1 /* Copyright (C) 2003, 2004
2 National Institute of Advanced Industrial Science and Technology (AIST)
3 Registration Number H15PRO112
4 See the end for copying conditions. */
8 @page mdbFLT Font Layout Table
10 @section flt-description DESCRIPTION
12 For simple scripts, the rendering engine converts character codes into glyph
13 codes one by one by consulting the encoding of each selected font.
14 But, to render text that requires complicated layout (e.g. Thai and
15 Indic scripts), one to one conversion is not sufficient. A sequence
16 of characters may have to be drawn as a single ligature. Some
17 glyphs may have to be drawn at 2-dimensionally shifted positions.
19 To handle those complicated scripts, the m17n library uses Font Layout
20 Tables (FLTs for short). The FLT driver interprets an FLT and
21 converts a character sequence into a glyph sequence that is ready to
22 be passed to the rendering engine.
24 An FLT can contain information to extract a grapheme cluster from a
25 character sequence and to reorder the characters in the cluster, in
26 addition to information found in OpenType Layout Tables (CMAP, GSUB,
29 An FLT is a cascade of one or more conversion stages. In each stage, a
30 sequence is converted into another sequence to be read in the
31 next stage. The length of sequences may differ from stage to
32 stage. Each element in a sequence has the following integer attributes.
37 In the first conversion stage, this is the character code in the
38 original character sequence. In the last stage, it is the glyph code
39 passed to the rendering engine. In other cases, it is an intermediate
44 This is the category code defined in the @c CATEGORY-TABLE
49 If nonzero, it specifies how to combine this (intermediate) glyph
50 with the previous one.
52 <li> left-padding-flag
54 If nonzero, it instructs the rendering function to insert a padding
55 space before this (intermediate) glyph so that the glyph does not
56 overlap with the previous one.
58 <li> right-padding-flag
60 If nonzero, it instructs the rendering function to insert a padding
61 space after this (intermediate) glyph so that the glyph does not
62 overlap with the next one.
66 When the layout engine draws text, it at first determines a font and
67 an FLT for each character in the text. For each subsequence of
68 characters that use use the same font and FLT, the layout engine
69 generates an intermediate glyph sequence from the character
70 subsequence. Each element in the intermediate glyph sequence
71 has the corresponding character code as the code attribute and zeroes
72 for other attributes. This sequence is processed in the
73 first stage of FLT as the current @e run (substring).
75 Each stage works as follows.
77 At first, if the stage has a @c CATEGORY-TABLE, the category of each glyph
78 in the current run is updated. If there is a glyph that has no category,
79 the current run ends before that glyph.
81 Then, the default values of code-offset, combining-spec, and left-padding-flag
82 of this stage are initialized to zero.
84 Next, the initial conversion rule of the stage is applied to the
87 Lastly, the current run is replaced with the newly produced
88 (intermediate) glyph sequence.
90 @section flt-syntax SYNTAX and SEMANTICS
92 The m17n library loads an FLT from the m17n database using the tag
93 \<font, layouter, FLT-NAME\>. The date format of an FLT is as follows:
96 FONT-LAYOUT-TABLE ::= STAGE0 STAGE *
98 STAGE0 ::= CATEGORY-TABLE GENERATOR
100 STAGE ::= CATEGORY-TABLE ? GENERATOR
102 CATEGORY-TABLE ::= '(' 'category' CATEGORY-SPEC + ')'
104 CATEGORY-SPEC ::= '(' CODE CATEGORY ')'
105 | '(' CODE CODE CATEGORY ')'
112 In the definition of @c CATEGORY-SPEC, @c CODE is a glyph code, and @c
113 CATEGORY is ASCII code of an upper or lower letter, i.e. one of 'A',
114 ... 'Z', 'a', .. 'z'.
116 The first form of @c CATEGORY-SPEC assigns @c CATEGORY to a glyph
117 whose code @c CODE. The second form assigns @c CATEGORY to glyphs
118 whose code falls between the two @c CODEs.
121 GENERATOR ::= '(' 'generator' RULE MACRO-DEF * ')'
123 RULE ::= REGEXP-BLOCK | MATCH-BLOCK | SUBST-BLOCK | COND-BLOCK
124 | DIRECT-CODE | COMBINING-SPEC | OTF-SPEC
125 | PREDEFINED-RULE | MACRO-NAME
127 MACOR-DEF ::= '(' MACRO-NAME RULE + ')'
130 Each @c RULE specifies glyphs to be consumed and glyphs to be produced. When
131 some glyphs are consumed, they are taken away from the current run. A
132 rule may fail in some condition. If not described explicitly to
133 fail, it should be regarded that the rule succeeds.
136 DIRECT-CODE ::= INTEGER
139 This rule consumes no glyph and produces a glyph which has the
140 following attributes:
143 <li> code : @c INTEGER plus the default code-offset
144 <li> combining-spec : default value
145 <li> left-padding-flag : default value
146 <li> right-padding-flag : zero
149 After having produced the glyph, the default code-offset,
150 combining-spec, and left-padding-flag are all reset to zero.
153 PREDEFINED-RULE ::= '=' | '*' | '<' | '>' | '|' | '[' | ']'
156 They perform actions as follows.
161 This rule consumes the first glyph in the current run and produces the
162 same glyph. It fails if the current run is empty.
166 This rule repeatedly executes the previous rule.
167 If the previous rule fails, this rule does nothing and fails.
171 This rule specifies the start of a grapheme cluster.
175 This rule specifies the end of a grapheme cluster.
179 This rule sets the default left-padding-flag to 1.
180 No glyph is consumed. No glyph is produced.
184 This rule changes the right-padding-flag of the lastly generated
186 No glyph is consumed. No glyph is produced.
190 This rule consumes no glyph and produces a special glyph whose
191 category is ' ' and other attributes are zero.
192 This is the only rule that produces that special glyph.
197 REGEXP-BLOCK ::= '(' REGEXP RULE * ')'
202 @c MTEXT is a regular expression that should match the sequence of
203 categories of the current run. If a match is found, this rule
204 executes @c RULEs temporarily limiting the current run to the matched
205 part. The matched part is consumed by this rule.
207 Parenthesized subexpressions, if any, are recorded to be used in @c
208 MATCH-BLOCK that may appear in one of @c RULEs.
210 If no match is found, this rule fails.
213 MATCH-BLOCK ::= '(' MATCH-INDEX RULE * ')'
215 MATCH-INDEX ::= INTEGER
218 MATCH-INDEX is an integer specifying a parenthesized subexpression
219 recoded by the previous @c REGEXP-BLOCK. If such a subexpression was
220 found by the previous regular expression matching, this rule executes @c
221 RULEs temporarily limiting the current run to the matched part
222 of the subexpression. The matched part is consumed by this rule.
224 If no match was found, this rule fails.
226 If this is the first rule of the stage, @c MATCH-INDEX must be 0, and
227 it matches the whole current run.
230 SUBST-BLOCK ::= '(' SOURCE-PATTERN RULE * ')'
232 SOURCE-PATTERN ::= '(' CODE + ')'
233 | (' 'range' CODE CODE ')'
236 If the sequence of codes of the current run matches @c SOURCE-PATTERN,
237 this rule executes @c RULEs temporarily limiting the current run to
238 the matched part. The matched part is consumed.
240 The first form of @c SOURCE-PATTERN specifies a sequence of glyph codes to be
241 matched. In this case, this rule resets the default code-offset to
244 The second form specifies a range of codes that should match the first
245 glyph code of the code sequence. In this case, this rule sets the
246 default code-offset to the first glyph code minus the first @c CODE in @c
249 If no match is found, this rule fails.
252 COND-BLOCK ::= '(' 'cond' RULE + ')'
255 This rule sequentially executes @c RULEs until one succeeds. If no
256 rule succeeds, this rule fails. Otherwise, it succeeds.
262 @c OTF-SPEC is a symbol whose name specifies an instruction to the OTF
263 driver. The name has the following syntax.
266 OTF-SPEC-NAME ::= 'otf:' SCRIPT LANGSYS ? GSUB-FEATURES ? GPOS-FEATURES ?
270 LANGSYS ::= '/' SYMBOL
272 GSUB-FEATURES ::= '=' FEATURE-LIST ?
274 GPOS-FEATURES ::= '+' FEATURE-LIST ?
276 FEATURE-LIST ::= ( SYMBOL ',' ) * [ SYMBOL | '*' ]
280 Each @c SYMBOL specifies a tag name defined in the OpenType
283 For @c SCRIPT, @c SYMBOL specifies a Script tag name (e.g. deva for
286 For @c LANGSYS, @c SYMBOL specifies a Language System tag name. If @c
287 LANGSYS is omitted, the Default Language System
290 For @c GSUB-FEATURES, each @c SYMBOL specifies a GSUB Feature tag name
291 to apply. '*' is allowed as the last item to specify all remaining
292 features. If @c SYMBOL is preceded by '~' and the last item is '*',
293 @c SYMBOL is excluded from the features to apply. If no @c SYMBOL is
294 specified, no GSUB feature is applied. If @c GSUB-FEATURES itself is
295 omitted, all GSUB features are applied.
297 The specification of @c GPOS-FEATURES is analogous to that of @c
300 See the following page for the OpenType specification.\n
301 <http://www.microsoft.com/typography/otspec/default.htm>
307 @c COMBINING is a symbol whose name specifies how
308 to combine the next glyph with the previous one. This rule sets the
309 default combining-spec to an integer code that is unique to the symbol
310 name. The name has the following syntax.
313 COMBINING-NAME ::= VPOS HPOS OFFSET VPOS HPOS
315 VPOS ::= 't' | 'c' | 'b' | 'B'
317 HPOS ::= 'l' | 'c' | 'r'
319 OFFSET :: = '.' | XOFF | YOFF XOFF ?
321 XOFF ::= ('<' | '>') INTEGER ?
323 YOFF ::= ('+' | '-') INTEGER ?
326 @c VPOS and @c HPOS specify the vertical and horizontal positions
332 0----1----2 <---- top 0 t l
336 9 10 11 <---- center 4 B c
338 --3----4----5-- <-- baseline 6 b l
340 6----7----8 <---- bottom 8 b r
343 left center right 11 c r
346 The left figure shows 12 reference points of a glyph by numbers 0 to
347 11. The rectangle 0-6-8-2 is the bounding box of the glyph, the
348 positions 3, 4, and 5 are on the baseline, 9 and 11 are on the center
349 of the lines 0-6 and 2-8 respectively, 1, 10, 4, and 7 are on the
350 center of the lines 1-2, 3-5, 9-11, and 6-8 respectively.
352 The right table shows how those reference points are specified by a
353 pair of @c VPOS and @c HPOS.
355 The first @c VPOS and @c HPOS in the definition of @c COMBINING-NAME
357 reference point of the previous glyph, and the second @c VPOS and @c
358 HPOS specify that of the next glyph.
359 The next glyph is drawn so that these two reference points align.
361 @c OFFSET specifies the way of alignment in detail. If it is '.', the
362 reference points are on the same position.
364 @c XOFF specifies how much the X position of the next reference point
365 should be shifted to the right ('<') or left ('>') from the previous
368 @c YOFF specifies how much the Y position of the next reference point
369 should be shifted upward ('+') or downward ('-') from the previous reference
372 In both cases, @c INTEGER is the amount of shift expressed as a
373 percentage of the font size, i.e., if @c INTEGER is 10, it means
374 10% (1/10) of the font size. If @c INTEGER is omitted, it is assumed that
377 Once the next glyph is combined with the previous one, they
378 are treated as a single combined glyph.
381 MACRO-NAME ::= SYMBOL
384 @c MACRO-NAME is a symbol that appears in one of @c MACRO-DEF. It is
385 exapanded to the sequence of the correponding @c RULSs.
387 @section flt-context-dependent CONTEXT DEPENDENT BEHAVIOR
389 So far, it has been assumed that each sequence, which is drawn with a
390 specific font, is context free, i.e. not affected by the glyphs
391 preceding or following that sequence. This is true when sequence S1
392 is drawn with font F1 while the preceding sequence S0 unconditionally
397 currently used font F0 F1
401 Sometimes, however, a clear separation of sequences is not possible.
402 Suppose that the preceding sequence S0 can be drawn not only with F0
407 currently used font F0 F1
408 usable font(s) F0,F1 F1
411 In this case, glyphs used to draw the preceding S0 may affect glyph
412 generation of S1. Therefore it is necessary to access information
413 about S0, which has already been processed, when processing S1.
414 Generation rules in the first stage (only in the first stage) accept a
415 special regular expression to access already processed parts.
421 @c RE0 and @c RE1 are regular expressions that match the preceding
422 sequence S0 and the following sequence S1, respectively.
424 Pay attention to the space between the two regular expressions. It
425 represents the special category ' ' (see above). Note that the
426 regular expression above belongs to glyph generation rules using font
427 F1, therefore not only RE1 but also RE0 must be expressed with the
428 categories for F1. This means when the preceding sequence S0 cannot
429 be expressed with the categories for F1 (as in the first example
430 above) generation rules having these patterns never match.
432 @section flt-seealso SEE ALSO
434 @ref mdbGeneral "mdbGeneral(5)",
435 @ref flt-list "FLTs provided by the m17n database"
439 Copyright (C) 2003, 2004
440 National Institute of Advanced Industrial Science and Technology (AIST)
441 Registration Number H15PRO112
443 This file is part of the m17n database; a sub-part of the m17n
446 The m17n library is free software; you can redistribute it and/or
447 modify it under the terms of the GNU Lesser General Public License
448 as published by the Free Software Foundation; either version 2.1 of
449 the License, or (at your option) any later version.
451 The m17n library is distributed in the hope that it will be useful,
452 but WITHOUT ANY WARRANTY; without even the implied warranty of
453 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
454 Lesser General Public License for more details.
456 You should have received a copy of the GNU Lesser General Public
457 License along with the m17n library; if not, write to the Free
458 Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA