Debian Codegraph @

upstream/debian-trixie/man7/groff_char.7

raw
1 '\" t
2 .TH groff_char 7 "3 June 2025" "groff 1.23.0"
3 .SH Name
4 groff_char \- GNU
5 .I roff
6 special character and glyph repertoire
7 .
8 .
9 .\" ====================================================================
10 .\" Legal Terms
11 .\" ====================================================================
12 .\"
13 .\" Copyright (C) 1989-2023 Free Software Foundation, Inc.
14 .\"
15 .\" This file is part of groff (GNU roff), which is a free software
16 .\" project.
17 .\"
18 .\" You can redistribute it and/or modify it under the terms of the GNU
19 .\" General Public License as published by the Free Software Foundation,
20 .\" either version 2 of the License, or (at your option) any later
21 .\" version.
22 .\"
23 .\" You should have received a copy of the GNU General Public License
24 .\" along with this program.
25 .\"
26 .\" If not, see <http://www.gnu.org/licenses/gpl-2.0.html>.
27 .
28 .
29 .\" Save and disable compatibility mode (for, e.g., Solaris 10/11).
30 .do nr *groff_groff_char_7_man_C \n[.cp]
31 .cp 0
32 .
33 .\" Define fallback for groff 1.23's MR macro if the system lacks it.
34 .nr do-fallback 0
35 .if !\n(.f .nr do-fallback 1 \" mandoc
36 .if \n(.g .if !d MR .nr do-fallback 1 \" older groff
37 .if !\n(.g .nr do-fallback 1 \" non-groff *roff
38 .if \n[do-fallback] \{\
39 . de MR
40 . ie \\n(.$=1 \
41 . I \%\\$1
42 . el \
43 . IR \%\\$1 (\\$2)\\$3
44 . .
45 .\}
46 .rr do-fallback
47 .
48 .
49 .\" ====================================================================
50 .SH Description
51 .\" ====================================================================
52 .
53 The GNU
54 .I roff
55 typesetting system has a large glyph repertoire suitable for production
56 of varied literary,
57 professional,
58 technical,
59 and mathematical documents.
60 .
61 .I groff
62 works with
63 .I characters;
64 an output device renders
65 .I glyphs.
66 .
67 .IR groff 's
68 input character set is restricted to that defined by the standards
69 ISO Latin-1
70 (ISO 8859-1)
71 and CCSID \[lq]code page\[rq] 1047
72 (an EBCDIC arrangement of Latin-1).
73 .
74 For ease of document maintenance in UTF-8 environments,
75 it is advisable to use only the Unicode basic Latin code points,
76 a subset of all of the foregoing historically referred to as \%US-ASCII,
77 .\" Yes, a subset, albeit a permutation as well in the cp1047 case.
78 which has only 94 visible,
79 printable code points.
80 .\" In groff, 0x20 SP is mapped to a space node, not a glyph node, and
81 .\" all kinds of special behavior attaches to such nodes, so we count
82 .\" only to 94 and not 95 as is often done in other ASCII contexts.
83 In
84 .I groff,
85 these are termed
86 .I "ordinary characters."
87 .
88 Often,
89 many more are desired in output.
90 .
91 .
92 .P
93 AT&T
94 .I troff
95 in the 1970s faced a similar problem:
96 the available typesetter's glyph repertoire differed from that of the
97 computers that controlled it.
98 .
99 .IR troff 's \" AT&T
100 solution was a form of escape sequence known as a
101 .I special character
102 to access several dozen additional glyphs available in the fonts
103 prepared for mounting in the phototypesetter.
104 .
105 These glyphs were mapped onto a two-character name space for a degree
106 of mnemonic convenience;
107 for example,
108 the escape sequence
109 .B \e(aa
110 encoded an acute accent and
111 .B \e(sc
112 a section sign.
113 .
114 .
115 .P
116 .I groff
117 has lifted historical
118 .I roff
119 limitations on special character name lengths,
120 but recognizes and retains compatibility with the historical names.
121 .
122 .I groff
123 expands the lexicon of glyphs available by name and permits users to
124 define their own special character escape sequences with the
125 .B char
126 request.
127 .
128 Special character names are
129 .I groff
130 identifiers;
131 see section \[lq]Identifiers\[rq] in
132 .MR groff 7 .
133 .
134 Our discussion uses the terms \[lq]glyph name\[rq] and \[lq]special
135 character name\[rq] interchangeably;
136 we assume no character translations or redefinitions.
137 .
138 .
139 .P
140 This document lists all of the glyph names predefined by
141 .IR groff 's
142 font description files and presents the systematic notation by which it
143 enables access to arbitrary Unicode code points and construction of
144 composite glyphs.
145 .
146 Glyphs listed may be unavailable,
147 or may vary in appearance,
148 depending on the output device and font chosen when the page was
149 formatted.
150 .
151 This page was rendered for device
152 .B \*[.T]
153 using font
154 .BR \n[.fn] .
155 .
156 .
157 .P
158 A few escape sequences that are not
159 .I groff
160 special characters also produce glyphs;
161 these exist for syntactical or historical reasons.
162 .
163 .BR \e\[aq] ,
164 .BR \e\[ga] ,
165 .BR \e\- ,
166 and
167 .B \e_
168 are translated on input to the special character escape sequences
169 .BR \e[aa] ,
170 .BR \e[ga] ,
171 .BR \e[\-] ,
172 and
173 .BR \e[ul] ,
174 respectively.
175 .
176 Others include
177 .BR \e\e ,
178 .B \e.\&
179 (backslash-dot),
180 and
181 .BR \ee ;
182 see
183 .MR groff 7 .
184 .
185 A small number of special characters represent glyphs that are not
186 encoded in Unicode;
187 examples include the baseline rule
188 .B \e[ru]
189 and the Bell System logo
190 .B \e[bs].
191 .
192 .
193 .P
194 In
195 .IR groff ,
196 you can test output device support for any character
197 (ordinary or special)
198 with the conditional expression operator
199 .RB \[lq] c \[rq].
200 .
201 .RS
202 .\" https://www.bell-labs.com/usr/dmr/www/ ("In 1984, ...")
203 .EX
204 \&.ie c \e[bs] \e{Welcome to the \e[bs] Bell System;
205 did you get the Wehrmacht helmet or the Death Star?\e}
206 \&.el No Bell System logo.
207 .EE
208 .RE
209 .
210 .
211 .P
212 For brevity in the remainder of this document,
213 we shall refer to systems conforming to the
214 ISO 646:1991 IRV,
215 ISO 8859,
216 or
217 ISO 10646 (\[lq]Unicode\[rq])
218 character encoding standards as \[lq]ISO\[rq] systems,
219 and those employing IBM code page 1047 as \[lq]EBCDIC\[rq] systems.
220 .
221 That said,
222 EBCDIC systems that support
223 .I groff
224 are known to also support UTF-8.
225 .
226 .
227 .P
228 While
229 .I groff
230 accepts eight-bit encoded input,
231 not all such code points are valid as input.
232 .
233 .\" src/libs/libgroff/invalid.cpp
234 On ISO platforms,
235 character codes
236 0,
237 11,
238 13\[en]31,
239 and
240 128\[en]159
241 are invalid.
242 .
243 (This is all C0 and C1 controls except for
244 SOH through LF
245 [Control+A to Control+J],
246 and FF
247 [Control+L].)
248 .
249 On EBCDIC platforms,
250 0,
251 8\[en]9,
252 11,
253 13\[en]20,
254 23\[en]31,
255 and
256 48\[en]63
257 are invalid.
258 .
259 Some of these code points are used by
260 .I groff
261 for internal purposes,
262 which is one reason it does not support UTF-8 natively.
263 .
264 .
265 .\" ====================================================================
266 .SS "Fundamental character set"
267 .\" ====================================================================
268 .
269 The ordinary characters catalogued above,
270 plus the space,
271 tab,
272 newline,
273 and leader (Control+A),
274 form the fundamental character
275 set for
276 .I groff
277 input;
278 anything in the language,
279 even over one million code points in Unicode,
280 can be expressed using it.
281 .
282 On ISO systems,
283 code points in the range 33\[en]126 comprise a common set of
284 printable glyphs in all of the aforementioned ISO character encoding
285 standards.
286 .
287 It is this character set and
288 (with some noteworthy exceptions)
289 the corresponding glyph repertoire for which AT&T
290 .I troff
291 was implemented.
292 .
293 On EBCDIC systems,
294 printable characters are in the range 66\[en]201 and 203\[en]254;
295 those without counterparts in the ISO range 33\[en]126 are discussed
296 in the next subsection.
297 .\" From this point, do not talk about numerical character assignments.
298 .
299 .
300 .P
301 All of the following characters map to glyphs as you would expect.
302 .
303 .TS
304 center box;
305 Lf(CR).
306 ! # $ % & ( ) * + , . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
307 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ ] _
308 a b c d e f g h i j k l m n o p q r s t u v w x y z { | }
309 .TE
310 .\" The bottom border of that box is practically kissin' the tittles.
311 .if t .sp 0.2v
312 .
313 The remaining ordinary characters surprise computing professionals and
314 others intimately familiar with the ISO character encodings.
315 .
316 The developers of AT&T
317 .I troff
318 chose mappings for them that would be useful for typesetting technical
319 literature in a broad range of scientific disciplines:
320 Bell Labs used the system for preparation of AT&T's patent filings with
321 the U.S.\& government.
322 .
323 Further,
324 the prevailing character encoding standard in the 1970s,
325 USAS X3.4-1968 (\[lq]ASCII\[rq]),
326 deliberately supported semantic ambiguity at some code points,
327 and outright substitution at several others,
328 to suit the localization demands of various national standards bodies.
329 .
330 .
331 .P
332 The table below presents the seven exceptional code points
333 with their typical keycap engravings,
334 their glyph mappings and semantics in
335 .I roff
336 systems,
337 and the escape sequences producing the Unicode basic Latin character
338 they replace.
339 .
340 The first,
341 the neutral double quote,
342 is a partial exception because it does represent itself,
343 but since the
344 .I roff
345 language also uses it to quote macro arguments,
346 .I groff
347 supports a special character escape sequence as an alternative form so
348 that the glyph can be easily included in macro arguments without
349 requiring the user to master the quoting rules that AT&T
350 .I troff
351 required in that context.
352 .
353 (Some requests,
354 like
355 .BR ds ,
356 also treat
357 .B \[dq]
358 non-literally.)
359 .
360 Furthermore,
361 not all of the special character escape sequences are portable to AT&T
362 .I troff
363 and all of its descendants;
364 these
365 .I groff
366 extensions are presented using its special character form
367 .BR \[rs][] ,
368 whereas portable special character escape sequences are shown in the
369 traditional
370 .B \[rs](
371 form.
