213.         <td>mark to ligature</td>
    214. 

  214. 

  215.         <td>This sub-table is very similar to the previous one
    216. 

  216.         except that the base glyph is a ligature and may have
    217. 

  217.         several different points at which the same type of accent
    218. 

  218.         may be placed.</td>
    219. 

  219. 

  220.         <td>These may be created with the <a href=
    221. 

  221.         "pointmenu.html#AddAnchor">Points-&gt;Add Anchor</a>
    222. 

  222.         command</td>
    223. 

  223.       </tr>
    224. 

  224. 

  225.       <tr>
    226. 

  226.         <td>6</td>
    227. 

  227. 

  228.         <td>mark to mark</td>
    229. 

  229. 

  230.         <td>This sub-table is very similar to the previous two
    231. 

  231.         except that the base glyph is itself a mark. This may be
    232. 

  232.         used when a glyph has two accents each of which would
    233. 

  233.         normally be placed at the same attachment point on a base
    234. 

  234.         glyph. The second accent will be place relative to the
    235. 

  235.         first accent rather than to the base glyph.</td>
    236. 

  236. 

  237.         <td>These may be created with the <a href=
    238. 

  238.         "pointmenu.html#AddAnchor">Points-&gt;Add Anchor</a>
    239. 

  239.         command</td>
    240. 

  240.       </tr>
    241. 

  241. 

  242.       <tr>
    243. 

  243.         <td>7</td>
    244. 

  244. 

  245.         <td>contextual positioning</td>
    246. 

  246. 

  247.         <td>This sub-table allows the font designer to control the
    248. 

  248.         positioning of glyphs when they occur within a specific
    249. 

  249.         string (or class of strings). For instance this table could
    250. 

  250.         say "when you see a digit followed by the string "th" then
    251. 

  251.         raise the "th" into a superscript position"</td>
    252. 

    253. 

  253.         <td>These may be created with the <a href=
    254. 

  254.         "lookups.html#contextual-subs">Element-&gt;Font
    255. 

  255.         Info-&gt;Contextual</a> command</td>
    256. 

  256.       </tr>
    257. 

  257. 

  258.       <tr>
    259. 

  259.         <td>8</td>
    260. 

  260. 

  261.         <td>chaining contextual positioning</td>
    262. 

  262. 

  263.         <td>This is a slightly more complex version of the above,
    264. 

  264.         it doesn't really add new capabilities, but it does provide
    265. 

  265.         a more logical approach to the issue.</td>
    266. 

  266. 

  267.         <td>These may be created with the <a href=
    268. 

  268.         "lookups.html#contextual-subs">Element-&gt;Font
    269. 

  269.         Info-&gt;Contextual</a> command</td>
    270. 

  270.       </tr>
    271. 

  271. 

  272.       <tr>
    273. 

  273.         <td>9</td>
    274. 

  274. 

  275.         <td>extension positioning</td>
    276. 

  276. 

  277.         <td>This is used to allow for a GPOS table which is bigger
    278. 

  278.         than 64k. Its use should be quite invisible to the font
    279. 

  279.         designer</td>
    280. 

  280. 

  281.         <td>FontForge uses this sub-table when needed.</td>
    282. 

  282.       </tr>
    283. 

  283. 

  284.       <tr>
    285. 

  285.         <td>10+</td>
    286. 

  286. 

  287.         <td>reserved for future use</td>
    288. 

  288. 

  289.         <td></td>
    290. 

  290. 

  291.         <td>FontForge does not support these sub-tables yet.<br />
    292. 

  292.         (nor does anyone)</td>
    293. 

  293.       </tr>
    294. 

  294.     </table>
    295. 

  295. 

  296.     <p>FontForge also has built into it knowledge on how to provide
    297. 

  297.     default values for some features that use these tables. See
    298. 

  298.     <a href="elementmenu.html#DefaultATT">Element-&gt;Typo.
    299. 

  299.     Features-&gt;Default ATT</a> command for that.</p>
    300. 

  300. 

  301.     <p>FontForge will retain the order of features in the GPOS
    302. 

  302.     table and when a font is generated the order should be the same
    303. 

  303.     as it was before.</p>
    304. 

  304. 

  305.     <h3>The <b><code><a name="GSUB" id="GSUB">GSUB</a></code></b>
    306. 

  306.     table</h3>
    307. 

  307. 

  308.     <p>FontForge will read the following sub tables of the GSUB
    309. 

  309.     table:</p>
    310. 

  310. 

  311.     <table border="1">
    312. 

  312.       <tr>
    313. 

  313.         <th></th>
    314. 

  314. 

  315.         <th>name</th>
    316. 

  316. 

  317.         <th>Reading support</th>
    318. 

  318. 

  319.         <th>Writing support</th>
    320. 

  320.       </tr>
    321. 

  321. 

  322.       <tr>
    323. 

  323.         <td>1</td>
    324. 

  324. 

  325.         <td>single substitution</td>
    326. 

  326. 

  327.         <td>This sub-table allows the font designer to change from
    328. 

  328.         one glyph to another, with a context provided by the
    329. 

  329.         feature tag. For example many scripts have letters which
    330. 

  330.         have a different form at the end of a word than they do
    331. 

  331.         within (this is true of every letter in arabic, several in
    332. 

  332.         hebrew, lower case sigma in greek, and the long-s/short-s
    333. 

  333.         pair in renaissance latin). So the 'fina' feature would map
    334. 

  334.         the normal form into the final form, and the word
    335. 

  335.         processing program would do a lookup at the end of each
    336. 

  336.         word to see if a transformation was needed.</td>
    337. 

  337. 

  338.         <td>These can be created with the <a href=
    339. 

  339.         "charinfo.html">Element-&gt;Char Info</a>-&gt;Substitution
    340. 

  340.         command.</td>
    341. 

  341.       </tr>
    342. 

  342. 

  343.       <tr>
    344. 

  344.         <td>2</td>
    345. 

  345. 

  346.         <td>multiple substitution</td>
    347. 

  347. 

  348.         <td>This sub-table allows the font designer to replace one
    349. 

  349.         glyph by a series of others. This is generally used for
    350. 

  350.         rather technical layout issues.</td>
    351. 

  351. 

  352.         <td>These can be created with the <a href=
    353. 

  353.         "charinfo.html">Element-&gt;Char Info</a>-&gt;Multiple
    354. 

  354.         Substitution command.</td>
    355. 

  355.       </tr>
    356. 

  356. 

  357.       <tr>
    358. 

  358.         <td>3</td>
    359. 

  359. 

  360.         <td>alternate substitution</td>
    361. 

  361. 

  362.         <td>This sub-table allows the font designer to have a
    363. 

  363.         series of "alternates" for each glyph. One common example
    364. 

  364.         would be an italic font which had several swash variants
    365. 

  365.         for each capital letter. The word processing program would
    366. 

  366.         allow the user to choose which variant was appropriate</td>
    367. 

  367. 

  368.         <td>These can be created with the <a href=
    369. 

  369.         "charinfo.html">Element-&gt;Char Info</a>-&gt;Alternate
    370. 

  370.         Substitution command.</td>
    371. 

  371.       </tr>
    372. 

  372. 

  373.       <tr>
    374. 

  374.         <td>4</td>
    375. 

  375. 

  376.         <td>ligature substitution</td>
    377. 

  377. 

  378.         <td>This sub-table allows the font designer to replace a
    379. 

  379.         string of glyphs with another glyph. A common example is a
    380. 

  380.         ligature where the string <img src="f+i.png" width="24"
    381. 

  381.         height="25" align="middle" /> is replaced by the <img src=
    382. 

  382.         "fi.png" width="20" height="25" align="middle" />
    383. 

  383.         ligature.</td>
    384. 

  384. 

  385.         <td>These can be created with the <a href=
    386. 

  386.         "charinfo.html">Element-&gt;Char Info</a>-&gt;Ligature
    387. 

  387.         command.</td>
    388. 

  388.       </tr>
    389. 

  389. 

  390.       <tr>
    391. 

  391.         <td>5</td>
    392. 

  392. 

  393.         <td>contextual substitution</td>
    394. 

  394. 

  395.         <td>This subtable allows for a string of glyphs to replace
    396. 

  396.         another string of glyphs (or class of strings of
    397. 

  397.         glyphs)</td>
    398. 

  398. 

  399.         <td>These may be created with the <a href=
    400. 

  400.         "lookups.html#contextual-subs">Element-&gt;Font
    401. 

  401.         Info-&gt;Contextual</a> command</td>
    402. 

  402.       </tr>
    403. 

  403. 

  404.       <tr>
    405. 

  405.         <td>6</td>
    406. 

  406. 

  407.         <td>chaining contextual substitution</td>
    408. 

  408. 

  409.         <td>This is a slightly more complex version of the above,
    410. 

  410.         it doesn't really add new capabilities, but it does provide
    411. 

  411.         a more logical approach to the issue.</td>
    412. 

  412. 

  413.         <td>These may be created with the <a href=
    414. 

  414.         "lookups.html#contextual-subs">Element-&gt;Font
    415. 

  415.         Info-&gt;Contextual</a> command</td>
    416. 

  416.       </tr>
    417. 

  417. 

  418.       <tr>
    419. 

  419.         <td>7</td>
    420. 

  420. 

  421.         <td>extension positioning</td>
    422. 

  422. 

  423.         <td>This is used to allow for a GSUB table which is bigger
    424. 

  424.         than 64k. Its use should be quite invisible to the font
    425. 

  425.         designer</td>
    426. 

  426. 

  427.         <td>FontForge uses this sub-table when needed.</td>
    428. 

  428.       </tr>
    429. 

  429. 

  430.       <tr>
    431. 

  431.         <td>8</td>
    432. 

  432. 

  433.         <td>reverse chaining contextual single substitution</td>
    434. 

  434. 

  435.         <td>This allows glyph substitutions to happen in reverse
    436. 

  436.         order, and it a variant of the chaining contextual
    437. 

  437.         subtable.</td>
    438. 

  438. 

  439.         <td>These may be created with the <a href=
    440. 

  440.         "lookups.html#contextual-subs">Element-&gt;Font
    441. 

  441.         Info-&gt;Contextual</a> command</td>
    442. 

  442.       </tr>
    443. 

  443. 

  444.       <tr>
    445. 

  445.         <td>9+</td>
    446. 

  446. 

  447.         <td>reserved for future use</td>
    448. 

  448. 

  449.         <td></td>
    450. 

  450. 

  451.         <td>FontForge does not support these sub-tables yet.<br />
    452. 

  452.         (nor does anyone)</td>
    453. 

  453.       </tr>
    454. 

  454.     </table>
    455. 

  455. 

  456.     <p>FontForge also has built into it knowledge on how to provide
    457. 

  457.     default values for some features that use these tables. See the
    458. 

  458.     [Populate] button of the various <a href="lookups.html">lookup
    459. 

  459.     subtable</a> dialogs.</p>
    460. 

  460. 

  461.     <p>FontForge can produce some of these tables, but the text
    462. 

  462.     layout/word processing program used has to look up the tables
    463. 

  463.     and do the actual work of rearranging the glyphs.</p>
    464. 

  464. 

  465.     <p>FontForge will retain the order of features in the GSUB
    466. 

  466.     table, and the user may adjust it with the <a href=
    467. 

  467.     "lookups.html#Order">Element-&gt;Font Info</a> command.</p>
    468. 

  468. 

  469.     <h3>The <b><code><a name="GDEF" id="GDEF">GDEF</a></code></b>
    470. 

  470.     table</h3>
    471. 

  471. 

  472.     <p>FontForge will read ligature carets out of a GDEF table.</p>
    473. 

  473. 

  474.     <p>It will generate a GDEF table containing a glyph class
    475. 

  475.     definition sub-table (if needed) or a ligature caret sub-table
    476. 

  476.     (if needed).</p>
    477. 

  477. 

  478.     <h2><a name="AAT" id="AAT">Apple</a> Advanced Typography</h2>
    479. 

  479. 

  480.     <p>As above I do not go deeply into the abilities of these
    481. 

  481.     tables, for more information see Apple's docs:</p>
    482. 

    483. 

  483.     <ul>
    484. 

  484.       <li><a href=
    485. 

  485.       "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6bsln.html">
    486. 

  486.       The 'bsln' (baseline) table</a></li>
    487. 

  487. 

  488.       <li><a href=
    489. 

  489.       "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6kern.html">
    490. 

  490.       The 'kern' table</a></li>
    491. 

  491. 

  492.       <li><a href=
    493. 

  493.       "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6lcar.html">
    494. 

  494.       The 'lcar' (ligature caret) table</a></li>
    495. 

  495. 

  496.       <li>
    497. 

  497.         <a href=
    498. 

  498.         "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6morx.html">
    499. 

  499.         The 'morx' (extended glyph metamorphosis) table</a>
    500. 

  500. 

  501.         <ul>
    502. 

  502.           <li><a href=
    503. 

  503.           "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6mort.html">
    504. 

  504.           The 'mort' (older version of 'morx') table</a></li>
    505. 

  505. 

  506.           <li><a href=
    507. 

  507.           "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6feat.html">
    508. 

  508.           The 'feat' (feature) table</a></li>
    509. 

  509. 

  510.           <li><a href=
    511. 

  511.           "http://developer.apple.com/fonts/Registry/index.html">Apple's
    512. 

  512.           Font Feature registry</a></li>
    513. 

  513. 

  514.           <li><a href=
    515. 

  515.           "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6Tables.html">
    516. 

  516.           Description of the subtables of the 'mort'/'morx'
    517. 

  517.           tables</a></li>
    518. 

  518.         </ul>
    519. 

  519.       </li>
    520. 

  520. 

  521.       <li><a href=
    522. 

  522.       "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6opbd.html">
    523. 

  523.       The 'opbd' (optical bounds) table</a></li>
    524. 

  524. 

  525.       <li><a href=
    526. 

  526.       "http://developer.apple.com/fonts/TTRefMan/RM06/Chap6prop.html">
    527. 

  527.       The 'prop' (glyph properties) table</a></li>
    528. 

  528.     </ul>
    529. 

  529. 

  530.     <p>FontForge will currently read and produce (if Apple mode is
    531. 

  531.     set in font generation) the following tables:</p>
    532. 

  532. 

  533.     <table border="1" cellpadding="2">
    534. 

  534.       <caption>
    535. 

  535.         Apple tables corresponding vaguely to BASE
    536. 

  536.       </caption>
    537. 

  537. 

  538.       <tr>
    539. 

  539.         <th>tag</th>
    540. 

  540. 

  541.         <th>name</th>
    542. 

  542. 

  543.         <th>Reading support</th>
    544. 

  544. 

  545.         <th>Writing support</th>
    546. 

  546.       </tr>
    547. 

  547. 

  548.       <tr>
    549. 

  549.         <td><code>'bsln'</code></td>
    550. 

  550. 

  551.         <td>baseline table</td>
    552. 

  552. 

  553.         <td>FontForge will read baseline data (except for Apple's
    554. 

  554.         ideographic centered baseline, for which there is no
    555. 

  555.         OpenType equivalent)</td>
    556. 

  556. 

  557.         <td>FontForge will produce this table if the user has
    558. 

  558.         specified baseline data which apple supports</td>
    559. 

  559.       </tr>
    560. 

  560.     </table>
    561. 

  561. 

  562.     <table border="1" cellpadding="2">
    563. 

  563.       <caption>
    564. 

  564.         Apple tables corresponding vaguely to <a href=
    565. 

  565.         "#GDEF">GDEF</a>
    566. 

  566.       </caption>
    567. 

  567. 

  568.       <tr>
    569. 

  569.         <th>tag</th>
    570. 

  570. 

  571.         <th>name</th>
    572. 

  572. 

  573.         <th>Reading support</th>
    574. 

  574. 

  575.         <th>Writing support</th>
    576. 

  576.       </tr>
    577. 

  577. 

  578.       <tr>
    579. 

  579.         <td><code>'lcar'</code></td>
    580. 

  580. 

  581.         <td>ligature caret table</td>
    582. 

  582. 

  583.         <td>FontForge will read ligature carets</td>
    584. 

  584. 

  585.         <td>FontForge will produce this table if the user has
    586. 

  586.         specified ligature carets</td>
    587. 

  587.       </tr>
    588. 

  588. 

  589.       <tr>
    590. 

  590.         <td><code>'prop'</code></td>
    591. 

  591. 

  592.         <td>glyph properties table</td>
    593. 

  593. 

  594.         <td>FontForge will read this table to figure out which
    595. 

  595.         glyphs are hebrew and arabic, and which have 'r2la'
    596. 

  596.         substitutions.</td>
    597. 

  597. 

  598.         <td>FontForge will generate this table if the font contains
    599. 

  599.         some right to left glyphs.</td>
    600. 

  600.       </tr>
    601. 

  601.     </table>
    602. 

  602. 

  603.     <table border="1" cellpadding="2">
    604. 

  604.       <caption>
    605. 

  605.         Apple tables corresponding vaguely to <a href=
    606. 

  606.         "#GPOS">GPOS</a>
    607. 

  607.       </caption>
    608. 

  608. 

  609.       <tr>
    610. 

  610.         <th>tag</th>
    611. 

  611. 

  612.         <th>name</th>
    613. 

  613. 

  614.         <th>Reading support</th>
    615. 

  615. 

  616.         <th>Writing support</th>
    617. 

  617.       </tr>
    618. 

  618. 

  619.       <tr>
    620. 

  620.         <td><code>'kern'</code></td>
    621. 

  621. 

  622.         <td>kerning table</td>
    623. 

  623. 

  624.         <td>FontForge will read horizontal/vertical kerning pairs
    625. 

  625.         and classes. FontForge can read contextual kerning
    626. 

  626.         information too into a state machine.</td>
    627. 

  627. 

  628.         <td>FontForge will produce this if the font contains
    629. 

  629.         kerning data -- kerning pairs, kerning by classes, and
    630. 

  630.         kerning by state machine.</td>
    631. 

  631.       </tr>
    632. 

  632. 

  633.       <tr>
    634. 

  634.         <td><code>'opbd'</code></td>
    635. 

  635. 

  636.         <td>Optical bounds table</td>
    637. 

  637. 

  638.         <td>FontForge will read optical bounds</td>
    639. 

  639. 

  640.         <td>FontForge will produce this table if the user has
    641. 

  641.         specified right and left bounds as simple positions ('lfbd'
    642. 

  642.         and 'rtbd').</td>
    643. 

  643.       </tr>
    644. 

  644.     </table>
    645. 

  645. 

  646.     <p>FontForge has support for the <code>'mort'</code> and
    647. 

  647.     <code>'<a name="morx" id="morx">morx</a>'</code> tables (Glyph
    648. 

  648.     metamorphosis and extended glyph metamorphosis tables). These
    649. 

  649.     correspond vaguely to the <a href="#GSUB">GSUB</a> table. Note:
    650. 

  650.     Any feature type/setting combinations which correspond directly
    651. 

  651.     to an open type feature will be converted to the opentype tag
    652. 

  652.     when read in. It will be converted back to a feature/setting
    653. 

  653.     when an apple font is generated (use <a href=
    654. 

  654.     "prefs.html#Mac">File-&gt;Preferences</a> to extend FontForge's
    655. 

  655.     mapping from feature type/setting to opentype tags).</p>
    656. 

  656. 

  657.     <table border="1" cellpadding="2">
    658. 

  658.       <caption>
    659. 

  659.         Sub tables of <code>'mort'</code> or <code>'morx'</code>
    660. 

  660.       </caption>
    661. 

  661. 

  662.       <tr>
    663. 

  663.         <th></th>
    664. 

  664. 

  665.         <th>name</th>
    666. 

  666. 

  667.         <th>Reading support</th>
    668. 

  668. 

  669.         <th>Writing support</th>
    670. 

  670.       </tr>
    671. 

  671. 

  672.       <tr>
    673. 

  673.         <td>0</td>
    674. 

  674. 

  675.         <td>Indic style rearrangement</td>
    676. 

  676. 

  677.         <td>FontForge can read these and stores them as state
    678. 

  678.         machines (which can be edited with <a href=
    679. 

  679.         "lookups.html#sm-subs">Font Info</a>)</td>
    680. 

  680. 

  681.         <td>Any indic state machines will be output in the
    682. 

  682.         generated font.</td>
    683. 

  683.       </tr>
    684. 

  684. 

  685.       <tr>
    686. 

  686.         <td>1</td>
    687. 

  687. 

  688.         <td>contextual glyph substitution</td>
    689. 

  689. 

  690.         <td>FontForge can read these and stores them as state
    691. 

  691.         machines (which can be edited with <a href=
    692. 

  692.         "lookups.html#sm-subs">Font Info</a>)</td>
    693. 

  693. 

  694.         <td>
    695. 

  695.           If the font contains any state machines they will be
    696. 

  696.           output here. If there are no state machines then the
    697. 

  697.           following conversions of opentype features will be done:
    698. 

  698. 

  699.           <ul>
    700. 

  700.             <li>FontForge will generate a <a name=
    701. 

  701.             "cursive-connection" id="cursive-connection">cursive
    702. 

  702.             connection</a> feature using this subtable type if the
    703. 

  703.             font contains 'init', 'medi', 'fina' or 'isol' simple
    704. 

  704.             substitutions.</li>
    705. 

  705. 

  706.             <li>In <a href="gposgsub.html#sometimes">some cases</a>
    707. 

  707.             FontForge is able to convert an OpenType
    708. 

  708.             Contextual/Chaining substitution table into an Apple
    709. 

  709.             contextual glyph substitution table.</li>
    710. 

  710.           </ul>
    711. 

  711.         </td>
    712. 

  712.       </tr>
    713. 

  713. 

  714.       <tr>
    715. 

  715.         <td>2</td>
    716. 

  716. 

  717.         <td>ligature substitution</td>
    718. 

  718. 

  719.         <td>FontForge can read the unconditional information from
    720. 

  720.         these and stores them as opentype ligatures (which can be
    721. 

  721.         edited with <a href="lookups.html">Font Info</a> or
    722. 

  722.         <a href="charinfo.html#lookups">Char Info</a>).</td>
    723. 

  723. 

  724.         <td>If there are any ligatures with an apple
    725. 

  725.         feature/setting (or which have an opentype tag which can be
    726. 

  726.         converted to an apple feature/setting) then this table will
    727. 

  727.         be output.</td>
    728. 

  728.       </tr>
    729. 

  729. 

  730.       <tr>
    731. 

  731.         <td>4</td>
    732. 

  732. 

  733.         <td>non-contextual glyph substitution</td>
    734. 

  734. 

  735.         <td>FontForge can read these and stores them as opentype
    736. 

  736.         simple substitutions (which can be edited with <a href=
    737. 

  737.         "lookups.html">Font Info</a> or <a href=
    738. 

  738.         "charinfo.html#lookups">Char Info</a>)</td>
    739. 

  739. 

  740.         <td>If there are any substitutions with an apple
    741. 

  741.         feature/setting (or which have an opentype tag which can be
    742. 

  742.         converted to an apple feature/setting) then this table will
    743. 

  743.         be output.</td>
    744. 

  744.       </tr>
    745. 

  745. 

  746.       <tr>
    747. 

  747.         <td>5</td>
    748. 

  748. 

  749.         <td>contextual glyph insertion</td>
    750. 

  750. 

  751.         <td>FontForge can read these and stores them as state
    752. 

  752.         machines (which can be edited with <a href=
    753. 

  753.         "lookups.html#sm-subs">Font Info</a>)</td>
    754. 

  754. 

  755.         <td>Any glyph insertion state machines will be output in
    756. 

  756.         the generated font.</td>
    757. 

  757.       </tr>
    758. 

  758.     </table>
    759. 

  759. 

  760.     <h3>What features can be <a name="Conversion" id=
    761. 

  761.     "Conversion">interconverted</a> between OpenType and AAT?</h3>
    762. 

  762. 

  763.     <p>Some features have almost the same meaning in OpenType and
    764. 

  764.     AAT (although they are expressed quite differently), others are
    765. 

  765.     similar enough that they can sometimes be converted, and others
    766. 

  766.     have essentially no common ground.</p>
    767. 

  767. 

  768.     <table border="1" cellpadding="2">
    769. 

  769.       <tr>
    770. 

  770.         <th>OT Table</th>
    771. 

  771. 

  772.         <th>AAT Table</th>
    773. 

  773. 

  774.         <th>
    775. 

  775.           <p class="c2">Description</p>
    776. 

  776.         </th>
    777. 

  777.       </tr>
    778. 

  778. 

  779.       <tr>
    780. 

  780.         <td valign="top">
    781. 

  781.           <p class="c3">GDEF</p>
    782. 

  782.         </td>
    783. 

  783. 

  784.         <td valign="top">
    785. 

  785.           <p class="c3">lcar</p>
    786. 

  786.         </td>
    787. 

  787. 

  788.         <td>The ligature caret information in both 'GDEF' and
    789. 

  789.         'lcar' is essentially identical and FontForge has no
    790. 

  790.         trouble reading both and converting from one to the
    791. 

  791.         other.</td>
    792. 

  792.       </tr>
    793. 

  793. 

  794.       <tr>
    795. 

  795.         <td valign="top">
    796. 

  796.           <p class="c3">BASE</p>
    797. 

  797.         </td>
    798. 

  798. 

  799.         <td valign="top">
    800. 

  800.           <p class="c3">bsln</p>
    801. 

  801.         </td>
    802. 

  802. 

  803.         <td>There is slightly different baseline data in the two
    804. 

  804.         formats. 'bsln' does not provide extent information. 'bsln'
    805. 

  805.         provides a baseline for every glyph, while 'BASE' provides
    806. 

  806.         a baseline for every script -- one hopes all glyphs in a
    807. 

  807.         script will have the same baseline, but it isn't
    808. 

  808.         guaranteed. Finally 'bsln' and 'BASE' provide a slightly
    809. 

  809.         different set of baseline tags, and FontForge only supports
    810. 

  810.         the OpenType ones. In particular Apple's ideographic
    811. 

  811.         centered baseline will be lost.</td>
    812. 

  812.       </tr>
    813. 

  813. 

  814.       <tr>
    815. 

  815.         <td>
    816. 

  816.           <p class="c3">GPOS</p>
    817. 

  817.         </td>
    818. 

  818. 

  819.         <td>
    820. 

  820.           <p class="c3">kern</p>
    821. 

  821.         </td>
    822. 

  822. 

  823.         <td>
    824. 

  824.           In most cases kerning information can be converted from
    825. 

  825.           one format to another. Both provide support for vertical
    826. 

  826.           kerning and right to left kerning. Both provide support
    827. 

  827.           for kerning by glyph pair and kerning by classes.
    828. 

  828. 

  829.           <p>OpenType allows kerning commands to be supplied via a
    830. 

  830.           contextual chaining feature, Apple allows them to be
    831. 

  831.           controled by a state machine. FontForge supports both,
    832. 

  832.           but does not interconvert.</p>
    833. 

  833.         </td>
    834. 

  834.       </tr>
    835. 

  835. 

  836.       <tr>
    837. 

  837.         <td>
    838. 

  838.           <p class="c3">GPOS</p>
    839. 

  839.         </td>
    840. 

  840. 

  841.         <td>
    842. 

  842.           <p class="c3">opbd</p>
    843. 

  843.         </td>
    844. 

  844. 

  845.         <td>The GPOS features 'lfbd' and 'rtbd' provide the
    846. 

  846.         information needed to generate an Apple opbd table. If
    847. 

  847.         FontForge reads a font with an opbd table it will generate
    848. 

  848.         appropriate 'lfbd' and 'rtbd' features. If FontForge
    849. 

  849.         generates a font in apple mode that has these features it
    850. 

  850.         will create an opbd table. Similarly when FontForge reads
    851. 

  851.         an opbd table it will create 'lfbd' and 'rtbd'
    852. 

  852.         features.</td>
    853. 

  853.       </tr>
    854. 

  854. 

  855.       <tr>
    856. 

  856.         <td>
    857. 

  857.           <p class="c3">GPOS</p>
    858. 

  858.         </td>
    859. 

  859. 

  860.         <td>
    861. 

  861.           <hr />
    862. 

  862.         </td>
    863. 

  863. 

  864.         <td>I am not aware of any way to convert other GPOS
    865. 

  865.         features to AAT.</td>
    866. 

  866.       </tr>
    867. 

  867. 

  868.       <tr>
    869. 

  869.         <td>
    870. 

  870.           <p class="c3">GSUB</p>
    871. 

  871.         </td>
    872. 

  872. 

  873.         <td>
    874. 

  874.           <p class="c3">morx</p>
    875. 

  875.         </td>
    876. 

  876. 

  877.         <td rowspan="2">The 'mort' and 'morx' tables have the same
    878. 

  878.         capabilities ('mort' tables are an old format and Apple
    879. 

  879.         currently encourages that 'morx' tables be used instead).
    880. 

  880.         FontForge can read either one, but only generates 'morx'
    881. 

  881.         tables. Interconversion depends on specific feature types
    882. 

  882.         and the sub-table formats, see below</td>
    883. 

  883.       </tr>
    884. 

  884. 

  885.       <tr>
    886. 

  886.         <td>
    887. 

  887.           <p class="c3">GSUB</p>
    888. 

  888.         </td>
    889. 

  889. 

  890.         <td>
    891. 

  891.           <p class="c3">mort</p>
    892. 

  892.         </td>
    893. 

  893.       </tr>
    894. 

  894.     </table>
    895. 

  895. 

  896.     <h4>An analysis of GSUB and morx sub-tables and feature
    897. 

  897.     tags</h4>
    898. 

  898. 

  899.     <p>OpenType uses a four character feature tag (like 'liga')
    900. 

  900.     while Apple uses two numbers to represent a feature setting
    901. 

  901.     (&lt;1,2&gt;). For FontForge to be able to inter-convert an
    902. 

  902.     OpenType feature into an Apple feature there must first be a
    903. 

  903.     correspondence between the two naming conventions. Sometimes
    904. 

  904.     there is an easy direct conversion (above 'liga' and
    905. 

  905.     &lt;1,2&gt; both represent "Common Ligatures") but far more
    906. 

  906.     often there is none. See <a href=
    907. 

  907.     "gposgsub.html#OT-Mac-features">below</a> for a list of the
    908. 

  908.     tags and feature settings that FontForge considers similar
    909. 

  909.     enough to interconvert.</p>
    910. 

  910. 

  911.     <p>GSUB tables have 7 sub-table formats, while morx tables have
    912. 

  912.     5.</p>
    913. 

  913. 

  914.     <table border="1" cellpadding="2">
    915. 

  915.       <tr>
    916. 

  916.         <th>GSUB<br />
    917. 

  917.         sub-<br />
    918. 

  918.         table</th>
    919. 

  919. 

  920.         <th>morx<br />
    921. 

  921.         sub-<br />
    922. 

  922.         table</th>
    923. 

  923. 

  924.         <th>
    925. 

  925.           <p class="c2">Description</p>
    926. 

  926.         </th>
    927. 

  927.       </tr>
    928. 

  928. 

  929.       <tr>
    930. 

  930.         <td valign="top">
    931. 

  931.           <p class="c3">Single</p>
    932. 

  932.         </td>
    933. 

  933. 

  934.         <td valign="top">
    935. 

  935.           <p class="c3">Non-<br />
    936. 

  936.           Contextual<br />
    937. 

  937.           Glyph</p>
    938. 

  938.         </td>
    939. 

  939. 

  940.         <td>These two sub-tables have almost exactly the same
    941. 

  941.         capabilities. Each allows one glyph to be substituted for
    942. 

  942.         another. The morx sub-table also allows a glyph to be
    943. 

  943.         deleted, while the GSUB sub-table does not.</td>
    944. 

  944.       </tr>
    945. 

  945. 

  946.       <tr>
    947. 

  947.         <td valign="top">
    948. 

  948.           <p class="c3">Multiple</p>
    949. 

  949.         </td>
    950. 

  950. 

  951.         <td></td>
    952. 

  952. 

  953.         <td>
    954. 

  954.           This GSUB subtable allows a single glyph to be replaced
    955. 

  955.           by multiple glyphs. It has some similarities to Apple's
    956. 

  956.           Glyph Insertion sub-table except:
    957. 

  957. 

  958.           <ul>
    959. 

  959.             <li>the 'morx' sub-table always leaves the current
    960. 

  960.             glyph in the glyph stream, while this sub-table need
    961. 

  961.             not</li>
    962. 

  962. 

  963.             <li>the 'morx' sub-table is contextual while this
    964. 

  964.             sub-table is never. (But if this sub-table is wrapped
    965. 

  965.             inside a Context or Chaining Context subtable the
    966. 

  966.             result can be contextual).</li>
    967. 

  967.           </ul>
    968. 

  968.         </td>
    969. 

  969.       </tr>
    970. 

  970. 

  971.       <tr>
    972. 

  972.         <td></td>
    973. 

  973. 

  974.         <td valign="top">
    975. 

  975.           <p class="c3">Glyph<br />
    976. 

  976.           Insertion</p>
    977. 

  977.         </td>
    978. 

  978. 

  979.         <td>This morx subtable allows a string of glyphs to be
    980. 

  980.         inserted before or after the current glyph (the current
    981. 

  981.         glyph always remains). This sub-table is contextual (ie.
    982. 

  982.         the insertion can be restricted to certain contexts). It
    983. 

  983.         bears some similarities to the GSUB Multiple subtable
    984. 

  984.         above.</td>
    985. 

  985.       </tr>
    986. 

  986. 

  987.       <tr>
    988. 

  988.         <td valign="top">Alternate</td>
    989. 

  989. 

  990.         <td></td>
    991. 

  991. 

  992.         <td>This GSUB subtable allows a single glyph to be replaced
    993. 

  993.         by any one of several alternatives (presumably with help
    994. 

  994.         from a word processor's UI). An example of this would be a
    995. 

  995.         character which had several swash variants. There is
    996. 

  996.         nothing like this in the 'morx' table.</td>
    997. 

  997.       </tr>
    998. 

  998. 

  999.       <tr>
    1000. 

  1000.         <td valign="top">
    1001. 

  1001.           <p class="c3">Ligature</p>
    1002. 

  1002.         </td>
    1003. 

  1003. 

  1004.         <td valign="top">
    1005. 

  1005.           <p class="c3">Ligature</p>
    1006. 

  1006.         </td>
    1007. 

  1007. 

  1008.         <td>
    1009. 

  1009.           Both formats have ligature sub-tables. The 'GSUB' version
    1010. 

  1010.           is unconditional (the ligature is always applied --
    1011. 

  1011.           though a ligature substitution could be embedded in an
    1012. 

  1012.           OpenType contextual substitution to make it condtional).
    1013. 

  1013.           The 'morx' version can be contextual (though in fonts I
    1014. 

  1014.           have examined it is usually uncontextual). FontForge only
    1015. 

  1015.           supports unconditional ligatures.
    1016. 

  1016. 

  1017.           <p>FontForge can read all the unconditional ligatures
    1018. 

  1018.           from a 'morx' sub-table. FontForge loses all contextual
    1019. 

  1019.           ligatures.</p>
    1020. 

  1020. 

  1021.           <p>In OpenType, contextual ligatures can be built by
    1022. 

  1022.           wrapping a ligature sub-table inside a Context or
    1023. 

  1023.           Chaining Context subtable.</p>
    1024. 

  1024.         </td>
    1025. 

  1025.       </tr>
    1026. 

  1026. 

  1027.       <tr>
    1028. 

  1028.         <td></td>
    1029. 

  1029. 

  1030.         <td valign="top">
    1031. 

  1031.           <p class="c3">Contextual<br />
    1032. 

  1032.           Glyph</p>
    1033. 

  1033.         </td>
    1034. 

  1034. 

  1035.         <td>This morx subtable allows single glyph substitutions to
    1036. 

  1036.         be applied within certain contexts. At first glance it
    1037. 

  1037.         seems that this could be converted into an opentype Context
    1038. 

  1038.         subtable, <a href="gposgsub.html#sometimes">but this is
    1039. 

  1039.         rarely the case</a>.</td>
    1040. 

  1040.       </tr>
    1041. 

  1041. 

  1042.       <tr>
    1043. 

  1043.         <td>
    1044. 

  1044.           <p class="c3">Context</p>
    1045. 

  1045.         </td>
    1046. 

  1046. 

  1047.         <td></td>
    1048. 

  1048. 

  1049.         <td rowspan="2" valign="top">These GSUB subtables allow any
    1050. 

  1050.         collection of other substitutions to be applied
    1051. 

  1051.         contextually. At first glance one might think that these
    1052. 

  1052.         (with appropriate nested substitutions) might be converted
    1053. 

  1053.         to 'morx' contextual glyph substitutions, contextual
    1054. 

  1054.         ligatures, or even glyph insertion. <a href=
    1055. 

  1055.         "gposgsub.html#sometimes">Unfortunately this is rarely the
    1056. 

  1056.         case</a>.</td>
    1057. 

  1057.       </tr>
    1058. 

  1058. 

  1059.       <tr>
    1060. 

  1060.         <td>
    1061. 

  1061.           <p class="c3">Chaining<br />
    1062. 

  1062.           Context</p>
    1063. 

  1063.         </td>
    1064. 

  1064. 

  1065.         <td></td>
    1066. 

  1066.       </tr>
    1067. 

  1067. 

  1068.       <tr valign="top">
    1069. 

  1069.         <td>
    1070. 

  1070.           <p class="c3">Reverse<br />
    1071. 

  1071.           Chaining<br />
    1072. 

  1072.           Context</p>
    1073. 

  1073.         </td>
    1074. 

  1074. 

  1075.         <td></td>
    1076. 

  1076. 

  1077.         <td>This GSUB subtable is applied backwards to the stream
    1078. 

  1078.         of glyphs, it allows a single glyph substitution per
    1079. 

  1079.         contextual match. There is nothing like it in 'morx'.</td>
    1080. 

  1080.       </tr>
    1081. 

  1081. 

  1082.       <tr valign="top">
    1083. 

  1083.         <td></td>
    1084. 

  1084. 

  1085.         <td>
    1086. 

  1086.           <p class="c3">Indic<br />
    1087. 

  1087.           Rearrange-<br />
    1088. 

  1088.           ment</p>
    1089. 

  1089.         </td>
    1090. 

  1090. 

  1091.         <td>This 'morx' subtable allows for several glyphs to
    1092. 

  1092.         interchange their positions in the glyph stream. There is
    1093. 

  1093.         nothing like it in GSUB (or GPOS for that matter).</td>
    1094. 

  1094.       </tr>
    1095. 

  1095.     </table>
    1096. 

  1096. 

  1097.     <h4>Why do contextual glyph substitutions only <a name=
    1098. 

  1098.     "sometimes" id="sometimes">sometimes</a> get generated in
    1099. 

  1099.     AAT?</h4>
    1100. 

  1100. 

  1101.     <p>Sadly OpenType and AAT provide disjoint capabilities when it
    1102. 

  1102.     comes to contextual matching. AAT is more capable in some
    1103. 

  1103.     areas, OpenType more capable in others. FontForge is able to
    1104. 

  1104.     convert an OpenType contextual substitution into an AAT one if
    1105. 

  1105.     FontForge can detect that the OpenType substitution does not
    1106. 

  1106.     use capabilities beyond those of AAT. Currently this means:</p>
    1107. 

  1107. 

  1108.     <ul>
    1109. 

  1109.       <li>There is an apple feature which matches the otf tag</li>
    1110. 

  1110. 

  1111.       <li>And one of the following is true:
    1112. 

  1112. 

  1113.         <ol>
    1114. 

  1114.           <li>Either
    1115. 

  1115. 

  1116.             <ul>
    1117. 

  1117.               <li>The sub-table is in coverage format</li>
    1118. 

  1118. 

  1119.               <li>The sub-table contains either exactly one nested
    1120. 

  1120.               single glyph replacement substitution, or<br />
    1121. 

  1121.               it contains exactly two single glyph replacements and
    1122. 

  1122.               one of them refers to the last glyph matched (and the
    1123. 

  1123.               other does not)</li>
    1124. 

  1124.             </ul>
    1125. 

  1125.           </li>
    1126. 

  1126. 

  1127.           <li>or
    1128. 

  1128. 

  1129.             <ul>
    1130. 

  1130.               <li>The sub-table is in either glyph or class
    1131. 

  1131.               format</li>
    1132. 

  1132. 

  1133.               <li>If in class format then either the backtrack and
    1134. 

  1134.               lookahead classes must be the same as the main class,
    1135. 

  1135.               or they must not be used.</li>
    1136. 

  1136. 

  1137.               <li>If a rule has a substitution at a given glyph
    1138. 

  1138.               position, then all rules which match the current rule
    1139. 

  1139.               up to that glyph position must also have a
    1140. 

  1140.               substitution at that position.</li>
    1141. 

  1141. 

  1142.               <li>A rule with exactly one substitution is
    1143. 

  1143.               acceptable<br />
    1144. 

  1144.                 A rule with one substitution in the middle and one
    1145. 

  1145.                 substitution on the last glyph is acceptable.<br />
    1146. 

  1146.                 A rule may contain more substitutions only if there
    1147. 

  1147.                 is another rule which matches it exactly up to the
    1148. 

  1148.                 internal substitution.<br />
    1149. 

  1149.                 So the following rule set is valid:
    1150. 

  1150. 

  1151.                 <table border="1" cellpadding="2">
    1152. 

  1152.                   <tr>
    1153. 

  1153.                     <th>Rule</th>
    1154. 

  1154. 

  1155.                     <td>a</td>
    1156. 

  1156. 

  1157.                     <td>b</td>
    1158. 

  1158. 

  1159.                     <td>c</td>
    1160. 

  1160. 

  1161.                     <td>d</td>
    1162. 

  1162. 

  1163.                     <td>e</td>
    1164. 

  1164. 

  1165.                     <td>f</td>
    1166. 

  1166.                   </tr>
    1167. 

  1167. 

  1168.                   <tr>
    1169. 

  1169.                     <th>Rule</th>
    1170. 

  1170. 

  1171.                     <td>a</td>
    1172. 

  1172. 

  1173.                     <td>b</td>
    1174. 

  1174. 

  1175.                     <td>c</td>
    1176. 

  1176. 

  1177.                     <td>d</td>
    1178. 

  1178. 

  1179.                     <td></td>
    1180. 

  1180. 

  1181.                     <td></td>
    1182. 

  1182.                   </tr>
    1183. 

  1183. 

  1184.                   <tr>
    1185. 

  1185.                     <th>Rule</th>
    1186. 

  1186. 

  1187.                     <td>a</td>
    1188. 

  1188. 

  1189.                     <td>b</td>
    1190. 

  1190. 

  1191.                     <td></td>
    1192. 

  1192. 

  1193.                     <td></td>
    1194. 

  1194. 

  1195.                     <td></td>
    1196. 

  1196. 

  1197.                     <td></td>
    1198. 

  1198.                   </tr>
    1199. 

  1199. 

  1200.                   <tr>
    1201. 

  1201.                     <th>Substitutions</th>
    1202. 

  1202. 

  1203.                     <td>A</td>
    1204. 

  1204. 

  1205.                     <td>B</td>
    1206. 

  1206. 

  1207.                     <td>C</td>
    1208. 

  1208. 

  1209.                     <td>D</td>
    1210. 

  1210. 

  1211.                     <td>E</td>
    1212. 

  1212. 

  1213.                     <td>F</td>
    1214. 

  1214.                   </tr>
    1215. 

  1215.                 </table>
    1216. 

  1216. 

  1217.                 <p>So the third rule will match an "ab" and convert
    1218. 

  1218.                 them to "AB" (and this is valid because we have one
    1219. 

  1219.                 internal and one final substitution and that's ok),
    1220. 

  1220.                 then if that "ab" is followed by "cd" then rule 2
    1221. 

  1221.                 kicks in and will replace the "cd" with "CD" (again
    1222. 

  1222.                 this has one internal and one final substitution,
    1223. 

  1223.                 which is ok), and if that is followed by "ef" then
    1224. 

  1224.                 they will be converted to "EF".</p>
    1225. 

  1225. 

  1226.                 <p>The following is not valid:</p>
    1227. 

  1227. 

  1228.                 <table border="1" cellpadding="2">
    1229. 

  1229.                   <tr>
    1230. 

  1230.                     <th>Substitution</th>
    1231. 

  1231. 

  1232.                     <td></td>
    1233. 

  1233. 

  1234.                     <td>B</td>
    1235. 

  1235. 

  1236.                     <td></td>
    1237. 

  1237.                   </tr>
    1238. 

  1238. 

  1239.                   <tr>
    1240. 

  1240.                     <th>Rule</th>
    1241. 

  1241. 

  1242.                     <td>a</td>
    1243. 

  1243. 

  1244.                     <td>b</td>
    1245. 

  1245. 

  1246.                     <td>c</td>
    1247. 

  1247.                   </tr>
    1248. 

  1248. 

  1249.                   <tr>
    1250. 

  1250.                     <td colspan="4">
    1251. 

  1251.                       <hr />
    1252. 

  1252.                     </td>
    1253. 

  1253.                   </tr>
    1254. 

  1254. 

  1255.                   <tr>
    1256. 

  1256.                     <th>Rule</th>
    1257. 

  1257. 

  1258.                     <td>a</td>
    1259. 

  1259. 

  1260.                     <td>b</td>
    1261. 

  1261. 

  1262.                     <td></td>
    1263. 

  1263.                   </tr>
    1264. 

  1264. 

  1265.                   <tr>
    1266. 

  1266.                     <th>Substitution</th>
    1267. 

  1267. 

  1268.                     <td>A</td>
    1269. 

  1269. 

  1270.                     <td></td>
    1271. 

  1271. 

  1272.                     <td></td>
    1273. 

  1273.                   </tr>
    1274. 

  1274.                 </table>
    1275. 

  1275. 

  1276.                 <p>The two rules have substitutions at different
    1277. 

  1277.                 places and that can't be expressed in an Apple
    1278. 

  1278.                 state machine given that they have the same
    1279. 

  1279.                 glyphs.</p>
    1280. 

  1280.               </li>
    1281. 

  1281.             </ul>
    1282. 

  1282.           </li>
    1283. 

  1283.         </ol>
    1284. 

  1284.       </li>
    1285. 

  1285.     </ul>
    1286. 

  1286. 

  1287.     <p>FontForge does not even try to convert an AAT contextual
    1288. 

  1288.     glyph substitution sub-table, too few of these can be expressed
    1289. 

  1289.     in OpenType to make it worth while.</p>
    1290. 

  1290. 

  1291.     <p>NOTE: It would be possible to convert more lookups to state
    1292. 

  1292.     machines if FontForge were willing to:</p>
    1293. 

  1293. 

  1294.     <ol>
    1295. 

  1295.       <li>Use several state machines to represent complicated
    1296. 

  1296.       lookups</li>
    1297. 

  1297. 

  1298.       <li>Add additional glyphs to the font to be used as temporary
    1299. 

  1299.       state flags.</li>
    1300. 

  1300.     </ol>
    1301. 

  1301. 

  1302.     <p>FontForge will do neither of these.</p>
    1303. 

  1303. 

  1304.     <h4 class="c4">BUG</h4>
    1305. 

  1305. 

  1306.     <p>There is a subtle bug involved in converting a chaining
    1307. 

  1307.     contextual substitution into an Apple contextual glyph
    1308. 

  1308.     substitution. AAT does not have the concept of a backtrack
    1309. 

  1309.     list, this means that substitutions may occur in a different
    1310. 

  1310.     order.</p>
    1311. 

  1311. 

  1312.     <h4>Why can't all contextual/chaining tables be <a name=
    1313. 

  1313.     "not-converted" id="not-converted">converted</a>?</h4>
    1314. 

  1314. 

  1315.     <p>Well, obviously there are some thing that just aren't
    1316. 

  1316.     present. The concept of contextual positioning is missing from
    1317. 

  1317.     AAT, while Indic rearrangement is missing from OpenType. So
    1318. 

  1318.     let's concentrate on contextual substitutions, which both
    1319. 

  1319.     appear to support. The argument that follows is based on the
    1320. 

  1320.     capabilities of contextual matching, it applies equally to
    1321. 

  1321.     contextual ligatures, glyph insertion, glyph substitution and
    1322. 

  1322.     kerning, the examples given are only of glyph substitution
    1323. 

  1323.     because it is easier to represent (and because FontForge is
    1324. 

  1324.     only willing to convert contextual glyph substitutions) But
    1325. 

  1325.     even here, there is a very basic mismatch in concepts between
    1326. 

  1326.     the way OpenType and Apple specify contextual substitutions.
    1327. 

  1327.     Consider the following contextual substitution in a glyph list
    1328. 

  1328.     format:</p>
    1329. 

  1329. 

  1330.     <table border="1" cellpadding="2">
    1331. 

  1331.       <tr>
    1332. 

  1332.         <th>Initial Sequence</th>
    1333. 

  1333. 

  1334.         <td>a</td>
    1335. 

  1335. 

  1336.         <td>b</td>
    1337. 

  1337. 

  1338.         <td>c</td>
    1339. 

  1339. 

  1340.         <td>d</td>
    1341. 

  1341.       </tr>
    1342. 

  1342. 

  1343.       <tr>
    1344. 

  1344.         <th>
    1345. 

  1345.           <p class="c2">Replace With</p>
    1346. 

  1346.         </th>
    1347. 

  1347. 

  1348.         <td>&nbsp;</td>
    1349. 

  1349. 

  1350.         <td>B</td>
    1351. 

  1351. 

  1352.         <td>C</td>
    1353. 

  1353. 

  1354.         <td>&nbsp;</td>
    1355. 

  1355.       </tr>
    1356. 

  1356.     </table>
    1357. 

  1357. 

  1358.     <p>Now in OpenType this means if you find the sequence "abcd"
    1359. 

  1359.     then replace the "b" with "B" and the "c" with "C". But this
    1360. 

  1360.     can't be expressed in an Apple state machine. In OpenType the
    1361. 

  1361.     match is done first, and then the substitutions occur. In a
    1362. 

  1362.     state machine the substitutions have to be done (almost)
    1363. 

  1363.     concurrently with the match and so must happen whether the
    1364. 

  1364.     final "d" is present or not. (Note I'm using a glyph sequence
    1365. 

  1365.     because it is easier to display in an example. The same problem
    1366. 

  1366.     applies if the substitution is expressed by classes or by
    1367. 

  1367.     coverage tables)</p>
    1368. 

  1368. 

  1369.     <p>Consider the following table with two glyph strings</p>
    1370. 

  1370. 

  1371.     <table border="1" cellpadding="2">
    1372. 

  1372.       <tr>
    1373. 

  1373.         <th>Initial Sequence</th>
    1374. 

  1374. 

  1375.         <td>a</td>
    1376. 

  1376. 

  1377.         <td>
    1378. 

  1378.           <p class="c3">b</p>
    1379. 

  1379.         </td>
    1380. 

  1380. 

  1381.         <td>
    1382. 

  1382.           <p class="c3">c</p>
    1383. 

  1383.         </td>
    1384. 

  1384. 

  1385.         <td>d</td>
    1386. 

  1386.       </tr>
    1387. 

  1387. 

  1388.       <tr>
    1389. 

  1389.         <th>
    1390. 

  1390.           <p class="c2">Replace With</p>
    1391. 

  1391.         </th>
    1392. 

  1392. 

  1393.         <td>&nbsp;</td>
    1394. 

  1394. 

  1395.         <td>
    1396. 

  1396.           <p class="c3">B</p>
    1397. 

  1397.         </td>
    1398. 

  1398. 

  1399.         <td></td>
    1400. 

  1400. 

  1401.         <td>&nbsp;</td>
    1402. 

  1402.       </tr>
    1403. 

  1403. 

  1404.       <tr>
    1405. 

  1405.         <th>Initial Sequence</th>
    1406. 

  1406. 

  1407.         <td>a</td>
    1408. 

  1408. 

  1409.         <td>
    1410. 

  1410.           <p class="c3">b</p>
    1411. 

  1411.         </td>
    1412. 

  1412. 

  1413.         <td>
    1414. 

  1414.           <p class="c3">c</p>
    1415. 

  1415.         </td>
    1416. 

  1416. 

  1417.         <td>e&nbsp;</td>
    1418. 

  1418.       </tr>
    1419. 

  1419. 

  1420.       <tr>
    1421. 

  1421.         <th>
    1422. 

  1422.           <p class="c2">Replace With</p>
    1423. 

  1423.         </th>
    1424. 

  1424. 

  1425.         <td>&nbsp;</td>
    1426. 

  1426. 

  1427.         <td></td>
    1428. 

  1428. 

  1429.         <td>
    1430. 

  1430.           <p class="c3">&nbsp;C</p>
    1431. 

  1431.         </td>
    1432. 

  1432. 

  1433.         <td>&nbsp;</td>
    1434. 

  1434.       </tr>
    1435. 

  1435.     </table>
    1436. 

  1436. 

  1437.     <p>So replace the "b" if the final "d" is present, otherwise
    1438. 

  1438.     replace the "c". Again this cannot be expressed in Apple's
    1439. 

  1439.     state machines.</p>
    1440. 

  1440. 

  1441.     <p>Finally consider</p>
    1442. 

  1442. 

  1443.     <table border="1" cellpadding="2">
    1444. 

  1444.       <tr>
    1445. 

  1445.         <th>Initial Sequence</th>
    1446. 

  1446. 

  1447.         <td>
    1448. 

  1448.           <p class="c3">a</p>
    1449. 

  1449.         </td>
    1450. 

  1450. 

  1451.         <td>
    1452. 

  1452.           <p class="c3">b</p>
    1453. 

  1453.         </td>
    1454. 

  1454. 

  1455.         <td>
    1456. 

  1456.           <p class="c3">c</p>
    1457. 

  1457.         </td>
    1458. 

  1458. 

  1459.         <td>d</td>
    1460. 

  1460.       </tr>
    1461. 

  1461. 

  1462.       <tr>
    1463. 

  1463.         <th>
    1464. 

  1464.           <p class="c2">Replace With</p>
    1465. 

  1465.         </th>
    1466. 

  1466. 

  1467.         <td>&nbsp;</td>
    1468. 

  1468. 

  1469.         <td>
    1470. 

  1470.           <p class="c3"></p>
    1471. 

  1471.         </td>
    1472. 

  1472. 

  1473.         <td>C</td>
    1474. 

  1474. 

  1475.         <td>&nbsp;</td>
    1476. 

  1476.       </tr>
    1477. 

  1477. 

  1478.       <tr>
    1479. 

  1479.         <th>Initial Sequence</th>
    1480. 

  1480. 

  1481.         <td>
    1482. 

  1482.           <p class="c3">b</p>
    1483. 

  1483.         </td>
    1484. 

  1484. 

  1485.         <td>
    1486. 

  1486.           <p class="c3">c</p>
    1487. 

  1487.         </td>
    1488. 

  1488. 

  1489.         <td>
    1490. 

  1490.           <p class="c3">e</p>
    1491. 

  1491.         </td>
    1492. 

  1492. 

  1493.         <td></td>
    1494. 

  1494.       </tr>
    1495. 

  1495. 

  1496.       <tr>
    1497. 

  1497.         <th>
    1498. 

  1498.           <p class="c2">Replace With</p>
    1499. 

  1499.         </th>
    1500. 

  1500. 

  1501.         <td>
    1502. 

  1502.           <p class="c3">B&nbsp;</p>
    1503. 

  1503.         </td>
    1504. 

  1504. 

  1505.         <td></td>
    1506. 

  1506. 

  1507.         <td>
    1508. 

  1508.           <p class="c3">&nbsp;</p>
    1509. 

  1509.         </td>
    1510. 

  1510. 

  1511.         <td>&nbsp;</td>
    1512. 

  1512.       </tr>
    1513. 

  1513.     </table>
    1514. 

  1514. 

  1515.     <p>If this substitution is given the sequence "abce" it cannot
    1516. 

  1516.     work in AAT. When it reads the "a" it will start down the
    1517. 

  1517.     "abcd" branch, the match will not fail until it looks for the
    1518. 

  1518.     "d" and finds "e" instead. At that point it is too late to
    1519. 

  1519.     switch to the "bce" substitution (which does match) because the
    1520. 

  1520.     "b" glyph will not have been marked as a substitution
    1521. 

  1521.     location.</p>
    1522. 

  1522.     <hr />
    1523. 

  1523.     On the other hand, Apple's state machines can express things
    1524. 

  1524.     that OpenType tables cannot (again I'm restricting myself to
    1525. 

  1525.     contextual glyph substitutions). Consider the case of a swash
    1526. 

  1526.     glyph at the beginning (or end) of a word. Let us say that a
    1527. 

  1527.     word begins when a letter appears at the start of input or
    1528. 

  1528.     following a whitespace character. But OpenType has no way of
    1529. 

  1529.     expressing "start of input" (or end of input) in a
    1530. 

  1530.     contextual/chaining context, whereas Apple's state machines do.
    1531. 

    1532. 

  1532.     <p>Since Apple's glyph substitutions can delete glyphs a
    1533. 

  1533.     contextual glyph substitution table can create two character
    1534. 

  1534.     ligatures (one glyph is converted to the ligature and the other
    1535. 

  1535.     is deleted), while OpenType tables must use a ligature
    1536. 

  1536.     substitution to do this.</p>
    1537. 

  1537. 

  1538.     <p>Finally an AAT state machine can match a general regular
    1539. 

  1539.     expression, while OpenType tables can only match fixed length
    1540. 

  1540.     strings. Suppose you were typesetting mathematics and wanted a
    1541. 

  1541.     substitution which would convert an arbitrary digit string
    1542. 

  1542.     following a variable into a subscript (so x23 should become
    1543. 

  1543.     x<small><sub>23</sub></small>). It is easy to write a state
    1544. 

  1544.     machine which will keep substituting as long as it gets more
    1545. 

  1545.     digits, but you'd need an infinite number of OpenType rules to
    1546. 

  1546.     have the same expressive power.</p>
    1547. 

  1547. 

  1548.     <p>These examples probably seem fairly contrived, and (except
    1549. 

  1549.     for the swash one) they are. But they illustrate the point that
    1550. 

  1550.     the two formats have very different expressive capabilities and
    1551. 

  1551.     it is NOT possible to write a converter which will take any
    1552. 

  1552.     input in one format and produce an equivalent output in the
    1553. 

  1553.     other.</p>
    1554. 

  1554. 

  1555.     <h3>Apple and OpenType features</h3>
    1556. 

  1556. 

  1557.     <table border="1" cellpadding="2">
    1558. 

  1558.       <caption>
    1559. 

  1559.         Correspondences between Apple and OpenType <a name=
    1560. 

  1560.         "OT-Mac-features" id="OT-Mac-features">features</a><br />
    1561. 

  1561.         (that I support)
    1562. 

  1562.       </caption>
    1563. 

  1563. 

  1564.       <tr>
    1565. 

  1565.         <th>Apple Feature Setting</th>
    1566. 

  1566. 

  1567.         <th>OpenType Feature Name</th>
    1568. 

  1568. 

  1569.         <th>OpenType Tag</th>
    1570. 

  1570.       </tr>
    1571. 

  1571. 

  1572.       <tr>
    1573. 

  1573.         <td>Required Ligatures</td>
    1574. 

  1574. 

  1575.         <td>Required Ligatures</td>
    1576. 

  1576. 

  1577.         <td>rlig</td>
    1578. 

  1578.       </tr>
    1579. 

  1579. 

  1580.       <tr>
    1581. 

  1581.         <td>Common Ligatures</td>
    1582. 

  1582. 

  1583.         <td>Standard Ligatures</td>
    1584. 

  1584. 

  1585.         <td>liga</td>
    1586. 

  1586.       </tr>
    1587. 

  1587. 

  1588.       <tr>
    1589. 

  1589.         <td>Rare Ligatures</td>
    1590. 

  1590. 

  1591.         <td>Discretionary</td>
    1592. 

  1592. 

  1593.         <td>dlig</td>
    1594. 

  1594.       </tr>
    1595. 

  1595. 

  1596.       <tr>
    1597. 

  1597.         <td>Fractions</td>
    1598. 

  1598. 

  1599.         <td>Fractions</td>
    1600. 

  1600. 

  1601.         <td>frac</td>
    1602. 

  1602.       </tr>
    1603. 

  1603. 

  1604.       <tr>
    1605. 

  1605.         <td colspan="3">
    1606. 

  1606.           <hr />
    1607. 

  1607.         </td>
    1608. 

  1608.       </tr>
    1609. 

  1609. 

  1610.       <tr>
    1611. 

  1611.         <td>Contextual Alternatives</td>
    1612. 

  1612. 

  1613.         <td>Cursive connection</td>
    1614. 

  1614. 

  1615.         <td>calt</td>
    1616. 

  1616.       </tr>
    1617. 

  1617. 

  1618.       <tr>
    1619. 

  1619.         <td colspan="3">
    1620. 

  1620.           <hr />
    1621. 

  1621.         </td>
    1622. 

  1622.       </tr>
    1623. 

  1623. 

  1624.       <tr>
    1625. 

  1625.         <td>Vertical Forms</td>
    1626. 

  1626. 

  1627.         <td>Vertical Rotation 2</td>
    1628. 

  1628. 

  1629.         <td>vrt2</td>
    1630. 

  1630.       </tr>
    1631. 

  1631. 

  1632.       <tr>
    1633. 

  1633.         <td>Monospace numbers</td>
    1634. 

  1634. 

  1635.         <td>Tabular numbers</td>
    1636. 

  1636. 

  1637.         <td>tnum</td>
    1638. 

  1638.       </tr>
    1639. 

  1639. 

  1640.       <tr>
    1641. 

  1641.         <td>Superscript</td>
    1642. 

  1642. 

  1643.         <td>Superscript</td>
    1644. 

  1644. 

  1645.         <td>sups</td>
    1646. 

  1646.       </tr>
    1647. 

  1647. 

  1648.       <tr>
    1649. 

  1649.         <td>Subscript</td>
    1650. 

  1650. 

  1651.         <td>Subscript</td>
    1652. 

  1652. 

  1653.         <td>subs</td>
    1654. 

  1654.       </tr>
    1655. 

  1655. 

  1656.       <tr>
    1657. 

  1657.         <td>Proportional Text</td>
    1658. 

  1658. 

  1659.         <td>Proportional Widths</td>
    1660. 

  1660. 

  1661.         <td>pwid</td>
    1662. 

  1662.       </tr>
    1663. 

  1663. 

  1664.       <tr>
    1665. 

  1665.         <td>Half-width Text</td>
    1666. 

  1666. 

  1667.         <td>Half Width</td>
    1668. 

  1668. 

  1669.         <td>hwid</td>
    1670. 

  1670.       </tr>
    1671. 

  1671. 

  1672.       <tr>
    1673. 

  1673.         <td>Full-width Text</td>
    1674. 

  1674. 

  1675.         <td>Full Width</td>
    1676. 

  1676. 

  1677.         <td>fwid</td>
    1678. 

  1678.       </tr>
    1679. 

  1679. 

  1680.       <tr>
    1681. 

  1681.         <td>Traditional Characters</td>
    1682. 

  1682. 

  1683.         <td>Traditional</td>
    1684. 

  1684. 

  1685.         <td>trad</td>
    1686. 

  1686.       </tr>
    1687. 

  1687. 

  1688.       <tr>
    1689. 

  1689.         <td>Simplified Characters</td>
    1690. 

  1690. 

  1691.         <td>Simplified</td>
    1692. 

  1692. 

  1693.         <td>smpl</td>
    1694. 

  1694.       </tr>
    1695. 

  1695. 

  1696.       <tr>
    1697. 

  1697.         <td>JIS 1978 Characters</td>
    1698. 

  1698. 

  1699.         <td>JIS 1978 Characters</td>
    1700. 

  1700. 

  1701.         <td>jp78</td>
    1702. 

  1702.       </tr>
    1703. 

  1703. 

  1704.       <tr>
    1705. 

  1705.         <td>JIS 1983 Characters</td>
    1706. 

  1706. 

  1707.         <td>JIS 1983 Characters</td>
    1708. 

  1708. 

  1709.         <td>jp83</td>
    1710. 

  1710.       </tr>
    1711. 

  1711. 

  1712.       <tr>
    1713. 

  1713.         <td>JIS 1990 Characters</td>
    1714. 

  1714. 

  1715.         <td>JIS 1990 Characters</td>
    1716. 

  1716. 

  1717.         <td>jp90</td>
    1718. 

  1718.       </tr>
    1719. 

  1719.     </table>
    1720. 

  1720. 

  1721.     <p>FontForge will retain the order of features in the morx
    1722. 

  1722.     table, and the user may adjust it with the <a href=
    1723. 

  1723.     "fontinfo.html#Lookups">Element-&gt;Font Info</a> command.
    1724. 

  1724.     (this is the same list as that used for GSUB table. GSUB
    1725. 

  1725.     features that don't correspond to mac features will be
    1726. 

  1726.     ignored).</p>
    1727. 

  1727. 

  1728.     <h2>What is <a name="Unsupported" id=
    1729. 

  1729.     "Unsupported">Unsupported</a>?</h2>
    1730. 

  1730. 

  1731.     <p>FontForge does not (yet) support all the advanced
    1732. 

  1732.     typographic features available in either opentype or apple
    1733. 

  1733.     advanced typography.</p>
    1734. 

  1734. 

  1735.     <h3>OpenType</h3>
    1736. 

  1736. 

  1737.     <ul>
    1738. 

  1738.       <li>FontForge does not provide an attachment list subtable
    1739. 

  1739.       nor a MarkAttachClassDef subtable of the GDEF table.</li>
    1740. 

  1740. 

  1741.       <li>FontForge does not support the VORG and JUST tables</li>
    1742. 

  1742.     </ul>
    1743. 

  1743. 

  1744.     <p><a href="TrueOpenTables.html">See here for a complete list
    1745. 

  1745.     of supported tables</a>.</p>
    1746. 

  1746. 

  1747.     <h3>Apple Advanced Typography</h3>
    1748. 

  1748. 

  1749.     <ul>
    1750. 

  1750.       <li>FontForge will never generate a 'mort' table. It can read
    1751. 

  1751.       'mort' tables, but it will only produce 'morx' tables.</li>
    1752. 

  1752. 

  1753.       <li>FontForge is unable to parse contextual ligatures. It can
    1754. 

  1754.       find ligatures which start from state 0, but it will not find
    1755. 

  1755.       ligatures which start from other states (that is, which are
    1756. 

  1756.       contextual)</li>
    1757. 

  1757. 

  1758.       <li>FontForge does not support the following Apple specific
    1759. 

  1759.       tables
    1760. 

  1760. 

  1761.         <ul>
    1762. 

  1762.           <li>acnt (accent attachment)</li>
    1763. 

  1763. 

  1764.           <li>bsln (baseline)</li…
    1765.

Large files files are truncated, but you can click here to view the full file