372 .
373 .B \[rs]\-
374 and
375 .B \[rs]e
376 are portable to all known
377 .IR troff s.
378 .
379 .B \[rs]e
380 means \[lq]the glyph of the current escape character\[rq];
381 it therefore can produce unexpected output if the
382 .B ec
383 request is used.
384 .
385 On devices with a limited glyph repertoire,
386 glyphs in the \[lq]keycap\[rq] and \[lq]appearance\[rq] columns on the
387 same row of the table may look identical;
388 except for the neutral double quote,
389 this will
390 .I not
391 be the case on more-capable devices.
392 .
393 Review your document using as many different output devices as possible.
394 .
395 .
396 .P
397 .TS
398 center box;
399 L L L.
400 Keycap Appearance and meaning Special character and meaning
401 _
402 " " neutral double quote \f[CR]\[rs][dq]\f[] neutral double quote
403 \[aq] \[cq] closing single quote \f[CR]\[rs][aq]\f[] neutral apostrophe
404 \- - hyphen \f[CR]\[rs]\-\f[] or \f[CR]\[rs][\-]\f[] minus sign/Unix dash
405 \[rs] (escape character) \f[CR]\[rs]e\f[] or \f[CR]\[rs][rs]\f[] reverse solidus
406 \[ha] \[u02C6] modifier circumflex \f[CR]\[rs](ha\f[] circumflex/caret/\[lq]hat\[rq]
407 \[ga] \[oq] opening single quote \f[CR]\[rs](ga\f[] grave accent
408 \[ti] \[u02DC] modifier tilde \f[CR]\[rs](ti\f[] tilde
409 .TE
410 .
411 .
412 .P
413 The hyphen-minus is a particularly unfortunate case of overloading.
414 .
415 Its awkward name in ISO 8859 and later standards reflects the many
416 distinguishable purposes to which it had already been put by the 1980s,
417 including
418 a hyphen,
419 a minus sign,
420 and
421 (alone or in repetition)
422 dashes of varying widths.
423 .
424 For best results in
425 .I roff
426 systems,
427 use the
428 .RB \[lq] \- \[rq]
429 character in input outside an escape sequence
430 .I only
431 to mean a hyphen,
432 as in the phrase \[lq]long-term\[rq].
433 .
434 For a minus sign in running text or a Unix command-line option dash,
435 use
436 .B \[rs]\-
437 (or
438 .B \[rs][\-]
439 in
440 .I groff
441 if you find it helps the clarity of the source document).
442 .
443 (Another minus sign,
444 for use in mathematical equations,
445 is available as
446 .BR \[rs][mi] ).
447 .
448 AT&T
449 .I troff
450 supported em-dashes as
451 .BR \[rs](em ,
452 as does
453 .IR groff .
454 .
455 .
456 .P
457 The special character escape sequence for the apostrophe as a neutral
458 single quote is typically needed only in technical content;
459 typing words like \[lq]can't\[rq] and \[lq]Anne's\[rq] in a natural way
460 will render correctly,
461 because in ordinary prose an apostrophe is typeset either as a closing
462 single quotation mark or as a neutral single quote,
463 depending on the capabilities of the output device.
464 .
465 By contrast,
466 special character escape sequences should be used for quotation marks
467 unless portability to limited or historical
468 .I troff
469 implementations is necessary;
470 on those systems,
471 the input convention is to pair the grave accent with the apostrophe for
472 single quotes,
473 and to double both characters for double quotes.
474 .
475 AT&T
476 .I troff
477 defined no special characters for quotation marks or the apostrophe.
478 .
479 Repeated single quotes
480 (\[oq]\[oq]thus\[cq]\[cq])
481 will be visually distinguishable from double quotes
482 (\[lq]thus\[rq])
483 on terminal devices,
484 and perhaps on others
485 (depending on the font selected).
486 .
487 .TS
488 tab(@) center box;
489 L L.
490 AT&T \f[I]troff\f[] input@recommended \f[I]groff\f[] input
491 _
492 .T&
493 Lf(CR) Lf(CR).
494 A Winter\[aq]s Tale@A Winter\[aq]s Tale
495 \[ga]U.K.\& outer quotes\[aq]@\[rs][oq]U.K.\& outer quotes\[rs][cq]
496 \[ga]U.K.\& \[ga]\[ga]inner\[aq]\[aq] quotes\[aq]\
497 @\[rs][oq]U.K.\& \[rs][lq]inner\[rs][rq] quotes\[rs][cq]
498 \[ga]\[ga]U.S.\& outer quotes\[aq]\[aq]\
499 @\[rs][lq]U.S.\& outer quotes\[rs][rq]
500 \[ga]\[ga]U.S.\& \[ga]inner\[aq] quotes\[aq]\[aq]\
501 @\[rs][lq]U.S.\& \[rs][oq]inner\[rs][cq] quotes\[rs][rq]
502 .TE
503 .\" Keep bottom border of box from sitting on the ascenders below.
504 .if t .sp 0.2v
505 .
506 If you frequently require quotation marks in your document,
507 see if the macro package you're using supplies strings or macros to
508 facilitate quotation,
509 or define them yourself
510 (except in man pages).
511 .
512 .
513 .P
514 Using Unicode basic Latin characters to compose boxes and lines is
515 ill-advised.
516 .
517 .I roff
518 systems have special characters for drawing horizontal and vertical
519 lines;
520 see subsection \[lq]Rules and lines\[rq] below.
521 .
522 Preprocessors like
523 .MR \%tbl 1
524 and
525 .MR \%pic 1
526 draw boxes and will produce the best possible output for the device,
527 falling back to basic Latin glyphs only when necessary.
528 .
529 .
530 .\" ====================================================================
531 .SS "Eight-bit encodings and Latin-1 supplement"
532 .\" ====================================================================
533 .
534 ISO 646 is a seven-bit code encoding 128 code points;
535 eight-bit codes are twice the size.
536 .
537 ISO 8859-1 and code page 1047 allocated the additional space to what
538 Unicode calls \[lq]C1 controls\[rq]
539 (control characters)
540 and the \[lq]Latin-1 supplement\[rq].
541 .
542 The C1 controls are neither printable nor usable as
543 .I groff
544 input.
545 .
546 .
547 .P
548 Two Latin-1 supplement characters are handled specially on input.
549 .
550 .I \%troff
551 never produces them as output.
552 .
553 .
554 .TP
555 NBSP
556 encodes a no-break space;
557 it is mapped to
558 .BR \[rs]\[ti] ,
559 the adjustable non-breaking space escape sequence.
560 .
561 .
562 .TP
563 SHY
564 encodes a soft hyphen;
565 it is mapped to
566 .BR \[rs]% ,
567 the hyphenation control escape sequence.
568 .
569 .
570 .P
571 The remaining characters in the Latin-1 supplement represent
572 themselves.
573 .
574 Although they can be specified directly with the keyboard on systems
575 configured to use Latin-1 as the character encoding,
576 it is more portable,
577 both to other
578 .I roff
579 systems and to UTF-8 environments,
580 to use their special character escape sequences,
581 shown below.
582 .
583 The glyph descriptions we use are non-standard in some cases,
584 for brevity.
585 .
586 .
587 .P
588 .TS
589 L2 Lf(CR)1 L L2 Lf(CR)1 L.
590 \[r!] \e[r!] inverted exclamation mark \[~N] \e[\[ti]N] N tilde
591 \[ct] \e[ct] cent sign \[`O] \e[\[ga]O] O grave
592 \[Po] \e[Po] pound sign \['O] \e[\[aq]O] O acute
593 \[Cs] \e[Cs] currency sign \[^O] \e[\[ha]O] O circumflex
594 \[Ye] \e[Ye] yen sign \[~O] \e[\[ti]O] O tilde
595 \[bb] \e[bb] broken bar \[:O] \e[:O] O dieresis
596 \[sc] \e[sc] section sign \[mu] \e[mu] multiplication sign
597 \[ad] \e[ad] dieresis accent \[/O] \e[/O] O slash
598 \[co] \e[co] copyright sign \[`U] \e[\[ga]U] U grave
599 \[Of] \e[Of] feminine ordinal indicator \['U] \e[\[aq]U] U acute
600 \[Fo] \e[Fo] left double chevron \[^U] \e[\[ha]U] U circumflex
601 \[no] \e[no] logical not \[:U] \e[:U] U dieresis
602 \[rg] \e[rg] registered sign \['Y] \e[\[aq]Y] Y acute
603 \[a-] \e[a\-] macron accent \[TP] \e[TP] uppercase thorn
604 \[de] \e[de] degree sign \[ss] \e[ss] lowercase sharp s
605 \[+-] \e[+\-] plus-minus \[`a] \e[\[ga]a] a grave
606 \[S2] \e[S2] superscript two \['a] \e[\[aq]a] a acute
607 \[S3] \e[S3] superscript three \[^a] \e[\[ha]a] a circumflex
608 \[aa] \e[aa] acute accent \[~a] \e[\[ti]a] a tilde
609 \[mc] \e[mc] micro sign \[:a] \e[:a] a dieresis
610 \[ps] \e[ps] pilcrow sign \[oa] \e[oa] a ring
611 \[pc] \e[pc] centered period \[ae] \e[ae] ae ligature
612 \[ac] \e[ac] cedilla accent \[,c] \e[,c] c cedilla
613 \[S1] \e[S1] superscript one \[`e] \e[\[ga]e] e grave
614 \[Om] \e[Om] masculine ordinal indicator \['e] \e[\[aq]e] e acute
615 \[Fc] \e[Fc] right double chevron \[^e] \e[\[ha]e] e circumflex
616 \[14] \e[14] one quarter symbol \[:e] \e[:e] e dieresis
617 \[12] \e[12] one half symbol \[`i] \e[\[ga]i] i grave
618 \[34] \e[34] three quarters symbol \['i] \e[\[aq]i] e acute
619 \[r?] \e[r?] inverted question mark \[^i] \e[\[ha]i] i circumflex
620 \[`A] \e[\[ga]A] A grave \[:i] \e[:i] i dieresis
621 \['A] \e[\[aq]A] A acute \[Sd] \e[Sd] lowercase eth
622 \[^A] \e[\[ha]A] A circumflex \[~n] \e[\[ti]n] n tilde
623 \[~A] \e[\[ti]A] A tilde \[`o] \e[\[ga]o] o grave
624 \[:A] \e[:A] A dieresis \['o] \e[\[aq]o] o acute
625 \[oA] \e[oA] A ring \[^o] \e[\[ha]o] o circumflex
626 \[AE] \e[AE] AE ligature \[~o] \e[\[ti]o] o tilde
627 \[,C] \e[,C] C cedilla \[:o] \e[:o] o dieresis
628 \[`E] \e[\[ga]E] E grave \[di] \e[di] division sign
629 \['E] \e[\[aq]E] E acute \[/o] \e[/o] o slash
630 \[^E] \e[\[ha]E] E circumflex \[`u] \e[\[ga]u] u grave
631 \[:E] \e[:E] E dieresis \['u] \e[\[aq]u] u acute
632 \[`I] \e[\[ga]I] I grave \[^u] \e[\[ha]u] u circumflex
633 \['I] \e[\[aq]I] I acute \[:u] \e[:u] u dieresis
634 \[^I] \e[\[ha]I] I circumflex \['y] \e[\[aq]y] y acute
635 \[:I] \e[:I] I dieresis \[Tp] \e[Tp] lowercase thorn
636 \[-D] \e[\-D] uppercase eth \[:y] \e[:y] y dieresis
637 .TE
638 .
639 .
640 .\" ====================================================================
641 .SS "Special character escape forms"
642 .\" ====================================================================
643 .
644 Glyphs that lack a character code in the basic Latin repertoire to
645 directly represent them are entered by one of several special character
646 escape forms.
647 .
648 Such glyphs can be simple or composite,
649 and accessed either by name or numerically by code point.
650 .
651 Code points and combining properties are determined by character
652 encoding standards,
653 whereas glyph names as used here originated in AT&T
654 .I troff \" AT&T
655 special character escape sequences.
656 .
657 Predefined glyph names use only characters in the basic Latin
658 repertoire.
659 .
660 .
661 .TP
662 .BI \[rs]( gl
663 is a special character escape sequence for the glyph with the
664 two-character name
665 .IR gl .
666 .
667 This is the original syntax form supported by AT&T
668 .IR troff .
669 .
670 The acute accent,
671 .BR \[rs](aa ,
672 is an example.
673 .
674 .
675 .TP
676 .BI \[rs]C\[aq] glyph-name \[aq]
677 is a special character escape sequence for
678 .IR glyph-name ,
679 which can be of arbitrary length.
680 .
681 The delimiter,
682 shown here as a neutral apostrophe,
683 can be any character not occurring in
684 .IR glyph-name .
685 .
686 This syntax form was introduced in later versions of AT&T
687 device-independent
688 .IR troff . \" AT&T
689 .
690 The foregoing acute accent example can be expressed
691 as
692 .BR \[rs]C\[aq]aa\[aq] .
693 .
694 .
695 .TP
696 .BI \[rs][ glyph-name ]
697 is a special character escape sequence for
698 .IR glyph-name ,
699 which can be of arbitrary length but must not contain a closing square
700 bracket
701 .RB \[lq] ] \[rq].
702 .
703 (No glyph names predefined by
704 .I groff
705 employ
706 .RB \[lq] ] \[rq].)
707 .
708 The foregoing acute accent example can be expressed in
709 .I groff
710 as
711 .BR \[rs][aa] .
712 .
713 .
714 .P
715 .BI \[rs]C\[aq] c \[aq]
716 and
717 .BI \[rs][ c ]
718 are not synonyms for the ordinary character
719 .RI \[lq] c \[rq],
720 but request the special character named
721 .RB \[lq] \[rs] \c
722 .IR c \[rq].
723 .
724 For example,
725 .RB \[lq] \[rs][a] \[rq]
726 is not \[lq]a\[rq],
727 but rather a special character with the internal glyph name
728 (used in font description files and diagnostic messages)
729 .BR \[rs]a ,
730 which is typically undefined.
731 .
732 The only such glyph name
733 .I groff
734 predefines is the minus sign,
735 which can therefore be accessed as
736 .B \[rs]C\[aq]\-\[aq]
737 or
738 .BR \[rs][\-] .
739 .
740 .
741 .TP
742 .BI \[rs][ "base-char composite-1 composite-2"\~\c
743 \&.\|.\|.\~\c
744 .IB composite-n ]
745 is a composite glyph.
746 .
747 Glyphs like a lowercase \[lq]e\[rq] with an acute accent,
748 as in the word \[lq]caf\[e aa]\[rq],
749 can be expressed as
750 .BR "\[rs][e aa]" .
751 .
752 See subsection \[lq]Accents\[rq] below for a table of combining glyph
753 names.
754 .
755 .
756 .P
757 Unicode encodes far more characters than
758 .I groff
759 has glyph names for;
760 special character escape forms based on numerical code points enable
761 access to any of them.
762 .
763 Frequently used glyphs or glyph combinations can be stored in strings,
764 and new glyph names can be created
765 .I "ad hoc"
766 with the
767 .B char
768 request;
769 see
770 .MR groff 7 .
771 .
772 .
773 .TP
774 .BI \[rs][u nnnn\c
775 .RI [ n\c
776 .RI [ n ]]\c
777 .B ]
778 is a Unicode numeric special character escape sequence.
779 .
780 Any Unicode code point can be accessed with four to six hexadecimal
781 digits,
782 with hexadecimal letters accepted in uppercase form only.
783 .
784 Thus,
785 .B \[rs][u02DA]
786 accesses the (spacing) ring accent,
787 producing \[lq]\[u02DA]\[rq].
788 .
789 .
790 .\" Use "GNU troff" in this paragraph because the contrast with AT&T
791 .\" troff, which antedated Unicode, is important, and that contrast is
792 .\" obscured with the default empty command prefix on "troff".
793 .P
794 Unicode code points can be composed as well;
795 when they are,
796 GNU
797 .I troff \" GNU
798 requires NFD
799 (Normalization Form D),
800 where all Unicode glyphs are maximally decomposed.
801 .
802 (Exception:
803 precomposed characters in the Latin-1 supplement described above are
804 also accepted.
805 .
806 Do not count on this exception remaining in a future
807 GNU
808 .I troff \" GNU
809 that accepts UTF-8 input directly.)
810 .
811 .
812 Thus,
813 GNU
814 .I troff \" GNU
815 accepts
816 .RB \[lq]caf \[rs][\[aq]e] \[rq],
817 .RB \[lq]caf \[rs][e\~aa] \[rq],
818 and
819 .RB \[lq]caf \[rs][u0065_0301] \[rq],
820 as ways to input \[lq]caf\['e]\[rq].
821 .
822 (Due to its legacy 8-bit encoding compatibility,
823 at present it also accepts
824 .RB \[lq]caf \[rs][u00E9] \[rq]
825 on ISO Latin-1 systems.)
826 .
827 .
828 .TP
829 .BI \[rs][u base-char\c
830 [\c
831 .BI _ combining-component\c
832 ].\|.\|.]
833 constructs a composite glyph from Unicode numeric special character
834 escape sequences.
835 .
836 The code points of the base glyph and the combining components are each
837 expressed in hexadecimal,
838 with an underscore
839 .RB ( _ )
840 separating each component.
841 .
842 Thus,
843 .B \[rs][u006E_0303]
844 produces \[lq]\[u006E_0303]\[rq].
845 .
846 .
847 .TP
848 .BI \[rs][char nnn ]
849 expresses an eight-bit code point where
850 .I nnn
851 is the code point of the character,
852 a decimal number between 0 and\~255
853 without leading zeroes.
854 .
855 This legacy numeric special character escape sequence is used to map
856 characters onto glyphs via the
857 .B trin
858 request in macro files loaded by
859 .MR grotty 1 .
860 .
861 .
862 .\" ====================================================================
863 .SH "Glyph tables"
864 .\" ====================================================================
865 .
866 In this section,
867 .IR groff 's
868 glyph name repertoire is presented in tabular form.
869 .
870 The meanings of the columns are as follows.
871 .
872 .
873 .TP 8n
874 .B Output
875 shows the glyph as it appears on the device used to render this
876 document;
877 although it can have a notably different shape on other devices
878 (and is subject to user-directed translation and replacement),
879 .I groff
880 attempts reasonable equivalency on all output devices.
881 .
882 .
883 .TP
884 .B Input
885 shows the
886 .I groff
887 character
888 (ordinary or special)
889 that normally produces the glyph.
890 .
891 Some code points have multiple glyph names.
892 .
893 .
894 .TP
895 .B Unicode
896 is the code point notation for the glyph or combining glyph sequence as
897 described in subsection \[lq]Special character escape forms\[rq] above.
898 .
899 It corresponds to the standard notation for Unicode short identifiers
900 such that
901 .IR groff 's
902 .BI u nnnn
903 is equivalent to Unicode's
904 .RI U+ nnnn .
905 .\" And thereby hangs a tale...
906 .\" https://unicode.org/mail-arch/unicode-ml/y2005-m11/0060.html
907 .
908 .
909 .TP
910 .B Notes
911 describes the glyph,
912 elucidating the mnemonic value of the glyph name where possible.
913 .
914 .
915 .IP
916 A plus sign \[lq]+\[rq] indicates that the glyph name appears in the
917 AT&T
918 .I troff
919 user's manual,
920 CSTR\~#54
921 (1992 revision).
922 .
923 When using the AT&T special character syntax
924 .BI \[rs]( xx\c
925 , widespread portability can be expected from such names.
926 .
927 .
928 .IP
929 Entries marked with \[lq]***\[rq] denote glyphs used for mathematical
930 purposes.
931 .
932 On typesetting devices,
933 such glyphs are typically drawn from a
934 .I special
935 font
936 (see
937 .MR groff_font 5 ).
938 .
939 Often,
940 such glyphs lack bold or italic style forms or have metrics that look
941 incongruous in ordinary prose.
942 .
943 A few which are not uncommon in running text have \[lq]text
944 variants\[rq],
945 which should work better in that context.
946 .
947 Conversely,
948 a handful of glyphs that are normally drawn from a text font may be
949 required in mathematical equations.
950 .
951 Both sets of exceptions are noted in the tables where they appear
952 (\[lq]Logical symbols\[rq] and \[lq]Mathematical symbols\[rq]).
953 .
954 .
955 .\" ====================================================================
956 .SS "Basic Latin"
957 .\" ====================================================================
958 .
959 Apart from basic Latin characters with special mappings,
960 described in subsection \[lq]Fundamental character set\[rq] above,
961 a few others in that range have special character glyph names.
962 .
963 .\" XXX: I surmise that...
964 These were defined for ease of input on non-U.S.\& keyboards lacking
965 keycaps for them,
966 or for symmetry with other special character glyph names serving a
967 similar purpose.
968 .
969 .
970 .P
971 The vertical bar is overloaded;
972 the
973 .B \[rs][ba]
974 and
975 .B \[rs][or]
976 escape sequences may render differently.
977 .
978 See subsection \[lq]Mathematical symbols\[rq] below for special variants
979 of the plus,
980 minus,
981 and equals
982 signs normally drawn from this range.
983 .
984 .
985 .P
986 .TS
987 L L L Lx.
988 Output Input Unicode Notes
989 _
990 .T&
991 L Lf(CR) L Lx.
992 \[dq] \e[dq] u0022 neutral double quote
993 \[sh] \e[sh] u0023 number sign
994 \[Do] \e[Do] u0024 dollar sign
995 \[aq] \e[aq] u0027 apostrophe, neutral single quote
996 \[sl] \e[sl] u002F slash, solidus +
997 \[at] \e[at] u0040 at sign
998 \[lB] \e[lB] u005B left square bracket
999 \[rs] \e[rs] u005C reverse solidus
1000 \[rB] \e[rB] u005D right square bracket
1001 \[ha] \e[ha] u005E circumflex, caret, \[lq]hat\[rq]
1002 \[lC] \e[lC] u007B left brace
1003 | | u007C bar
1004 \[ba] \e[ba] u007C bar
1005 \[or] \e[or] u007C bitwise or +
1006 \[rC] \e[rC] u007D right brace
1007 \[ti] \e[ti] u007E tilde
1008 .TE
1009 .
1010 .
1011 .\" ====================================================================
1012 .SS "Supplementary Latin letters"
1013 .\" ====================================================================
1014 .
1015 Historically,
1016 .B \[rs][ss]
1017 could be considered a ligature of \[lq]sz\[rq].
1018 .
1019 An uppercase form is available as
1020 .BR \[rs][u1E9E] ,
1021 but in the German language it is of specialized use;
1022 \[ss] does
1023 .I not
1024 normally uppercase-transform to it,
1025 but rather to \[lq]SS\[rq].
1026 .
1027 \[lq]Lowercase f with hook\[rq] is also used as a function symbol;
1028 see subsection \[lq]Mathematical symbols\[rq] below.
1029 .
1030 .
1031 .P
1032 .if t .ne 2v
1033 .if n .ne 3v \" account for horizontal rule
1034 .TS
1035 L L L Lx.
1036 Output Input Unicode Notes
1037 _
1038 .T&
1039 L Lf(CR) L Lx.
1040 \[-D] \e[\-D] u00D0 uppercase eth
1041 \[Sd] \e[Sd] u00F0 lowercase eth
1042 \[TP] \e[TP] u00DE uppercase thorn
1043 \[Tp] \e[Tp] u00FE lowercase thorn
1044 \[ss] \e[ss] u00DF lowercase sharp s
1045 \[.i] \e[.i] u0131 i without tittle
1046 \[.j] \e[.j] u0237 j without tittle
1047 \[Fn] \e[Fn] u0192 lowercase f with hook, function
1048 \[/L] \e[/L] u0141 L with stroke
1049 \[/l] \e[/l] u0142 l with stroke
1050 \[/O] \e[/O] u00D8 O with stroke
1051 \[/o] \e[/o] u00F8 o with stroke
1052 .TE
1053 .
1054 .
1055 .br
1056 .if t .ne 4v
1057 .if n .ne 5v \" account for horizontal rule
1058 .\" ====================================================================
1059 .SS "Ligatures and digraphs"
1060 .\" ====================================================================
1061 .
1062 .TS
1063 L L L Lx.
1064 Output Input Unicode Notes
1065 _
1066 .T&
1067 L Lf(CR) L Lx.
1068 \[ff] \e[ff] u0066_0066 ff ligature +
1069 \[fi] \e[fi] u0066_0069 fi ligature +
1070 \[fl] \e[fl] u0066_006C fl ligature +
1071 \[Fi] \e[Fi] u0066_0066_0069 ffi ligature +
1072 \[Fl] \e[Fl] u0066_0066_006C ffl ligature +
1073 \[AE] \e[AE] u00C6 AE ligature
1074 \[ae] \e[ae] u00E6 ae ligature
1075 \[OE] \e[OE] u0152 OE ligature
1076 \[oe] \e[oe] u0153 oe ligature
1077 \[IJ] \e[IJ] u0132 IJ digraph
1078 \[ij] \e[ij] u0133 ij digraph
1079 .TE
1080 .
1081 .
1082 .\" ====================================================================
1083 .SS Accents
1084 .\" ====================================================================
1085 .
1086 Normally,
1087 the formatting of a special character advances the drawing position as
1088 an ordinary character does.
1089 .
1090 .IR groff 's
1091 .B composite
1092 request designates a special character as combining.
1093 .
1094 The
1095 .I composite.tmac
1096 macro file,
1097 loaded automatically by the default
1098 .IR troffrc ,
1099 maps the following special characters to the combining characters shown
1100 below.
1101 .
1102 The non-combining code point in parentheses is used when the special
1103 character occurs in isolation
1104 (compare
1105 .RB \[lq] "caf\[rs][e aa]" \[rq]
1106 and
1107 .RB \[lq] "caf\[rs][aa]e" \[rq]).
1108 .
1109 .
1110 .P
1111 .if t .ne 3v
1112 .if n .ne 4v \" account for horizontal rule
1113 .TS
1114 L L L Lx.
1115 Output Input Unicode Notes
1116 _
1117 .T&
1118 L Lf(CR) L Lx.
1119 \[a"] \e[a"] u030B (u02DD) double acute accent
1120 \[a-] \e[a\-] u0304 (u00AF) macron accent
1121 \[a.] \e[a.] u0307 (u02D9) dot accent
1122 \[a^] \e[a\[ha]] u0302 (u005E) circumflex accent
1123 \[aa] \e[aa] u0301 (u00B4) acute accent +
1124 \[ga] \e[ga] u0300 (u0060) grave accent +
1125 \[ab] \e[ab] u0306 (u02D8) breve accent
1126 \[ac] \e[ac] u0327 (u00B8) cedilla accent
1127 \[ad] \e[ad] u0308 (u00A8) dieresis accent
1128 \[ah] \e[ah] u030C (u02C7) caron accent
1129 \[ao] \e[ao] u030A (u02DA) ring accent
1130 \[a~] \e[a\[ti]] u0303 (u007E) tilde accent
1131 \[ho] \e[ho] u0328 (u02DB) hook accent
1132 .TE
1133 .
1134 .
1135 .\" ====================================================================
1136 .SS "Accented characters"
1137 .\" ====================================================================
1138 .
1139 All of these glyphs can be composed using combining glyph names as
1140 described in subsection \[lq]Special character escape forms\[rq] above;
1141 the names below are short aliases for convenience.
1142 .
1143 .
1144 .P
1145 .if t .ne 2v
1146 .if n .ne 3v \" account for horizontal rule
1147 .TS
1148 L L L Lx.
1149 Output Input Unicode Notes
1150 _
1151 .T&
1152 L Lf(CR) L Lx.
1153 \['A] \e[\[aq]A] u0041_0301 A acute
1154 \['C] \e[\[aq]C] u0043_0301 C acute
1155 \['E] \e[\[aq]E] u0045_0301 E acute
1156 \['I] \e[\[aq]I] u0049_0301 I acute
1157 \['O] \e[\[aq]O] u004F_0301 O acute
1158 \['U] \e[\[aq]U] u0055_0301 U acute
1159 \['Y] \e[\[aq]Y] u0059_0301 Y acute
1160 \['a] \e[\[aq]a] u0061_0301 a acute
1161 \['c] \e[\[aq]c] u0063_0301 c acute
1162 \['e] \e[\[aq]e] u0065_0301 e acute
1163 \['i] \e[\[aq]i] u0069_0301 i acute
1164 \['o] \e[\[aq]o] u006F_0301 o acute
1165 \['u] \e[\[aq]u] u0075_0301 u acute
1166 \['y] \e[\[aq]y] u0079_0301 y acute
1167
1168 \[:A] \e[:A] u0041_0308 A dieresis
1169 \[:E] \e[:E] u0045_0308 E dieresis
1170 \[:I] \e[:I] u0049_0308 I dieresis
1171 \[:O] \e[:O] u004F_0308 O dieresis
1172 \[:U] \e[:U] u0055_0308 U dieresis
1173 \[:Y] \e[:Y] u0059_0308 Y dieresis
1174 \[:a] \e[:a] u0061_0308 a dieresis
1175 \[:e] \e[:e] u0065_0308 e dieresis
1176 \[:i] \e[:i] u0069_0308 i dieresis
1177 \[:o] \e[:o] u006F_0308 o dieresis
1178 \[:u] \e[:u] u0075_0308 u dieresis
1179 \[:y] \e[:y] u0079_0308 y dieresis
1180
1181 \[^A] \e[\[ha]A] u0041_0302 A circumflex
1182 \[^E] \e[\[ha]E] u0045_0302 E circumflex
1183 \[^I] \e[\[ha]I] u0049_0302 I circumflex
1184 \[^O] \e[\[ha]O] u004F_0302 O circumflex
1185 \[^U] \e[\[ha]U] u0055_0302 U circumflex
1186 \[^a] \e[\[ha]a] u0061_0302 a circumflex
1187 \[^e] \e[\[ha]e] u0065_0302 e circumflex
1188 \[^i] \e[\[ha]i] u0069_0302 i circumflex
1189 \[^o] \e[\[ha]o] u006F_0302 o circumflex
1190 \[^u] \e[\[ha]u] u0075_0302 u circumflex
1191
1192 \[`A] \e[\[ga]A] u0041_0300 A grave
1193 \[`E] \e[\[ga]E] u0045_0300 E grave
1194 \[`I] \e[\[ga]I] u0049_0300 I grave
1195 \[`O] \e[\[ga]O] u004F_0300 O grave
1196 \[`U] \e[\[ga]U] u0055_0300 U grave
1197 \[`a] \e[\[ga]a] u0061_0300 a grave
1198 \[`e] \e[\[ga]e] u0065_0300 e grave
1199 \[`i] \e[\[ga]i] u0069_0300 i grave
1200 \[`o] \e[\[ga]o] u006F_0300 o grave
1201 \[`u] \e[\[ga]u] u0075_0300 u grave
1202
1203 \[~A] \e[\[ti]A] u0041_0303 A tilde
1204 \[~N] \e[\[ti]N] u004E_0303 N tilde
1205 \[~O] \e[\[ti]O] u004F_0303 O tilde
1206 \[~a] \e[\[ti]a] u0061_0303 a tilde
1207 \[~n] \e[\[ti]n] u006E_0303 n tilde
1208 \[~o] \e[\[ti]o] u006F_0303 o tilde
1209
1210 \[vS] \e[vS] u0053_030C S caron
1211 \[vs] \e[vs] u0073_030C s caron
1212 \[vZ] \e[vZ] u005A_030C Z caron
1213 \[vz] \e[vz] u007A_030C z caron
1214
1215 \[,C] \e[,C] u0043_0327 C cedilla
1216 \[,c] \e[,c] u0063_0327 c cedilla
1217
1218 \[oA] \e[oA] u0041_030A A ring
1219 \[oa] \e[oa] u0061_030A a ring
1220 .TE
1221 .
1222 .
1223 .\" ====================================================================
1224 .SS "Quotation marks"
1225 .\" ====================================================================
1226 .
1227 The neutral double quote,
1228 often useful when documenting programming languages,
1229 is also available as a special character for convenient embedding in
1230 macro arguments;
1231 see subsection \[lq]Fundamental character set\[rq] above.
1232 .
1233 .
1234 .P
1235 .if t .ne 2v
1236 .if n .ne 3v \" account for horizontal rule
1237 .TS
1238 L L L Lx.
1239 Output Input Unicode Notes
1240 _
1241 .T&
1242 L Lf(CR) L Lx.
1243 \[Bq] \e[Bq] u201E low double comma quote
1244 \[bq] \e[bq] u201A low single comma quote
1245 \[lq] \e[lq] u201C left double quote
1246 \[rq] \e[rq] u201D right double quote
1247 \[oq] \e[oq] u2018 single opening (left) quote
1248 \[cq] \e[cq] u2019 single closing (right) quote
1249 \[aq] \e[aq] u0027 apostrophe, neutral single quote
1250 \[dq] " u0022 neutral double quote
1251 \[dq] \e[dq] u0022 neutral double quote
1252 \[Fo] \e[Fo] u00AB left double chevron
1253 \[Fc] \e[Fc] u00BB right double chevron
1254 \[fo] \e[fo] u2039 left single chevron
1255 \[fc] \e[fc] u203A right single chevron
1256 .TE
1257 .
1258 .
1259 .\" ====================================================================
1260 .SS Punctuation
1261 .\" ====================================================================
1262 .
1263 The Unicode name for U+00B7 is \[lq]middle dot\[rq],
1264 which is unfortunately confusable with the
1265 .I groff
1266 mnemonic for the visually similar but semantically distinct
1267 multiplication dot;
1268 see subsection \[lq]Mathematical symbols\[rq] below.
1269 .
1270 .
1271 .P
1272 .if t .ne 2v
1273 .if n .ne 3v \" account for horizontal rule
1274 .TS
1275 L L L Lx.
1276 Output Input Unicode Notes
1277 _
1278 .T&
1279 L Lf(CR) L Lx.
1280 \[r!] \e[r!] u00A1 inverted exclamation mark
1281 \[r?] \e[r?] u00BF inverted question mark
1282 \[pc] \e[pc] u00B7 centered period
1283 \[em] \e[em] u2014 em-dash +
1284 \[en] \e[en] u2013 en-dash
1285 \[hy] \e[hy] u2010 hyphen +
1286 .TE
1287 .
1288 .
1289 .\" ====================================================================
1290 .SS Brackets
1291 .\" ====================================================================
1292 .
1293 On typesetting devices,
1294 the bracket extensions are font-invariant glyphs;
1295 that is,
1296 they are rendered the same way regardless of font
1297 (with a drawing escape sequence).
1298 .
1299 On terminals,
1300 they are
1301 .I not
1302 font-invariant;
1303 .I groff
1304 maps them rather arbitrarily to U+23AA
1305 (\[lq]curly bracket extension\[rq]).
1306 .
1307 In AT&T
1308 .IR troff ,
1309 only one glyph was available to vertically extend
1310 brackets,
1311 braces,
1312 and
1313 parentheses:
1314 .BR \[rs](bv .
1315 .
1316 .
1317 .
1318 .P
1319 Not all devices supply bracket pieces that can be piled up with
1320 .B \[rs]b
1321 due to the restrictions of the escape's piling algorithm.
1322 .
1323 A general solution to build brackets out of pieces is the following
1324 macro:
1325 .
1326 .
1327 .RS
1328 .EX
1329 \&.\e" Make a pile centered vertically 0.5em above the baseline.
1330 \&.\e" The first argument is placed at the top.
1331 \&.\e" The pile is returned in string \[aq]pile\[aq].
1332 \&.eo
1333 \&.de pile\-make
1334 \&.\& nr pile\-wd 0
1335 \&.\& nr pile\-ht 0
1336 \&.\& ds pile\-args
1337 \&.\&
1338 \&.\& nr pile\-# \en[.$]
1339 \&.\& while \en[pile\-#] \e{\e
1340 \&.\& nr pile\-wd (\en[pile\-wd] >? \ew\[aq]\e$[\en[pile\-#]]\[aq])
1341 \&.\& nr pile\-ht +(\en[rst] \- \en[rsb])
1342 \&.\& as pile\-args \ev\[aq]\en[rsb]u\[aq]\e"
1343 \&.\& as pile\-args \eZ\[aq]\e$[\en[pile\-#]]\[aq]\e"
1344 \&.\& as pile\-args \ev\[aq]\-\en[rst]u\[aq]\e"
1345 \&.\& nr pile\-# \-1
1346 \&.\& \e}
1347 \&.\&
1348 \&.\& ds pile \ev\[aq](\-0.5m + (\en[pile\-ht]u / 2u))\[aq]\e"
1349 \&.\& as pile \e*[pile\-args]\e"
1350 \&.\& as pile \ev\[aq]((\en[pile\-ht]u / 2u) + 0.5m)\[aq]\e"
1351 \&.\& as pile \eh\[aq]\en[pile\-wd]u\[aq]\e"
1352 \&..
1353 \&.ec
1354 .EE
1355 .RE
1356 .
1357 .
1358 .P
1359 Another complication is the fact that some glyphs which represent
1360 bracket pieces in AT&T
1361 .I troff
1362 can be used for other mathematical symbols as well,
1363 for example
1364 .B \[rs](lf
1365 and
1366 .BR \[rs](rf ,
1367 which provide the floor operator.
1368 .
1369 Some output devices,
1370 such as
1371 .BR dvi ,
1372 don't unify such glyphs.
1373 .
1374 For this reason,
1375 the glyphs
1376 .BR \[rs][lf] ,
1377 .BR \[rs][rf] ,
1378 .BR \[rs][lc] ,
1379 and
1380 .B \[rs][rc]
1381 are not unified with similar-looking bracket pieces.
1382 .
1383 In
1384 .IR groff ,
1385 only glyphs with long names are guaranteed to pile up correctly for all
1386 devices\[em]provided those glyphs are available.
1387 .
1388 .
1389 .P
1390 .if t .ne 2v
1391 .if n .ne 3v \" account for horizontal rule
1392 .TS
1393 L L L Lx.
1394 Output Input Unicode Notes
1395 _
1396 .T&
1397 L Lf(CR) L Lx.
1398 \[lB] [ u005B left square bracket
1399 \[lB] \e[lB] u005B left square bracket
1400 \[rB] ] u005D right square bracket
1401 \[rB] \e[rB] u005D right square bracket
1402 \[lC] { u007B left brace
1403 \[lC] \e[lC] u007B left brace
1404 \[rC] } u007D right brace
1405 \[rC] \e[rC] u007D right brace
1406 \[la] \e[la] u27E8 left angle bracket
1407 \[ra] \e[ra] u27E9 right angle bracket
1408 \[bv] \e[bv] u23AA brace vertical extension + ***
1409 \[braceex] \e[braceex] u23AA brace vertical extension
1410
1411 \[bracketlefttp] \e[bracketlefttp] u23A1 left square bracket top
1412 \[bracketleftex] \e[bracketleftex] u23A2 left square bracket extension
1413 \[bracketleftbt] \e[bracketleftbt] u23A3 left square bracket bottom
1414
1415 \[bracketrighttp] \e[bracketrighttp] u23A4 right square bracket top
1416 \[bracketrightex] \e[bracketrightex] u23A5 right square bracket extension
1417 \[bracketrightbt] \e[bracketrightbt] u23A6 right square bracket bottom
1418
1419 \[lt] \e[lt] u23A7 left brace top +
1420 \[lk] \e[lk] u23A8 left brace middle +
1421 \[lb] \e[lb] u23A9 left brace bottom +
1422 \[bracelefttp] \e[bracelefttp] u23A7 left brace top
1423 \[braceleftmid] \e[braceleftmid] u23A8 left brace middle
1424 \[braceleftbt] \e[braceleftbt] u23A9 left brace bottom
1425 \[braceleftex] \e[braceleftex] u23AA left brace extension
1426
1427 \[rt] \e[rt] u23AB right brace top +
1428 \[rk] \e[rk] u23AC right brace middle +
1429 \[rb] \e[rb] u23AD right brace bottom +
1430 \[bracerighttp] \e[bracerighttp] u23AB right brace top
1431 \[bracerightmid] \e[bracerightmid] u23AC right brace middle
1432 \[bracerightbt] \e[bracerightbt] u23AD right brace bottom
1433 \[bracerightex] \e[bracerightex] u23AA right brace extension
1434
1435 \[parenlefttp] \e[parenlefttp] u239B left parenthesis top
1436 \[parenleftex] \e[parenleftex] u239C left parenthesis extension
1437 \[parenleftbt] \e[parenleftbt] u239D left parenthesis bottom
1438 \[parenrighttp] \e[parenrighttp] u239E right parenthesis top
1439 \[parenrightex] \e[parenrightex] u239F right parenthesis extension
1440 \[parenrightbt] \e[parenrightbt] u23A0 right parenthesis bottom
1441 .TE
1442 .
1443 .
1444 .br
1445 .if t .ne 4v
1446 .if n .ne 5v \" account for horizontal rule
1447 .\" ====================================================================
1448 .SS Arrows
1449 .\" ====================================================================
1450 .
1451 .TS
1452 L L L Lx.
1453 Output Input Unicode Notes
1454 _
1455 .T&
1456 L Lf(CR) L Lx.
1457 \[<-] \e[<\-] u2190 horizontal arrow left +
1458 \[->] \e[\->] u2192 horizontal arrow right +
1459 \[<>] \e[<>] u2194 bidirectional horizontal arrow
1460 \[da] \e[da] u2193 vertical arrow down +
1461 \[ua] \e[ua] u2191 vertical arrow up +
1462 \[va] \e[va] u2195 bidirectional vertical arrow
1463 \[lA] \e[lA] u21D0 horizontal double arrow left
1464 \[rA] \e[rA] u21D2 horizontal double arrow right
1465 \[hA] \e[hA] u21D4 bidirectional horizontal double arrow
1466 \[dA] \e[dA] u21D3 vertical double arrow down
1467 \[uA] \e[uA] u21D1 vertical double arrow up
1468 \[vA] \e[vA] u21D5 bidirectional vertical double arrow
1469 \[an] \e[an] u23AF horizontal arrow extension
1470 .TE
1471 .
1472 .
1473 .\" ====================================================================
1474 .SS "Rules and lines"
1475 .\" ====================================================================
1476 .
1477 On typesetting devices,
1478 the font-invariant glyphs
1479 (see subsection \[lq]Brackets\[rq] above)
1480 .BR \[rs][br] ,
1481 .BR \[rs][ul] ,
1482 and
1483 .B \[rs][rn]
1484 form corners when adjacent;
1485 they can be used to build boxes.
1486 .
1487 On terminal devices,
1488 they are mapped as shown in the table.
1489 .
1490 The Unicode-derived names of these three glyphs are approximations.
1491 .
1492 .
1493 .P
1494 The input character
1495 .B _
1496 always accesses the underscore glyph in a font;
1497 .\" unless one isn't available, but this seems to be only a theoretical
1498 .\" concern--what font doesn't support every ASCII codepoint these days?
1499 .BR \[rs][ul] ,
1500 by contrast,
1501 may be font-invariant on typesetting devices.
1502 .
1503 .
1504 .P
1505 The baseline rule
1506 .B \[rs][ru]
1507 is a font-invariant glyph,
1508 namely a rule of one-half em.
1509 .
1510 .
1511 .P
1512 In AT&T
1513 .IR troff , \" AT&T
1514 .B \[rs][rn]
1515 also served as a one\~en extension of the square root symbol.
1516 .
1517 .I groff
1518 favors
1519 .B \[rs][radicalex]
1520 for this purpose;
1521 see subsection \[lq]Mathematical symbols\[rq] below.
1522 .
1523 .
1524 .P
1525 .if t .ne 2v
1526 .if n .ne 3v \" account for horizontal rule
1527 .TS
1528 L L L Lx.
1529 Output Input Unicode Notes
1530 _
1531 .T&
1532 L Lf(CR) L Lx.
1533 | | u007C bar
1534 \[ba] \e[ba] u007C bar
1535 \[br] \e[br] u2502 box rule +
1536 \&_ \&_ u005F underscore, low line +
1537 \[ul] \e[ul] --- underrule +
1538 \[rn] \e[rn] u203E overline +
1539 \[ru] \e[ru] --- baseline rule +
1540 \[bb] \e[bb] u00A6 broken bar
1541 \[sl] / u002F slash, solidus +
1542 \[sl] \e[sl] u002F slash, solidus +
1543 \[rs] \e[rs] u005C reverse solidus
1544 .TE
1545 .
1546 .
1547 .br
1548 .if t .ne 4v
1549 .if n .ne 5v \" account for horizontal rule
1550 .\" ====================================================================
1551 .SS "Text markers"
1552 .\" ====================================================================
1553 .
1554 .TS
1555 L L L Lx.
1556 Output Input Unicode Notes
1557 _
1558 .T&
1559 L Lf(CR) L Lx.
1560 \[ci] \e[ci] u25CB circle +
1561 \[bu] \e[bu] u2022 bullet +
1562 \[dg] \e[dg] u2020 dagger +
1563 \[dd] \e[dd] u2021 double dagger +
1564 \[lz] \e[lz] u25CA lozenge, diamond
1565 \[sq] \e[sq] u25A1 square +
1566 \[ps] \e[ps] u00B6 pilcrow sign
1567 \[sc] \e[sc] u00A7 section sign +
1568 \[lh] \e[lh] u261C hand pointing left +
1569 \[rh] \e[rh] u261E hand pointing right +
1570 \[at] @ u0040 at sign
1571 \[at] \e[at] u0040 at sign
1572 \[sh] # u0023 number sign
1573 \[sh] \e[sh] u0023 number sign
1574 \[CR] \e[CR] u21B5 carriage return
1575 \[OK] \e[OK] u2713 check mark
1576 .TE
1577 .
1578 .\" ====================================================================
1579 .SS "Legal symbols"
1580 .\" ====================================================================
1581 .
1582 The Bell System logo is not supported in
1583 .IR groff .
1584 .
1585 .
1586 .P
1587 .if t .ne 2v
1588 .if n .ne 3v \" account for horizontal rule
1589 .TS
1590 L L L Lx.
1591 Output Input Unicode Notes
1592 _
1593 .T&
1594 L Lf(CR) L Lx.
1595 \[co] \e[co] u00A9 copyright sign +
1596 \[rg] \e[rg] u00AE registered sign +
1597 \[tm] \e[tm] u2122 trade mark sign
1598 \[bs] \e[bs] --- Bell System logo +
1599 .TE
1600 .
1601 .
1602 .br
1603 .if t .ne 4v
1604 .if n .ne 5v \" account for horizontal rule
1605 .\" ====================================================================
1606 .SS "Currency symbols"
1607 .\" ====================================================================
1608 .
1609 .TS
1610 L L L Lx.
1611 Output Input Unicode Notes
1612 _
1613 .T&
1614 L Lf(CR) L Lx.
1615 \[Do] $ u0024 dollar sign
1616 \[Do] \e[Do] u0024 dollar sign
1617 \[ct] \e[ct] u00A2 cent sign +
1618 \[eu] \e[eu] u20AC Euro sign
1619 \[Eu] \e[Eu] u20AC variant Euro sign
1620 \[Ye] \e[Ye] u00A5 yen sign
1621 \[Po] \e[Po] u00A3 pound sign
1622 \[Cs] \e[Cs] u00A4 currency sign
1623 .TE
1624 .
1625 .
1626 .br
1627 .if t .ne 4v
1628 .if n .ne 5v \" account for horizontal rule
1629 .\" ====================================================================
1630 .SS Units
1631 .\" ====================================================================
1632 .
1633 .TS
1634 L L L Lx.
1635 Output Input Unicode Notes
1636 _
1637 .T&
1638 L Lf(CR) L Lx.
1639 \[de] \e[de] u00B0 degree sign +
1640 \[%0] \e[%0] u2030 per thousand, per mille sign
1641 \[fm] \e[fm] u2032 arc minute sign, foot mark +
1642 \[sd] \e[sd] u2033 arc second sign
1643 \[mc] \e[mc] u00B5 micro sign
1644 \[Of] \e[Of] u00AA feminine ordinal indicator
1645 \[Om] \e[Om] u00BA masculine ordinal indicator
1646 .TE
1647 .
1648 .
1649 .\" ====================================================================
1650 .SS "Logical symbols"
1651 .\" ====================================================================
1652 .
1653 The variants of the not sign may differ in appearance or spacing
1654 depending on the device and font selected.
1655 .
1656 Unicode does not encode a discrete \[lq]bitwise or\[rq] sign:
1657 on typesetting devices,
1658 it is drawn shorter than the bar,
1659 about the same height as a capital letter.
1660 .
1661 Terminal devices unify
1662 .B \[rs][ba]
1663 and
1664 .BR \[rs][or] .
1665 .
1666 .
1667 .P
1668 .if t .ne 2v
1669 .if n .ne 3v \" account for horizontal rule
1670 .TS
1671 L L L Lx.
1672 Output Input Unicode Notes
1673 _
1674 .T&
1675 L Lf(CR) L Lx.
1676 \[AN] \e[AN] u2227 logical and
1677 \[OR] \e[OR] u2228 logical or
1678 \[no] \e[no] u00AC logical not + ***
1679 \[tno] \e[tno] u00AC text variant of \f[B]\e[no]\f[]
1680 \[te] \e[te] u2203 there exists
1681 \[fa] \e[fa] u2200 for all
1682 \[st] \e[st] u220B such that
1683 \[3d] \e[3d] u2234 therefore
1684 \[tf] \e[tf] u2234 therefore
1685 | | u007C bar
1686 \[or] \e[or] u007C bitwise or +
1687 .TE
1688 .
1689 .
1690 .\" ====================================================================
1691 .SS "Mathematical symbols"
1692 .\" ====================================================================
1693 .
1694 .B \[rs][Fn]
1695 also appears in subsection \[lq]Supplementary Latin letters\[rq] above.
1696 .
1697 Observe the two varieties of the
1698 plus-minus,
1699 multiplication,
1700 and division signs;
1701 .BR \[rs][+\-] ,
1702 .BR \[rs][mu] ,
1703 and
1704 .B \[rs][di]
1705 are normally drawn from the special font,
1706 but have text font variants.
1707 .
1708 Also be aware of three glyphs available in special font variants that
1709 are normally drawn from text fonts:
1710 the plus,
1711 minus,
1712 and equals signs.
1713 .
1714 These variants may differ in appearance or spacing depending on the
1715 device and font selected.
1716 .
1717 .
1718 .P
1719 In AT&T
1720 .IR troff ,
1721 .B \[rs](rn
1722 (\[lq]root en extender\[rq])
1723 served as the horizontal extension of the radical
1724 (square root)
1725 sign,
1726 .BR \[rs](sr ,
1727 and was drawn at the maximum height of the typeface's bounding box;
1728 this enabled the special character to double as an overline
1729 (see subsection \[lq]Rules and lines\[rq] above).
1730 .
1731 A contemporary font's radical sign might not ascend to such an extreme.
1732 .
1733 In
1734 .IR groff ,
1735 you can instead use
1736 .B \[rs][radicalex]
1737 to continue the radical sign
1738 .BR \[rs][sr] ;
1739 these special characters are intended for use with text fonts.
1740 .
1741 .B \[rs][sqrt]
1742 and
1743 .B \[rs][sqrtex]
1744 are their counterparts with mathematical spacing.
1745 .
1746 .
1747 .P
1748 .if t .ne 2v
1749 .if n .ne 3v \" account for horizontal rule
1750 .TS
1751 L L L Lx.
1752 Output Input Unicode Notes
1753 _
1754 .T&
1755 L Lf(CR) L Lx.
1756 \[12] \e[12] u00BD one half symbol +
1757 \[14] \e[14] u00BC one quarter symbol +
1758 \[34] \e[34] u00BE three quarters symbol +
1759 \[18] \e[18] u215B one eighth symbol
1760 \[38] \e[38] u215C three eighths symbol
1761 \[58] \e[58] u215D five eighths symbol
1762 \[78] \e[78] u215E seven eighths symbol
1763 \[S1] \e[S1] u00B9 superscript one
1764 \[S2] \e[S2] u00B2 superscript two
1765 \[S3] \e[S3] u00B3 superscript three
1766
1767 + + u002B plus
1768 \[pl] \e[pl] u002B special variant of plus + ***
1769 \- \e[\-] u002D minus
1770 \[mi] \e[mi] u2212 special variant of minus + ***
1771 \[-+] \e[\-+] u2213 minus-plus
1772 \[+-] \e[+\-] u00B1 plus-minus + ***
1773 \[t+-] \e[t+\-] u00B1 text variant of \f[B]\e[+\-]\f[]
1774 \[md] \e[md] u22C5 multiplication dot
1775 \[mu] \e[mu] u00D7 multiplication sign + ***
1776 \[tmu] \e[tmu] u00D7 text variant of \f[B]\e[mu]\f[]
1777 \[c*] \e[c*] u2297 circled times
1778 \[c+] \e[c+] u2295 circled plus
1779 \[di] \e[di] u00F7 division sign + ***
1780 \[tdi] \e[tdi] u00F7 text variant of \f[B]\e[di]\f[]
1781 \[f/] \e[f/] u2044 fraction slash
1782 * * u002A asterisk
1783 \[**] \e[**] u2217 mathematical asterisk +
1784
1785 \[<=] \e[<=] u2264 less than or equal to +
1786 \[>=] \e[>=] u2265 greater than or equal to +
1787 \[<<] \e[<<] u226A much less than
1788 \[>>] \e[>>] u226B much greater than
1789 \&= \&= u003D equals
1790 \[eq] \e[eq] u003D special variant of equals + ***
1791 \[!=] \e[!=] u003D_0338 not equals +
1792 \[==] \e[==] u2261 equivalent +
1793 \[ne] \e[ne] u2261_0338 not equivalent
1794 \[=~] \e[=\[ti]] u2245 approximately equal to
1795 \[|=] \e[|=] u2243 asymptotically equal to +
1796 \[ti] \e[ti] u007E tilde +
1797 \[ap] \e[ap] u223C similar to, tilde operator +
1798 \[~~] \e[\[ti]\[ti]] u2248 almost equal to
1799 \[~=] \e[\[ti]=] u2248 almost equal to
1800 \[pt] \e[pt] u221D proportional to +
1801
1802 \[es] \e[es] u2205 empty set +
1803 \[mo] \e[mo] u2208 element of a set +
1804 \[nm] \e[nm] u2208_0338 not element of set
1805 \[sb] \e[sb] u2282 proper subset +
1806 \[nb] \e[nb] u2282_0338 not subset
1807 \[sp] \e[sp] u2283 proper superset +
1808 \[nc] \e[nc] u2283_0338 not superset
1809 \[ib] \e[ib] u2286 subset or equal +
1810 \[ip] \e[ip] u2287 superset or equal +
1811 \[ca] \e[ca] u2229 intersection, cap +
1812 \[cu] \e[cu] u222A union, cup +
1813
1814 \[/_] \e[/_] u2220 angle
1815 \[pp] \e[pp] u22A5 perpendicular
1816 \[is] \e[is] u222B integral +
1817 \[integral] \e[integral] u222B integral ***
1818 \[sum] \e[sum] u2211 summation ***
1819 \[product] \e[product] u220F product ***
1820 \[coproduct] \e[coproduct] u2210 coproduct ***
1821 \[gr] \e[gr] u2207 gradient +
1822 \[sr] \e[sr] u221A radical sign, square root +
1823 \[rn] \e[rn] u203E overline +
1824 \[radicalex] \e[radicalex] --- radical extension
1825 \[sqrt] \e[sqrt] u221A radical sign, square root ***
1826 \[sqrtex] \e[sqrtex] --- radical extension ***
1827
1828 \[lc] \e[lc] u2308 left ceiling +
1829 \[rc] \e[rc] u2309 right ceiling +
1830 \[lf] \e[lf] u230A left floor +
1831 \[rf] \e[rf] u230B right floor +
1832
1833 \[if] \e[if] u221E infinity +
1834 \[Ah] \e[Ah] u2135 aleph symbol
1835 \[Fn] \e[Fn] u0192 lowercase f with hook, function
1836 \[Im] \e[Im] u2111 blackletter I, imaginary part
1837 \[Re] \e[Re] u211C blackletter R, real part
1838 \[wp] \e[wp] u2118 Weierstrass p
1839 \[pd] \e[pd] u2202 partial differential
1840 \[-h] \e[\-h] u210F h bar
1841 \[hbar] \e[hbar] u210F h bar
1842 .TE
1843 .
1844 .
1845 .\" ====================================================================
1846 .SS "Greek glyphs"
1847 .\" ====================================================================
1848 .
1849 These glyphs are intended for technical use,
1850 not for typesetting Greek language text;
1851 normally,
1852 the uppercase letters have upright shape,
1853 and the lowercase ones are slanted.
1854 .
1855 .
1856 .P
1857 .if t .ne 2v
1858 .if n .ne 3v \" account for horizontal rule
1859 .TS
1860 L L L Lx.
1861 Output Input Unicode Notes
1862 _
1863 .T&
1864 L Lf(CR) L Lx.
1865 \[*A] \e[*A] u0391 uppercase alpha +
1866 \[*B] \e[*B] u0392 uppercase beta +
1867 \[*G] \e[*G] u0393 uppercase gamma +
1868 \[*D] \e[*D] u0394 uppercase delta +
1869 \[*E] \e[*E] u0395 uppercase epsilon +
1870 \[*Z] \e[*Z] u0396 uppercase zeta +
1871 \[*Y] \e[*Y] u0397 uppercase eta +
1872 \[*H] \e[*H] u0398 uppercase theta +
1873 \[*I] \e[*I] u0399 uppercase iota +
1874 \[*K] \e[*K] u039A uppercase kappa +
1875 \[*L] \e[*L] u039B uppercase lambda +
1876 \[*M] \e[*M] u039C uppercase mu +
1877 \[*N] \e[*N] u039D uppercase nu +
1878 \[*C] \e[*C] u039E uppercase xi +
1879 \[*O] \e[*O] u039F uppercase omicron +
1880 \[*P] \e[*P] u03A0 uppercase pi +
1881 \[*R] \e[*R] u03A1 uppercase rho +
1882 \[*S] \e[*S] u03A3 uppercase sigma +
1883 \[*T] \e[*T] u03A4 uppercase tau +
1884 \[*U] \e[*U] u03A5 uppercase upsilon +
1885 \[*F] \e[*F] u03A6 uppercase phi +
1886 \[*X] \e[*X] u03A7 uppercase chi +
1887 \[*Q] \e[*Q] u03A8 uppercase psi +
1888 \[*W] \e[*W] u03A9 uppercase omega +
1889
1890 \[*a] \e[*a] u03B1 lowercase alpha +
1891 \[*b] \e[*b] u03B2 lowercase beta +
1892 \[*g] \e[*g] u03B3 lowercase gamma +
1893 \[*d] \e[*d] u03B4 lowercase delta +
1894 \[*e] \e[*e] u03B5 lowercase epsilon +
1895 \[*z] \e[*z] u03B6 lowercase zeta +
1896 \[*y] \e[*y] u03B7 lowercase eta +
1897 \[*h] \e[*h] u03B8 lowercase theta +
1898 \[*i] \e[*i] u03B9 lowercase iota +
1899 \[*k] \e[*k] u03BA lowercase kappa +
1900 \[*l] \e[*l] u03BB lowercase lambda +
1901 \[*m] \e[*m] u03BC lowercase mu +
1902 \[*n] \e[*n] u03BD lowercase nu +
1903 \[*c] \e[*c] u03BE lowercase xi +
1904 \[*o] \e[*o] u03BF lowercase omicron +
1905 \[*p] \e[*p] u03C0 lowercase pi +
1906 \[*r] \e[*r] u03C1 lowercase rho +
1907 \[*s] \e[*s] u03C3 lowercase sigma +
1908 \[*t] \e[*t] u03C4 lowercase tau +
1909 \[*u] \e[*u] u03C5 lowercase upsilon +
1910 \[*f] \e[*f] u03D5 lowercase phi +
1911 \[*x] \e[*x] u03C7 lowercase chi +
1912 \[*q] \e[*q] u03C8 lowercase psi +
1913 \[*w] \e[*w] u03C9 lowercase omega +
1914
1915 \[+e] \e[+e] u03F5 variant epsilon (lunate)
1916 \[+h] \e[+h] u03D1 variant theta (cursive form)
1917 \[+p] \e[+p] u03D6 variant pi (similar to omega)
1918 \[+f] \e[+f] u03C6 variant phi (curly shape)
1919 \[ts] \e[ts] u03C2 terminal lowercase sigma +
1920 .TE
1921 .
1922 .
1923 .br
1924 .if t .ne 4v
1925 .if n .ne 5v \" account for horizontal rule
1926 .\" ====================================================================
1927 .SS "Playing card symbols"
1928 .\" ====================================================================
1929 .
1930 .TS
1931 L L L Lx.
1932 Output Input Unicode Notes
1933 _
1934 .T&
1935 L Lf(CR) L Lx.
1936 \[CL] \e[CL] u2663 solid club suit
1937 \[SP] \e[SP] u2660 solid spade suit
1938 \[HE] \e[HE] u2665 solid heart suit
1939 \[DI] \e[DI] u2666 solid diamond suit
1940 .TE
1941 .
1942 .
1943 .\" ====================================================================
1944 .SH History
1945 .\" ====================================================================
1946 .
1947 A consideration of the typefaces originally available to AT&T
1948 .I nroff \" AT&T
1949 and
1950 .I troff \" AT&T
1951 illuminates many conventions that one might regard as idiosyncratic
1952 fifty years afterward.
1953 .
1954 (See section \[lq]History\[rq] of
1955 .MR roff 7
1956 for more context.)
1957 .
1958 The face used by the Teletype Model\~37 terminals of the Murray Hill
1959 Unix Room was based on ASCII,
1960 but assigned multiple meanings to several code points,
1961 as suggested by that standard.
1962 .
1963 Decimal 34
1964 .RB ( \[dq] )
1965 served as a dieresis accent and neutral double quotation mark;
1966 decimal 39
1967 .RB ( \[aq] )
1968 as an acute accent,
1969 apostrophe,
1970 and closing (right) single quotation mark;
1971 decimal 45
1972 .RB ( \[-] )
1973 as a hyphen and a minus sign;
1974 decimal 94
1975 .RB ( \[ha] )
1976 as a circumflex accent and caret;
1977 decimal 96
1978 .RB ( \[ga] )
1979 as a grave accent and opening (left) single quotation mark;
1980 and decimal 126
1981 .RB ( \[ti] )
1982 as a tilde accent and
1983 (with a half-line motion)
1984 swung dash.
1985 .
1986 The Model\~37 bore an optional extended character set offering upright
1987 Greek letters and several mathematical symbols;
1988 these were documented as early as the
1989 .IR kbd (VII)
1990 man page of the
1991 (First Edition)
1992 .I Unix Programmer's Manual.
1993 .
1994 .
1995 .br
1996 .ne 2v
1997 .P
1998 At the time Graphic Systems delivered the C/A/T phototypesetter to AT&T,
1999 the ASCII character set was not considered a standard basis for a glyph
2000 repertoire by traditional typographers.
2001 .
2002 In the stock Times roman,
2003 italic,
2004 and bold styles available,
2005 several ASCII characters were not present at all,
2006 nor was most of the Teletype's extended character set.
2007 .
2008 AT&T commissioned a \[lq]special\[rq] font to ensure no loss of
2009 repertoire.
2010 .
2011 .
2012 .br
2013 .ne 2v
2014 .P
2015 A representation of the coverage of the C/A/T's text fonts follows.
2016 .
2017 The glyph resembling an underscore is a baseline rule,
2018 and that resembling a vertical line is a box rule.
2019 .
2020 In italics,
2021 the box rule was not slanted.
2022 .
2023 We also observe that the hyphen and minus sign were already
2024 \[lq]de-unified\[rq] by the fonts provided;
2025 a decision whither to map an input \[lq]\-\[rq] therefore had to be
2026 taken.
2027 .
2028 .
2029 .br
2030 .if t .ne 5v
2031 .if t .ne 7v \" account for box border
2032 .P
2033 .TS
2034 center box;
2035 Lf(R).
2036 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
2037 a b c d e f g h i j k l m n o p q r s t u v w x y z
2038 0 1 2 3 4 5 6 7 8 9 \[fi] \[fl] \[Fi] \[Fl]
2039 ! $ % & ( ) \[oq] \[cq] * + \- . , / : ; = ? [ ] \[br]
2040 \[bu] \[sq] \[em] \[hy] \[ru] \[14] \[12] \[34] \
2041 \[de] \[dg] \[fm] \[ct] \[rg] \[co]
2042 .TE
2043 .
2044 .
2045 .P
2046 The special font supplied the missing ASCII and Teletype extended
2047 glyphs,
2048 among several others.
2049 .
2050 The plus,
2051 minus,
2052 and equals signs appeared in the special font despite availability in
2053 text fonts \[lq]to insulate the appearance of equations from the choice
2054 of standard [read: text] fonts\[rq]\[em]a priority since
2055 .I troff \" AT&T
2056 was turned to the task of mathematical typesetting as soon as it was
2057 developed.
2058 .
2059 .
2060 .P
2061 We note that AT&T took the opportunity to de-unify the apostrophe/right
2062 single quotation mark from the acute accent
2063 (a choice ISO later duplicated in its 8859 series of standards).
2064 .
2065 A slash intended to be mirror-symmetric with the backslash was also
2066 included,
2067 as was the Bell System logo;
2068 we do not attempt to depict the latter.
2069 .
2070 .
2071 .br
2072 .if t .ne 5v
2073 .if t .ne 7v \" account for box border
2074 .P
2075 .TS
2076 center box;
2077 Lf(I),Lf(R).
2078 \[*a] \[*b] \[*g] \[*d] \[*e] \[*z] \[*y] \[*h] \[*i] \[*k] \[*l] \
2079 \[*m] \[*n] \[*c] \[*o] \[*p] \[*r] \[*s] \[ts] \[*t] \[*u] \[*f] \
2080 \[*x] \[*q] \[*w]
2081 \[*G] \[*D] \[*H] \[*L] \[*C] \[*P] \[*S] \[*U] \[*F] \[*Q] \[*W]
2082 \[dq] \[aa] \[rs] \[ha] \[ul] \[ga] \[ti] \[sl] < > { } # @ \
2083 \[pl] \[mi] \[eq] \[**]
2084 .\" We use \[radicalex] instead of \[rn] for more reliable simulation of
2085 .\" the typeface shown in Table I of CSTR #54 (1976); see subsection
2086 .\" "Mathematical symbols" above.
2087 \[>=] \[<=] \[==] \[~=] \[ap] \[!=] \
2088 \[ua] \[da] \[<-] \[->] \[mu] \[di] \[+-] \
2089 \[if] \[pd] \[gr] \[no] \[is] \[pt] \[sr] \[radicalex] \
2090 \[cu] \[ca] \[sb] \[sp] \[ib] \[ip] \[es] \[mo]
2091 \[sc] \[dd] \[lh] \[rh] \[or] \[ci] \
2092 \[lt] \[lb] \[rt] \[rb] \[lk] \[rk] \[bv] \[lf] \[rf] \[lc] \[rc]
2093 .TE
2094 .
2095 .
2096 .P
2097 One ASCII character as rendered by the Model 37 was apparently
2098 abandoned.
2099 .
2100 That device printed decimal 124 (\[or]) as a broken vertical line,
2101 like Unicode U+00A6 (\[bb]).
2102 .
2103 No equivalent was available on the C/A/T;
2104 the box rule
2105 .BR \[rs][br] ,
2106 brace vertical extension
2107 .\" CSTR #54 (1976 edition) called this the "bold vertical", probably
2108 .\" because it was thicker than the box rule and matched the thickness
2109 .\" of the bracket pieces \(lt, \(lb, \(rt, \(rb, \(lk, \(rk, and so on.
2110 .\" Saying "bold" could be misleading because it appeared only in the
2111 .\" special font, not a bold text font.
2112 .BR \[rs][bv] ,
2113 and \[lq]or\[rq] operator
2114 .B \[rs][or]
2115 were used as contextually appropriate.
2116 .
2117 .
2118 .P
2119 .\" In the Holt, Reinhart, Winston edition of the _Unix Programmer's
2120 .\" Manual_, Revised and Expanded Version, Volume 2 (1983), the square
2121 .\" \(sq in Times bold is _not_ shown as filled on page 226.
2122 .\"
2123 .\" ...but in the AT&T USG Unix 4.0 manual (ca. 1981), typeset on the
2124 .\" Autologic APS-5, the Times bold \(sq _is_ filled.
2125 .\"
2126 .\" https://www.tuhs.org/Archive/Documentation/Manuals/Unix_4.0/
2127 .\" Volume_1/00_Annotated_Table_of_Contents.pdf
2128 .\" Volume_1/C.1.2_NROFF_TROFF_Users_Manual.pdf
2129 .\" -- GBR
2130 Devices supported by AT&T device-independent
2131 .I troff
2132 exhibited some differences in glyph detail.
2133 .
2134 For example,
2135 on the Autologic APS-5 phototypesetter,
2136 the square
2137 .B \[rs](sq
2138 became filled in the Times bold face.
2139 .
2140 .
2141 .\" ====================================================================
2142 .SH Files
2143 .\" ====================================================================
2144 .
2145 The files below are loaded automatically by the default
2146 .IR troffrc .
2147 .
2148 .
2149 .TP
2150 .I /usr/\:\%share/\:\%groff/\:\%1.23.0/\:\%tmac/\:\%composite\:.tmac
2151 assigns alternate mappings for identifiers after the first in a
2152 composite special character escape sequence.
2153 .
2154 See subsection \[lq]Accents\[rq] above.
2155 .
2156 .
2157 .TP
2158 .I /usr/\:\%share/\:\%groff/\:\%1.23.0/\:\%tmac/\:\%fallbacks\:.tmac
2159 defines fallback mappings for Unicode code points such as the increment
2160 sign (U+2206) and upper- and lowercase Roman numerals.
2161 .
2162 .
2163 .\" ====================================================================
2164 .SH Authors
2165 .\" ====================================================================
2166 .
2167 This document was written by
2168 .MT jjc@\:jclark\:.com
2169 James Clark
2170 .ME ,
2171 with additions by
2172 .MT wl@\:gnu\:.org
2173 Werner Lemberg
2174 .ME
2175 and
2176 .MT groff\-bernd\:.warken\-72@\:web\:.de
2177 Bernd Warken
2178 .ME ,
2179 revised to use
2180 .MR \%tbl 1
2181 by
2182 .MT esr@\:thyrsus\:.com
2183 Eric S.\& Raymond
2184 .ME ,
2185 and largely rewritten by
2186 .MT g.branden\:.robinson@\:gmail\:.com
2187 G.\& Branden Robinson
2188 .ME .
2189 .
2190 .
2191 .\" ====================================================================
2192 .SH "See also"
2193 .\" ====================================================================
2194 .
2195 .IR "Groff: The GNU Implementation of troff" ,
2196 by Trent A.\& Fisher and Werner Lemberg,
2197 is the primary
2198 .I groff
2199 manual.
2200 .
2201 Section \[lq]Using Symbols\[rq] may be of particular note.
2202 .
2203 You can browse it interactively with \[lq]info \[aq](groff) Using
2204 Symbols\[aq]\[rq].
2205 .
2206 .
2207 .P
2208 \[lq]An extension to the
2209 .I troff
2210 character set for Europe\[rq],
2211 E.G.\& Keizer,
2212 K.J.\& Simonsen,
2213 J.\& Akkerhuis;
2214 EUUG Newsletter,
2215 Volume 9,
2216 No.\& 2,
2217 Summer 1989
2218 .
2219 .
2220 .P
2221 .UR http://\:www\:.unicode\:.org
2222 The Unicode Standard
2223 .UE
2224 .
2225 .
2226 .br
2227 .ne 2v
2228 .P
2229 .UR https://\:www\:.aivosto\:.com/\:articles/\:charsets\-7bit\:.html
2230 \[lq]7-bit Character Sets\[rq]
2231 .UE
2232 by Tuomas Salste documents the inherent ambiguity and configurable code
2233 points of the ASCII encoding standard.
2234 .
2235 .
2236 .P
2237 \[lq]Nroff/Troff User's Manual\[rq]
2238 by Joseph F.\& Ossanna,
2239 1976,
2240 AT&T Bell Laboratories Computing Science Technical Report No.\& 54,
2241 features two tables that throw light on the glyph repertoire available
2242 to \[lq]typesetter
2243 .IR roff \[rq]
2244 when it was first written.
2245 .
2246 Be careful of re-typeset versions of this document that can be found on
2247 the Internet.
2248 .
2249 Some do not accurately represent the original document:
2250 several glyphs are obviously missing.
2251 .
2252 More subtly,
2253 lowercase Greek letters are rendered upright,
2254 not slanted as they appeared in the C/A/T's special font and as expected
2255 by
2256 .I troff \" AT&T
2257 users.
2258 .
2259 .
2260 .P
2261 .MR groff_rfc1345 7
2262 describes an alternative set of special character glyph names,
2263 which extends and in some cases overrides the definitions listed above.
2264 .
2265 .
2266 .P
2267 .MR groff 1 ,
2268 .MR troff 1 ,
2269 .MR groff 7
2270 .
2271 .
2272 .\" Restore compatibility mode (for, e.g., Solaris 10/11).
2273 .cp \n[*groff_groff_char_7_man_C]
2274 .do rr *groff_groff_char_7_man_C
2275 .
2276 .
2277 .\" Local Variables:
2278 .\" fill-column: 72
2279 .\" mode: nroff
2280 .\" tab-width: 20
2281 .\" End:
2282 .\" vim: set filetype=groff tabstop=20 textwidth=72: