Debian Codegraph @

upstream/debian-unstable/man7/groff_char.7

raw
1 '\" t
2 .TH groff_char 7 "2026-04-22" "groff 1.24.1"
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 1989-2025 Free Software Foundation, Inc.
14 .\"
15 .\" This file is part of groff, the GNU roff typesetting system.
16 .\"
17 .\" Permission is granted to copy, distribute and/or modify this
18 .\" document under the terms of the GNU Free Documentation License,
19 .\" Version 1.3 or any later version published by the Free Software
20 .\" Foundation; with no Invariant Sections, with no Front-Cover Texts,
21 .\" and with no Back-Cover Texts.
22 .\"
23 .\" A copy of the Free Documentation License is included as a file
24 .\" called FDL in the main directory of the groff source package.
25 .
26 .
27 .\" Save and disable compatibility mode (for, e.g., Solaris 10/11).
28 .do nr *groff_groff_char_7_man_C \n[.cp]
29 .cp 0
30 .
31 .\" Define fallback for groff 1.23's MR macro if the system lacks it.
32 .nr do-fallback 0
33 .if !\n(.f .nr do-fallback 1 \" mandoc
34 .if \n(.g .if !d MR .nr do-fallback 1 \" older groff
35 .if !\n(.g .nr do-fallback 1 \" non-groff *roff
36 .if \n[do-fallback] \{\
37 . de MR
38 . ie \\n(.$=1 \
39 . I \%\\$1
40 . el \
41 . IR \%\\$1 (\\$2)\\$3
42 . .
43 .\}
44 .rr do-fallback
45 .
46 .\" It is important for this page not to tell lies beloved of
47 .\" indifferent man page authors. See, for example,
48 .\" <https://lwn.net/Articles/947941/>.
49 .tr ''--^^``~~
50 .do char ' '
51 .do char - -
52 .do char ^ ^
53 .do char ` `
54 .do char ~ ~
55 .
56 .
57 .\" ====================================================================
58 .SH Description
59 .\" ====================================================================
60 .
61 The GNU
62 .I roff
63 typesetting system has a large glyph repertoire suitable for production
64 of varied literary,
65 professional,
66 technical,
67 and mathematical documents.
68 .
69 .I groff
70 works with
71 .I characters;
72 an output device renders
73 .I glyphs.
74 .
75 .IR groff 's
76 input character set is restricted to that defined by the ISO\~Latin-1
77 (ISO\~8859-1)
78 standard.
79 .
80 For ease of document maintenance in UTF-8 environments,
81 it is advisable to use only the Unicode basic Latin code points;
82 these correspond to ISO\~646:1991\~IRV (US-ASCII),
83 a subset of all of the foregoing
84 which has only 94 visible,
85 printable code points.
86 .\" In groff, 0x20 SP is mapped to a space node, not a glyph node, and
87 .\" all kinds of special behavior attaches to such nodes, so we count
88 .\" only to 94 and not 95 as is often done in other ASCII contexts.
89 In
90 .I groff,
91 these are termed
92 .I "ordinary characters."
93 .
94 Often,
95 many more are desired in output.
96 .
97 .
98 .P
99 AT&T
100 .I troff
101 in the 1970s faced a similar problem:
102 the available typesetter's glyph repertoire differed from that of the
103 computers that controlled it.
104 .
105 .IR troff 's \" AT&T
106 solution was a form of escape sequence known as a
107 .I special character
108 to access several dozen additional glyphs available in the fonts
109 prepared for mounting in the phototypesetter.
110 .
111 These glyphs were mapped onto a two-character name space for a degree
112 of mnemonic convenience;
113 for example,
114 the escape sequence
115 .B \e(aa
116 encoded an acute accent and
117 .B \e(sc
118 a section sign.
119 .
120 .
121 .P
122 .I groff
123 has lifted historical
124 .I roff
125 limitations on special character name lengths,
126 but recognizes and retains compatibility with the historical names.
127 .
128 .I groff
129 expands the lexicon of glyphs available by name and permits users to
130 define their own special character escape sequences with the
131 .B char
132 request.
133 .
134 Special character names are
135 .I groff
136 identifiers;
137 see section \[lq]Identifiers\[rq] in
138 .MR groff 7 .
139 .
140 Our discussion uses the terms \[lq]glyph name\[rq] and \[lq]special
141 character name\[rq] interchangeably;
142 we assume no character translations or redefinitions.
143 .
144 .
145 .P
146 This document lists all of the glyph names predefined by
147 .IR groff 's
148 font description files and presents the systematic notation by which it
149 enables access to arbitrary Unicode code points and construction of
150 composite glyphs.
151 .
152 Glyphs listed may be unavailable,
153 or may vary in appearance,
154 depending on the output device and font chosen when the page was
155 formatted.
156 .
157 This page was rendered for device
158 .B \*[.T]
159 primarily using font
160 .BR \n[.fn] .
161 .
162 .
163 .P
164 A few escape sequences that are not
165 .I groff
166 special characters also produce glyphs;
167 these exist for syntactical or historical reasons.
168 .
169 .BR \e\[aq] ,
170 .BR \e\[ga] ,
171 .BR \e\- ,
172 and
173 .B \e_
174 are translated on input to the special character escape sequences
175 .BR \e[aa] ,
176 .BR \e[ga] ,
177 .BR \e[\-] ,
178 and
179 .BR \e[ul] ,
180 respectively.
181 .
182 Others include
183 .BR \e\e ,
184 .B \e.\&
185 (backslash-dot),
186 and
187 .BR \ee ;
188 see
189 .MR groff 7 .
190 .
191 A small number of special characters represent glyphs that are not
192 encoded in Unicode;
193 examples include the baseline rule
194 .B \e[ru]
195 and the Bell System logo
196 .B \e[bs].
197 .
198 .
199 .P
200 In
201 .IR groff ,
202 test support for any character
203 (ordinary or special)
204 with the conditional expression operator
205 .RB \[lq] c \[rq].
206 .
207 .RS
208 .\" https://www.bell-labs.com/usr/dmr/www/ ("In 1984, ...")
209 .EX
210 \&.ie c \e[bs] \e{Welcome to the \e[bs] Bell System;
211 did you get the Wehrmacht helmet or the Death Star?\e}
212 \&.el No Bell System logo.
213 .EE
214 .RE
215 .
216 .
217 .P
218 While
219 .I groff
220 accepts eight-bit encoded input,
221 not all such code points are valid as input.
222 .
223 .\" src/libs/libgroff/invalid.cpp
224 Character codes
225 0,
226 11,
227 13\[en]31,
228 and
229 128\[en]159
230 are invalid.
231 .
232 (This is all C0 and C1 controls except for
233 SOH through LF
234 [Control+A to Control+J],
235 and FF
236 [Control+L].)
237 .
238 Some of these code points are used by
239 .I groff
240 for internal purposes,
241 which is one reason it does not support UTF-8 natively.
242 .
243 .
244 .\" ====================================================================
245 .SS "Fundamental character set"
246 .\" ====================================================================
247 .
248 The ordinary characters catalogued above,
249 plus the space,
250 tab,
251 newline,
252 and leader (Control+A),
253 form the fundamental character
254 set for
255 .I groff
256 input;
257 anything in the language,
258 even over one million code points in Unicode,
259 can be expressed using it.
260 .
261 Code points in the range 33\[en]126 comprise a common set of
262 printable glyphs in all of the aforementioned ISO character encoding
263 standards.
264 .
265 It is this character set and
266 (with some noteworthy exceptions)
267 the corresponding glyph repertoire for which AT&T
268 .I troff
269 was implemented.
270 .\" From this point, do not talk about numerical character assignments.
271 .
272 .
273 .P
274 All of the following characters map to glyphs as you would expect.
275 .
276 .TS
277 center box;
278 Lf(CR).
279 ! # $ % & ( ) * + , . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
280 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 [ ] _
281 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 { | }
282 .TE
283 .\" The bottom border of that box is practically kissin' the tittles.
284 .if t .sp 0.2v
285 .
286 The remaining ordinary characters surprise computing professionals and
287 others intimately familiar with the ISO character encodings.
288 .
289 The developers of AT&T
290 .I troff
291 chose mappings for them that would be useful for typesetting technical
292 literature in a broad range of scientific disciplines:
293 Bell Labs used the system to prepare AT&T's patent filings with the
294 U.S.\& government.
295 .
296 Further,
297 the prevailing character encoding standard in the 1970s,
298 USAS\~X3.4-1968 (ASCII),
299 deliberately supported semantic ambiguity at some code points,
300 and outright substitution at several others,
301 to suit the localization demands of various national standards bodies.
302 .
303 .
304 .P
305 The table below presents the seven exceptional code points
306 with their typical keycap engravings,
307 their glyph mappings and semantics in
308 .I roff
309 systems,
310 and the escape sequences producing the Unicode basic Latin character
311 they replace.
312 .
313 The first,
314 the neutral double quote,
315 is a partial exception because it does represent itself,
316 but since the
317 .I roff
318 language also uses it to quote macro arguments,
319 .I groff
320 supports a special character escape sequence as an alternative form so
321 that the glyph can be easily included in macro arguments without
322 requiring the user to master the quoting rules that AT&T
323 .I troff
324 required in that context.
325 .
326 (Some requests,
327 like
328 .BR ds ,
329 also treat
330 .B \[dq]
331 non-literally at the beginning of an argument.)
332 .
333 Furthermore,
334 not all of the special character escape sequences are portable to AT&T
335 .I troff
336 and all of its descendants;
337 these
338 .I groff
339 extensions are presented using its special character form
340 .BR \[rs][] ,
341 whereas portable special character escape sequences are shown in the
342 traditional
343 .B \[rs](
344 form.
345 .
346 .B \[rs]\-
347 and
348 .B \[rs]e
349 are portable to all known
350 .IR troff s.
351 .
352 .B \[rs]e
353 means \[lq]the glyph of the current escape character\[rq];
354 it therefore can produce unexpected output if the
355 .B ec
356 request is used.
357 .
358 On devices with a limited glyph repertoire,
359 glyphs in the \[lq]keycap\[rq] and \[lq]appearance\[rq] columns on the
360 same row of the table may look identical;
361 except for the neutral double quote,
362 this will
363 .I not
364 be the case on more-capable devices.
365 .
366 Review your document using as many different output devices as possible.
367 .
368 .
369 .P
370 .TS
371 center box;
372 L L L.
373 Keycap Appearance and meaning Special character and meaning
374 _
375 " " neutral double quote \f[CR]\[rs][dq]\f[] neutral double quote
376 \[aq] \[cq] closing single quote \f[CR]\[rs][aq]\f[] neutral apostrophe
377 \- - hyphen \f[CR]\[rs]\-\f[] or \f[CR]\[rs][\-]\f[] minus sign/Unix dash
378 \[rs] (escape character) \f[CR]\[rs]e\f[] or \f[CR]\[rs][rs]\f[] reverse solidus
379 \[ha] \[u02C6] modifier circumflex \f[CR]\[rs][ha]\f[] circumflex/caret/\[lq]hat\[rq]
380 \[ga] \[oq] opening single quote \f[CR]\[rs](ga\f[] grave accent
381 \[ti] \[u02DC] modifier tilde \f[CR]\[rs][ti]\f[] tilde
382 .TE
383 .
384 .
385 .P
386 The hyphen-minus is a particularly unfortunate case of overloading.
387 .
388 Its awkward name in ISO\~8859 and later standards reflects the many
389 distinguishable purposes to which it had already been put by the 1980s,
390 including
391 a hyphen,
392 a minus sign,
393 and
394 (alone or in repetition)
395 dashes of varying widths.
396 .
397 For best results in
398 .I roff
399 systems,
400 use the
401 .RB \[lq] \- \[rq]
402 character in input outside an escape sequence
403 .I only
404 to mean a hyphen,
405 as in the phrase \[lq]long-term\[rq].
406 .
407 For a minus sign in running text or a Unix file name or command-line
408 option dash,
409 use
410 .B \[rs]\-
411 (or
412 .B \[rs][\-]
413 in
414 .I groff
415 if you find it helps the clarity of the source document).
416 .
417 (Another minus sign,
418 for use in mathematical expressions,
419 is available as
420 .BR \[rs](mi .)
421 .
422 AT&T
423 .I troff
424 supported em-dashes as
425 .BR \[rs](em ,
426 as does
427 .IR groff .
428 .
429 .
430 .P
431 The special character escape sequence for the apostrophe as a neutral
432 single quote is typically needed only in technical content;
433 typing words like \[lq]can't\[rq] and \[lq]Anne's\[rq] in a natural way
434 will render correctly,
435 because in ordinary prose an apostrophe is typeset either as a closing
436 single quotation mark or as a neutral single quote,
437 depending on the capabilities of the output device.
438 .
439 By contrast,
440 special character escape sequences should be used for quotation marks
441 unless portability to limited or historical
442 .I troff
443 implementations is necessary;
444 on those systems,
445 the input convention is to pair the grave accent with the apostrophe for
446 single quotes,
447 and to double both characters for double quotes.
448 .
449 AT&T
450 .I troff
451 defined no special characters for quotation marks or the apostrophe.
452 .
453 Repeated single quotes
454 (\[oq]\[oq]thus\[cq]\[cq])
455 will be visually distinguishable from double quotes
456 (\[lq]thus\[rq])
457 on terminal devices,
458 and perhaps on others
459 (depending on the font selected and its kerning configuration).
460 .
461 .TS
462 tab(@) center box;
463 L L.
464 AT&T \f[I]troff\f[] input@recommended \f[I]groff\f[] input
465 _
466 .T&
467 Lf(CR) Lf(CR).
468 A Winter\[aq]s Tale@A Winter\[aq]s Tale
469 \[ga]U.K.\& outer quotes\[aq]@\[rs][oq]U.K.\& outer quotes\[rs][cq]
470 \[ga]U.K.\& \[ga]\[ga]inner\[aq]\[aq] quotes\[aq]\
471 @\[rs][oq]U.K.\& \[rs][lq]inner\[rs][rq] quotes\[rs][cq]
472 \[ga]\[ga]U.S.\& outer quotes\[aq]\[aq]\
473 @\[rs][lq]U.S.\& outer quotes\[rs][rq]
474 \[ga]\[ga]U.S.\& \[ga]inner\[aq] quotes\[aq]\[aq]\
475 @\[rs][lq]U.S.\& \[rs][oq]inner\[rs][cq] quotes\[rs][rq]
476 .TE
477 .\" Keep bottom border of box from sitting on the ascenders below.
478 .if t .sp 0.2v
479 .
480 If you frequently require quotation marks in your document,
481 see if the macro package you're using supplies strings or macros to
482 facilitate quotation,
483 or define them yourself
484 (except in man pages).
485 .
486 .
487 .P
488 Using Unicode basic Latin characters to compose boxes and lines is
489 ill-advised.
490 .
491 .I roff
492 systems have dedicated features for drawing horizontal and vertical
493 lines,
494 the
495 .BR \[rs]l ,
496 .BR \[rs]L ,
497 and
498 .B \[rs]D
499 escape sequences.
500 .
501 Also see subsection \[lq]Rules and lines\[rq] below.
502 .
503 Preprocessors like
504 .MR \%tbl 1
505 and
506 .MR \%pic 1
507 draw boxes and will produce the best possible output for the device,
508 falling back to basic Latin glyphs only when necessary.
509 .
510 .
511 .\" ====================================================================
512 .SS "Eight-bit encodings and Latin-1 supplement"
513 .\" ====================================================================
514 .
515 ISO 646 is a seven-bit code encoding 128 code points;
516 eight-bit codes are twice the size.
517 .
518 ISO\~Latin-1 (8859-1) allocated the additional space to what
519 Unicode calls \[lq]C1 controls\[rq]
520 (control characters)
521 and the \[lq]Latin-1 supplement\[rq].
522 .
523 The C1 controls are neither printable nor usable as
524 GNU
525 .I troff \" GNU
526 input.
527 .
528 .
529 .P
530 GNU
531 .I troff \" GNU
532 handles two Latin-1 supplement characters specially,
533 and never produces them as output.
534 .
535 .
536 .TP
537 NBSP
538 encodes a no-break space;
539 it maps to
540 .BR \[rs]\[ti] ,
541 the adjustable non-breaking space escape sequence.
542 .
543 .
544 .TP
545 SHY
546 encodes a soft hyphen;
547 it maps to
548 .BR \[rs]% ,
549 the hyphenation control escape sequence.
550 .
551 .
552 .P
553 The remaining characters in the Latin-1 supplement represent
554 themselves.
555 .
556 Although they can be specified directly with the keyboard on systems
557 configured to use Latin-1 as the character encoding,
558 it is more portable,
559 both to other
560 .I roff
561 systems and to UTF-8 environments,
562 to use their special character escape sequences,
563 shown below.
564 .
565 The glyph descriptions we use are non-standard in some cases,
566 for brevity.
567 .
568 .
569 .P
570 .TS
571 L2 Lf(CR)1 L L2 Lf(CR)1 L.
572 \[r!] \e[r!] inverted exclamation mark \[~N] \e[\[ti]N] N tilde
573 \[ct] \e[ct] cent sign \[`O] \e[\[ga]O] O grave
574 \[Po] \e[Po] pound sign \['O] \e[\[aq]O] O acute
575 \[Cs] \e[Cs] currency sign \[^O] \e[\[ha]O] O circumflex
576 \[Ye] \e[Ye] yen sign \[~O] \e[\[ti]O] O tilde
577 \[bb] \e[bb] broken bar \[:O] \e[:O] O dieresis
578 \[sc] \e[sc] section sign \[mu] \e[mu] multiplication sign
579 \[ad] \e[ad] dieresis accent \[/O] \e[/O] O slash
580 \[co] \e[co] copyright sign \[`U] \e[\[ga]U] U grave
581 \[Of] \e[Of] feminine ordinal indicator \['U] \e[\[aq]U] U acute
582 \[Fo] \e[Fo] left double chevron \[^U] \e[\[ha]U] U circumflex
583 \[no] \e[no] logical not \[:U] \e[:U] U dieresis
584 \[rg] \e[rg] registered sign \['Y] \e[\[aq]Y] Y acute
585 \[a-] \e[a\-] macron accent \[TP] \e[TP] uppercase thorn
586 \[de] \e[de] degree sign \[ss] \e[ss] lowercase sharp s
587 \[+-] \e[+\-] plus-minus \[`a] \e[\[ga]a] a grave
588 \[S2] \e[S2] superscript two \['a] \e[\[aq]a] a acute
589 \[S3] \e[S3] superscript three \[^a] \e[\[ha]a] a circumflex
590 \[aa] \e[aa] acute accent \[~a] \e[\[ti]a] a tilde
591 \[mc] \e[mc] micro sign \[:a] \e[:a] a dieresis
592 \[ps] \e[ps] pilcrow sign \[oa] \e[oa] a ring
593 \[pc] \e[pc] centered period \[ae] \e[ae] ae ligature
594 \[ac] \e[ac] cedilla accent \[,c] \e[,c] c cedilla
595 \[S1] \e[S1] superscript one \[`e] \e[\[ga]e] e grave
596 \[Om] \e[Om] masculine ordinal indicator \['e] \e[\[aq]e] e acute
597 \[Fc] \e[Fc] right double chevron \[^e] \e[\[ha]e] e circumflex
598 \[14] \e[14] one quarter symbol \[:e] \e[:e] e dieresis
599 \[12] \e[12] one half symbol \[`i] \e[\[ga]i] i grave
600 \[34] \e[34] three quarters symbol \['i] \e[\[aq]i] e acute
601 \[r?] \e[r?] inverted question mark \[^i] \e[\[ha]i] i circumflex
602 \[`A] \e[\[ga]A] A grave \[:i] \e[:i] i dieresis
603 \['A] \e[\[aq]A] A acute \[Sd] \e[Sd] lowercase eth
604 \[^A] \e[\[ha]A] A circumflex \[~n] \e[\[ti]n] n tilde
605 \[~A] \e[\[ti]A] A tilde \[`o] \e[\[ga]o] o grave
606 \[:A] \e[:A] A dieresis \['o] \e[\[aq]o] o acute
607 \[oA] \e[oA] A ring \[^o] \e[\[ha]o] o circumflex
608 \[AE] \e[AE] AE ligature \[~o] \e[\[ti]o] o tilde
609 \[,C] \e[,C] C cedilla \[:o] \e[:o] o dieresis
610 \[`E] \e[\[ga]E] E grave \[di] \e[di] division sign
611 \['E] \e[\[aq]E] E acute \[/o] \e[/o] o slash
612 \[^E] \e[\[ha]E] E circumflex \[`u] \e[\[ga]u] u grave
613 \[:E] \e[:E] E dieresis \['u] \e[\[aq]u] u acute
614 \[`I] \e[\[ga]I] I grave \[^u] \e[\[ha]u] u circumflex
615 \['I] \e[\[aq]I] I acute \[:u] \e[:u] u dieresis
616 \[^I] \e[\[ha]I] I circumflex \['y] \e[\[aq]y] y acute
617 \[:I] \e[:I] I dieresis \[Tp] \e[Tp] lowercase thorn
618 \[-D] \e[\-D] uppercase eth \[:y] \e[:y] y dieresis
619 .TE
620 .
621 .
622 .\" ====================================================================
623 .SS "Special character escape forms"
624 .\" ====================================================================
625 .
626 Glyphs that lack a character code in the basic Latin repertoire to
627 directly represent them are entered by one of several special character
628 escape forms.
629 .
630 Such glyphs can be simple or composite,
631 and accessed either by name or numerically by code point.
632 .
633 Code points and combining properties are determined by character
634 encoding standards,
635 whereas glyph names as used here originated in AT&T
636 .I troff \" AT&T
637 special character escape sequences.
638 .
639 Predefined glyph names use only characters in the basic Latin
640 repertoire.
641 .
642 .
643 .TP
644 .BI \[rs]( gl
645 is a special character escape sequence for the glyph with the
646 two-character name
647 .IR gl .
648 .
649 This is the original syntax form supported by AT&T
650 .IR troff .
651 .
652 The acute accent,
653 .BR \[rs](aa ,
654 is an example.
655 .
656 .
657 .TP
658 .BI \[rs]C\[aq] glyph-name \[aq]
659 is a special character escape sequence for
660 .IR glyph-name ,
661 which can be of arbitrary length.
662 .
663 The delimiter,
664 shown here as a neutral apostrophe,
665 can be any character not occurring in
666 .IR glyph-name .
667 .
668 This syntax form was introduced in later versions of AT&T
669 device-independent
670 .IR troff . \" AT&T
671 .
672 The foregoing acute accent example can be expressed
673 as
674 .BR \[rs]C\[aq]aa\[aq] .
675 .
676 .
677 .TP
678 .BI \[rs][ glyph-name ]
679 is a special character escape sequence for
680 .IR glyph-name ,
681 which can be of arbitrary length but must not contain a closing square
682 bracket
683 .RB \[lq] ] \[rq].
684 .
685 (No glyph names predefined by
686 .I groff
687 employ
688 .RB \[lq] ] \[rq].)
689 .
690 The foregoing acute accent example can be expressed in
691 in GNU
692 .I troff \" GNU
693 as
694 .BR \[rs][aa] .
695 .
696 .
697 .P
698 .BI \[rs]C\[aq] c \[aq]
699 and
700 .BI \[rs][ c ]
701 are not synonyms for the ordinary character
702 .RI \[lq] c \[rq],
703 but request the special character named
704 .RB \[lq] \[rs] \c
705 .IR c \[rq].
706 .
707 For example,
708 .RB \[lq] \[rs][a] \[rq]
709 is not \[lq]a\[rq],
710 but rather a special character with the internal glyph name
711 (used in font description files and diagnostic messages)
712 .BR \[rs]a ,
713 which is typically undefined.
714 .
715 The only such glyph name
716 .I groff
717 predefines is the minus sign,
718 which can therefore be accessed as
719 .B \[rs]C\[aq]\-\[aq]
720 or
721 .BR \[rs][\-] .
722 .
723 .
724 .TP
725 .BI \[rs][ "base-char composite-1 composite-2"\~\c
726 \&.\|.\|.\~\c
727 .IB composite-n ]
728 is a composite glyph.
729 .
730 Glyphs like a lowercase \[lq]e\[rq] with an acute accent,
731 as in the word \[lq]caf\[e aa]\[rq],
732 can be expressed as
733 .BR "\[rs][e aa]" .
734 .
735 .
736 .P
737 Normally,
738 the formatter advances the drawing position after setting a special
739 character,
740 as it does for ordinary ones.
741 .
742 .IR groff 's
743 .B \%composite
744 request designates a special character as combining,
745 suppressing advancement.
746 .
747 .
748 .P
749 You can obtain a report of mappings defined by the
750 .B \%composite
751 request on the standard error stream with the
752 .B \%pcomposite
753 request.
754 .
755 The
756 .I \%composite.tmac
757 macro file,
758 loaded automatically by the default
759 .IR \%troffrc ,
760 maps certain special characters to combining characters as shown
761 in subsection \[lq]Accents\[rq] below.
762 .
763 .
764 .P
765 Unicode encodes far more characters than
766 .I groff
767 has names for;
768 special character escape forms based on numerical code points enable
769 access to any of them.
770 .
771 Frequently used glyphs or glyph combinations can be stored in strings,
772 and new glyph names can be created
773 .I "ad hoc"
774 with the
775 .B char
776 request;
777 see
778 .MR groff 7 .
779 .
780 .
781 .TP
782 .BI \[rs][u nnnn\c
783 .RI [ n\c
784 .RI [ n ]]\c
785 .B ]
786 is a Unicode numeric special character escape sequence.
787 .
788 Any Unicode code point can be accessed with four to six hexadecimal
789 digits,
790 with hexadecimal letters accepted in uppercase form only.
791 .
792 Thus,
793 .B \[rs][u02DA]
794 accesses the (spacing) ring accent,
795 producing \[lq]\^\[u02DA]\^\[rq].
796 .
797 .
798 .\" Use "GNU troff" in this paragraph because the contrast with AT&T
799 .\" troff, which antedated Unicode, is important, and that contrast is
800 .\" obscured with the default empty command prefix on "troff".
801 .br
802 .ne 2v
803 .P
804 Unicode code points can be composed as well;
805 when they are,
806 GNU
807 .I troff \" GNU
808 requires NFD
809 (Normalization Form D),
810 where all Unicode glyphs are maximally decomposed.
811 .
812 (Exception:
813 precomposed characters in the Latin-1 supplement described above are
814 also accepted.
815 .
816 Do not count on this exception remaining in a future
817 GNU
818 .I troff \" GNU
819 that accepts UTF-8 input directly.)
820 .
821 .
822 Thus,
823 GNU
824 .I troff \" GNU
825 accepts
826 .RB \[lq]caf \[rs][\[aq]e] \[rq],
827 .RB \[lq]caf \[rs][e\~aa] \[rq],
828 and
829 .RB \[lq]caf \[rs][u0065_0301] \[rq],
830 as ways to input \[lq]caf\['e]\[rq].
831 .
832 (Due to its legacy 8-bit encoding compatibility,
833 at present it also accepts
834 .RB \[lq]caf \[rs][u00E9] \[rq]
835 on ISO\~Latin-1 systems.)
836 .
837 .
838 .TP
839 .BI \[rs][u base-char\c
840 [\c
841 .BI _ combining-component\c
842 ].\|.\|.]
843 constructs a composite glyph from Unicode numeric special character
844 escape sequences.
845 .
846 The code points of the base glyph and the combining components are each
847 expressed in hexadecimal,
848 with an underscore
849 .RB ( _ )
850 separating each component.
851 .
852 Thus,
853 .B \[rs][u006E_0303]
854 produces \[lq]\[u006E_0303]\[rq].
855 .
856 .
857 .TP
858 .BI \[rs][char nnn ]
859 expresses an eight-bit code point where
860 .I nnn
861 is the code point of the character,
862 a decimal number between 0 and\~255
863 without leading zeroes.
864 .
865 This legacy numeric special character escape sequence is used to map
866 characters onto glyphs via the
867 .B trin
868 request in macro files loaded by
869 .MR grotty 1 .
870 .
871 .
872 .\" ====================================================================
873 .SH "Glyph tables"
874 .\" ====================================================================
875 .
876 In this section,
877 .IR groff 's
878 glyph name repertoire is presented in tabular form.
879 .
880 The meanings of the columns are as follows.
881 .
882 .
883 .TP 9n \" "Unicode" + 2n
884 .B Output
885 shows the glyph as it appears on the device used to render this
886 document;
887 although it can have a notably different shape on other devices
888 (and is subject to user-directed translation and replacement),
889 .I groff
890 attempts reasonable equivalency on all output devices.
891 .
892 .
893 .TP
894 .B Input
895 shows the
896 .I groff
897 character
898 (ordinary or special)
899 that normally produces the glyph.
900 .
901 Some code points have multiple glyph names.
902 .
903 .
904 .TP
905 .B Unicode
906 is the code point notation for the glyph or combining glyph sequence as
907 described in subsection \[lq]Special character escape forms\[rq] above.
908 .
909 It corresponds to the standard notation for Unicode short identifiers
910 such that
911 .IR groff 's
912 .BI u nnnn
913 is equivalent to Unicode's
914 .RI U+ nnnn .
915 .\" And thereby hangs a tale...
916 .\" https://unicode.org/mail-arch/unicode-ml/y2005-m11/0060.html
917 .
918 .
919 .TP
920 .B Notes
921 describes the glyph,
922 elucidating the mnemonic value of the glyph name where possible.
923 .
924 .
925 .IP
926 A plus sign \[lq]+\[rq] indicates that the glyph name appears in the
927 AT&T
928 .I troff
929 user's manual,
930 CSTR\~#54
931 (1992 revision).
932 .
933 When using the AT&T special character syntax
934 .BI \[rs]( xx\c
935 , widespread portability can be expected from such names.
936 .
937 .
938 .IP
939 Entries marked with \[lq]***\[rq] denote glyphs used for mathematical
940 purposes.
941 .
942 On typesetting devices,
943 such glyphs are typically drawn from a
944 .I special
945 font
946 (see
947 .MR groff_font 5 ).
948 .
949 Often,
950 such glyphs lack bold or italic style forms or have metrics that look
951 incongruous in ordinary prose.
952 .
953 A few that are not uncommon in running text have \[lq]text
954 variants\[rq],
955 which should work better in that context.
956 .
957 Conversely,
958 a handful of glyphs that are normally drawn from a text font may be
959 required in mathematical expressions.
960 .
961 Both sets of exceptions are noted in the tables where they appear
962 (\[lq]Logical symbols\[rq] and \[lq]Mathematical symbols\[rq]).
963 .
964 .
965 .\" ====================================================================
966 .SS "Basic Latin"
967 .\" ====================================================================
968 .
969 Apart from basic Latin characters with special mappings,
970 described in subsection \[lq]Fundamental character set\[rq] above,
971 a few others in that range have special character glyph names.
972 .
973 .\" XXX: I surmise that...
974 These were defined for ease of input on non-U.S.\& keyboards lacking
975 keycaps for them,
976 or for symmetry with other special character glyph names serving a
977 similar purpose.
978 .
979 .
980 .P
981 The vertical bar is overloaded;
982 the
983 .B \[rs][ba]
984 and
985 .B \[rs][or]
986 escape sequences may render differently.
987 .
988 See subsection \[lq]Mathematical symbols\[rq] below
989 for unstyled variants
990 of the plus,
991 minus,
992 and equals
993 signs normally drawn from this range.
994 .
995 .
996 .br
997 .ne 5v
998 .P
999 .TS
1000 L L L Lx.
1001 Output Input Unicode Notes
1002 _
1003 .T&
1004 L Lf(CR) L Lx.
1005 \[dq] \e[dq] u0022 neutral double quote
1006 \[sh] \e[sh] u0023 number sign
1007 \[Do] \e[Do] u0024 dollar sign
1008 \[aq] \e[aq] u0027 apostrophe, neutral single quote
1009 \[sl] \e[sl] u002F slash, solidus +
1010 \[at] \e[at] u0040 at sign
1011 \[lB] \e[lB] u005B left square bracket
1012 \[rs] \e[rs] u005C reverse solidus
1013 \[rB] \e[rB] u005D right square bracket
1014 \[ha] \e[ha] u005E circumflex, caret, \[lq]hat\[rq]
1015 \[lC] \e[lC] u007B left brace
1016 | | u007C bar +
1017 \[ba] \e[ba] u007C bar
1018 \[or] \e[or] u007C bitwise or +
1019 \[rC] \e[rC] u007D right brace
1020 \[ti] \e[ti] u007E tilde
1021 .TE
1022 .
1023 .
1024 .\" ====================================================================
1025 .SS "Supplementary Latin letters"
1026 .\" ====================================================================
1027 .
1028 Historically,
1029 .B \[rs][ss]
1030 developed from a ligature of \[lq]sz\[rq].
1031 .\" ...though the s was a "long s" and the z was a "Blackletter z".
1032 .
1033 An uppercase form is available as
1034 .BR \[rs][u1E9E] ,
1035 but in the German language it is of specialized use;
1036 \[ss] does
1037 .I not
1038 normally uppercase-transform to it,
1039 but rather to \[lq]SS\[rq].
1040 .
1041 \[lq]Lowercase f with hook\[rq] is also used as a function symbol;
1042 see subsection \[lq]Mathematical symbols\[rq] below.
1043 .
1044 .
1045 .P
1046 .if t .ne 2v
1047 .if n .ne 3v \" account for horizontal rule
1048 .TS
1049 L L L Lx.
1050 Output Input Unicode Notes
1051 _
1052 .T&
1053 L Lf(CR) L Lx.
1054 \[-D] \e[\-D] u00D0 uppercase eth
1055 \[Sd] \e[Sd] u00F0 lowercase eth
1056 \[TP] \e[TP] u00DE uppercase thorn
1057 \[Tp] \e[Tp] u00FE lowercase thorn
1058 \[ss] \e[ss] u00DF lowercase sharp s
1059 \[.i] \e[.i] u0131 i without tittle
1060 \[.j] \e[.j] u0237 j without tittle
1061 \[Fn] \e[Fn] u0192 lowercase f with hook, function
1062 \[/L] \e[/L] u0141 L with stroke
1063 \[/l] \e[/l] u0142 l with stroke
1064 \[/O] \e[/O] u00D8 O with stroke
1065 \[/o] \e[/o] u00F8 o with stroke
1066 .TE
1067 .
1068 .
1069 .br
1070 .if t .ne 4v
1071 .if n .ne 5v \" account for horizontal rule
1072 .\" ====================================================================
1073 .SS "Ligatures and digraphs"
1074 .\" ====================================================================
1075 .
1076 .TS
1077 L L L Lx.
1078 Output Input Unicode Notes
1079 _
1080 .T&
1081 L Lf(CR) L Lx.
1082 \[ff] \e[ff] u0066_0066 ff ligature +
1083 \[fi] \e[fi] u0066_0069 fi ligature +
1084 \[fl] \e[fl] u0066_006C fl ligature +
1085 \[Fi] \e[Fi] u0066_0066_0069 ffi ligature +
1086 \[Fl] \e[Fl] u0066_0066_006C ffl ligature +
1087 \[AE] \e[AE] u00C6 AE ligature
1088 \[ae] \e[ae] u00E6 ae ligature
1089 \[OE] \e[OE] u0152 OE ligature
1090 \[oe] \e[oe] u0153 oe ligature
1091 \[IJ] \e[IJ] u0132 IJ digraph
1092 \[ij] \e[ij] u0133 ij digraph
1093 .TE
1094 .
1095 .
1096 .\" ====================================================================
1097 .SS Accents
1098 .\" ====================================================================
1099 .
1100 The non-combining code point in parentheses is used when the special
1101 character occurs in isolation
1102 (compare
1103 .RB \[lq] "caf\[rs][e aa]" \[rq]
1104 and
1105 .RB \[lq] "caf\[rs][aa]e" \[rq]).
1106 .
1107 .
1108 .P
1109 .if t .ne 3v
1110 .if n .ne 4v \" account for horizontal rule
1111 .TS
1112 L L L Lx.
1113 Output Input Unicode Notes
1114 _
1115 .T&
1116 L Lf(CR) L Lx.
1117 \[a"] \e[a"] u030B (u02DD) double acute accent
1118 \[a-] \e[a\-] u0304 (u00AF) macron accent
1119 \[a.] \e[a.] u0307 (u02D9) dot accent
1120 \[a^] \e[a\[ha]] u0302 (u005E) circumflex accent
1121 \[aa] \e[aa] u0301 (u00B4) acute accent +
1122 \[ga] \e[ga] u0300 (u0060) grave accent +
1123 \[ab] \e[ab] u0306 (u02D8) breve accent
1124 \[ac] \e[ac] u0327 (u00B8) cedilla accent
1125 \[ad] \e[ad] u0308 (u00A8) dieresis accent
1126 \[ah] \e[ah] u030C (u02C7) caron accent
1127 \[ao] \e[ao] u030A (u02DA) ring accent
1128 \[a~] \e[a\[ti]] u0303 (u007E) tilde accent
1129 \[ho] \e[ho] u0328 (u02DB) hook accent
1130 .TE
1131 .
1132 .
1133 .\" ====================================================================
1134 .SS "Accented characters"
1135 .\" ====================================================================
1136 .
1137 All of these glyphs can be composed using combining glyph names as
1138 described in subsection \[lq]Special character escape forms\[rq] above;
1139 the names below are short aliases for convenience.
1140 .
1141 .
1142 .P
1143 .if t .ne 2v
1144 .if n .ne 3v \" account for horizontal rule
1145 .TS
1146 L L L Lx.
1147 Output Input Unicode Notes
1148 _
1149 .T&
1150 L Lf(CR) L Lx.
1151 \['A] \e[\[aq]A] u0041_0301 A acute
1152 \['C] \e[\[aq]C] u0043_0301 C acute
1153 \['E] \e[\[aq]E] u0045_0301 E acute
1154 \['I] \e[\[aq]I] u0049_0301 I acute
1155 \['O] \e[\[aq]O] u004F_0301 O acute
1156 \['U] \e[\[aq]U] u0055_0301 U acute
1157 \['Y] \e[\[aq]Y] u0059_0301 Y acute
1158 \['a] \e[\[aq]a] u0061_0301 a acute
1159 \['c] \e[\[aq]c] u0063_0301 c acute
1160 \['e] \e[\[aq]e] u0065_0301 e acute
1161 \['i] \e[\[aq]i] u0069_0301 i acute
1162 \['o] \e[\[aq]o] u006F_0301 o acute
1163 \['u] \e[\[aq]u] u0075_0301 u acute
1164 \['y] \e[\[aq]y] u0079_0301 y acute
1165
1166 \[:A] \e[:A] u0041_0308 A dieresis
1167 \[:E] \e[:E] u0045_0308 E dieresis
1168 \[:I] \e[:I] u0049_0308 I dieresis
1169 \[:O] \e[:O] u004F_0308 O dieresis
1170 \[:U] \e[:U] u0055_0308 U dieresis
1171 \[:Y] \e[:Y] u0059_0308 Y dieresis
1172 \[:a] \e[:a] u0061_0308 a dieresis
1173 \[:e] \e[:e] u0065_0308 e dieresis
1174 \[:i] \e[:i] u0069_0308 i dieresis
1175 \[:o] \e[:o] u006F_0308 o dieresis
1176 \[:u] \e[:u] u0075_0308 u dieresis
1177 \[:y] \e[:y] u0079_0308 y dieresis
1178
1179 \[^A] \e[\[ha]A] u0041_0302 A circumflex
1180 \[^E] \e[\[ha]E] u0045_0302 E circumflex
1181 \[^I] \e[\[ha]I] u0049_0302 I circumflex
1182 \[^O] \e[\[ha]O] u004F_0302 O circumflex
1183 \[^U] \e[\[ha]U] u0055_0302 U circumflex
1184 \[^a] \e[\[ha]a] u0061_0302 a circumflex
1185 \[^e] \e[\[ha]e] u0065_0302 e circumflex
1186 \[^i] \e[\[ha]i] u0069_0302 i circumflex
1187 \[^o] \e[\[ha]o] u006F_0302 o circumflex
1188 \[^u] \e[\[ha]u] u0075_0302 u circumflex
1189
1190 \[`A] \e[\[ga]A] u0041_0300 A grave
1191 \[`E] \e[\[ga]E] u0045_0300 E grave
1192 \[`I] \e[\[ga]I] u0049_0300 I grave
1193 \[`O] \e[\[ga]O] u004F_0300 O grave
1194 \[`U] \e[\[ga]U] u0055_0300 U grave
1195 \[`a] \e[\[ga]a] u0061_0300 a grave
1196 \[`e] \e[\[ga]e] u0065_0300 e grave
1197 \[`i] \e[\[ga]i] u0069_0300 i grave
1198 \[`o] \e[\[ga]o] u006F_0300 o grave
1199 \[`u] \e[\[ga]u] u0075_0300 u grave
1200
1201 \[~A] \e[\[ti]A] u0041_0303 A tilde
1202 \[~N] \e[\[ti]N] u004E_0303 N tilde
1203 \[~O] \e[\[ti]O] u004F_0303 O tilde
1204 \[~a] \e[\[ti]a] u0061_0303 a tilde
1205 \[~n] \e[\[ti]n] u006E_0303 n tilde
1206 \[~o] \e[\[ti]o] u006F_0303 o tilde
1207
1208 \[vS] \e[vS] u0053_030C S caron
1209 \[vs] \e[vs] u0073_030C s caron
1210 \[vZ] \e[vZ] u005A_030C Z caron
1211 \[vz] \e[vz] u007A_030C z caron
1212
1213 \[,C] \e[,C] u0043_0327 C cedilla
1214 \[,c] \e[,c] u0063_0327 c cedilla
1215
1216 \[oA] \e[oA] u0041_030A A ring
1217 \[oa] \e[oa] u0061_030A a ring
1218 .TE
1219 .
1220 .
1221 .\" ====================================================================
1222 .SS "Quotation marks"
1223 .\" ====================================================================
1224 .
1225 The neutral double quote,
1226 useful in documenting programming languages,
1227 is also available as a special character for convenient embedding in
1228 macro arguments;
1229 see subsection \[lq]Fundamental character set\[rq] above.
1230 .
1231 .
1232 .P
1233 .if t .ne 2v
1234 .if n .ne 3v \" account for horizontal rule
1235 .TS
1236 L L L Lx.
1237 Output Input Unicode Notes
1238 _
1239 .T&
1240 L Lf(CR) L Lx.
1241 \[Bq] \e[Bq] u201E low double comma quote
1242 \[bq] \e[bq] u201A low single comma quote
1243 \[lq] \e[lq] u201C left double quote
1244 \[rq] \e[rq] u201D right double quote
1245 \[oq] \e[oq] u2018 single opening (left) quote
1246 \[cq] \e[cq] u2019 single closing (right) quote
1247 \[aq] \e[aq] u0027 apostrophe, neutral single quote
1248 \[dq] " u0022 neutral double quote
1249 \[dq] \e[dq] u0022 neutral double quote
1250 \[Fo] \e[Fo] u00AB left double chevron
1251 \[Fc] \e[Fc] u00BB right double chevron
1252 \[fo] \e[fo] u2039 left single chevron
1253 \[fc] \e[fc] u203A right single chevron
1254 .TE
1255 .
1256 .
1257 .\" ====================================================================
1258 .SS Punctuation
1259 .\" ====================================================================
1260 .
1261 The Unicode name for U+00B7 is \[lq]middle dot\[rq],
1262 which is unfortunately confusable with the
1263 .I groff
1264 mnemonic for the visually similar but semantically distinct
1265 multiplication dot;
1266 see subsection \[lq]Mathematical symbols\[rq] below.
1267 .
1268 .
1269 .P
1270 .if t .ne 2v
1271 .if n .ne 3v \" account for horizontal rule
1272 .TS
1273 L L L Lx.
1274 Output Input Unicode Notes
1275 _
1276 .T&
1277 L Lf(CR) L Lx.
1278 \[r!] \e[r!] u00A1 inverted exclamation mark
1279 \[r?] \e[r?] u00BF inverted question mark
1280 \[pc] \e[pc] u00B7 centered period
1281 \[em] \e[em] u2014 em-dash +
1282 \[en] \e[en] u2013 en-dash
1283 - \- u2010 hyphen +
1284 \[hy] \e[hy] u2010 hyphen +
1285 .TE
1286 .
1287 .
1288 .\" ====================================================================
1289 .SS Brackets
1290 .\" ====================================================================
1291 .
1292 On typesetting devices,
1293 the bracket extensions are font-invariant glyphs;
1294 that is,
1295 they are rendered the same way regardless of font
1296 (with a drawing escape sequence).
1297 .
1298 On terminals,
1299 they are
1300 .I not
1301 font-invariant;
1302 .I groff
1303 maps them arbitrarily to U+23AA
1304 (\[lq]curly bracket extension\[rq]).
1305 .
1306 In AT&T
1307 .IR troff ,
1308 only one glyph was available to vertically extend
1309 brackets,
1310 braces,
1311 and
1312 parentheses:
1313 .BR \[rs](bv .
1314 .
1315 .
1316 .
1317 .P
1318 Not all devices supply bracket pieces that can be piled up with
1319 .B \[rs]b
1320 due to the restrictions of the formatter's piling algorithm.
1321 .
1322 The following macro offers a more general bracket-building solution.
1323 .
1324 .
1325 .RS
1326 .EX
1327 \&.\e" Make a pile centered vertically 0.5em above the baseline.
1328 \&.\e" The first argument is placed at the top.
1329 \&.\e" The pile is returned in string \[aq]pile\[aq].
1330 \&.eo
1331 \&.de pile\-make
1332 \&.\& nr pile\-wd 0
1333 \&.\& nr pile\-ht 0
1334 \&.\& ds pile\-args
1335 \&.\&
1336 \&.\& nr pile\-# \en[.$]
1337 \&.\& while \en[pile\-#] \e{\e
1338 \&.\& nr pile\-wd (\en[pile\-wd] >? \ew\[aq]\e$[\en[pile\-#]]\[aq])
1339 \&.\& nr pile\-ht +(\en[rst] \- \en[rsb])
1340 \&.\& as pile\-args \ev\[aq]\en[rsb]u\[aq]\e"
1341 \&.\& as pile\-args \eZ\[aq]\e$[\en[pile\-#]]\[aq]\e"
1342 \&.\& as pile\-args \ev\[aq]\-\en[rst]u\[aq]\e"
1343 \&.\& nr pile\-# \-1
1344 \&.\& \e}
1345 \&.\&
1346 \&.\& ds pile \ev\[aq](\-0.5m + (\en[pile\-ht]u / 2u))\[aq]\e"
1347 \&.\& as pile \e*[pile\-args]\e"
1348 \&.\& as pile \ev\[aq]((\en[pile\-ht]u / 2u) + 0.5m)\[aq]\e"
1349 \&.\& as pile \eh\[aq]\en[pile\-wd]u\[aq]\e"
1350 \&..
1351 \&.ec
1352 .EE
1353 .RE
1354 .
1355 .
1356 .P
1357 Further complicating matters is that some glyphs representing bracket
1358 pieces in AT&T
1359 .I troff
1360 can be used for other mathematical symbols as well;
1361 for example,
1362 .B \[rs](lf
1363 and
1364 .B \[rs](rf
1365 provide the floor operator.
1366 .
1367 Some output devices,
1368 such as
1369 .BR dvi ,
1370 don't unify such glyphs.
1371 .
1372 For this reason,
1373 the glyphs
1374 .BR \[rs][lf] ,
1375 .BR \[rs][rf] ,
1376 .BR \[rs][lc] ,
1377 and
1378 .B \[rs][rc]
1379 are not unified with similar-looking bracket pieces.
1380 .
1381 In
1382 .IR groff ,
1383 only glyphs with long names are guaranteed to pile up correctly for all
1384 devices\[em]provided those glyphs are available.
1385 .
1386 .
1387 .P
1388 .if t .ne 2v
1389 .if n .ne 3v \" account for horizontal rule
1390 .TS
1391 L L L Lx.
1392 Output Input Unicode Notes
1393 _
1394 .T&
1395 L Lf(CR) L Lx.
1396 \[lB] [ u005B left square bracket
1397 \[lB] \e[lB] u005B left square bracket
1398 \[rB] ] u005D right square bracket
1399 \[rB] \e[rB] u005D right square bracket
1400 \[lC] { u007B left brace
1401 \[lC] \e[lC] u007B left brace
1402 \[rC] } u007D right brace
1403 \[rC] \e[rC] u007D right brace
1404 \[la] \e[la] u27E8 left angle bracket
1405 \[ra] \e[ra] u27E9 right angle bracket
1406 \[bv] \e[bv] u23AA brace vertical extension + ***
1407 \[braceex] \e[braceex] u23AA brace vertical extension
1408
1409 \[bracketlefttp] \e[bracketlefttp] u23A1 left square bracket top
1410 \[bracketleftex] \e[bracketleftex] u23A2 left square bracket extension
1411 \[bracketleftbt] \e[bracketleftbt] u23A3 left square bracket bottom
1412
1413 \[bracketrighttp] \e[bracketrighttp] u23A4 right square bracket top
1414 \[bracketrightex] \e[bracketrightex] u23A5 right square bracket extension
1415 \[bracketrightbt] \e[bracketrightbt] u23A6 right square bracket bottom
1416
1417 \[lt] \e[lt] u23A7 left brace top +
1418 \[lk] \e[lk] u23A8 left brace middle +
1419 \[lb] \e[lb] u23A9 left brace bottom +
1420 \[bracelefttp] \e[bracelefttp] u23A7 left brace top
1421 \[braceleftmid] \e[braceleftmid] u23A8 left brace middle
1422 \[braceleftbt] \e[braceleftbt] u23A9 left brace bottom
1423 \[braceleftex] \e[braceleftex] u23AA left brace extension
1424
1425 \[rt] \e[rt] u23AB right brace top +
1426 \[rk] \e[rk] u23AC right brace middle +
1427 \[rb] \e[rb] u23AD right brace bottom +
1428 \[bracerighttp] \e[bracerighttp] u23AB right brace top
1429 \[bracerightmid] \e[bracerightmid] u23AC right brace middle
1430 \[bracerightbt] \e[bracerightbt] u23AD right brace bottom
1431 \[bracerightex] \e[bracerightex] u23AA right brace extension
1432
1433 \[parenlefttp] \e[parenlefttp] u239B left parenthesis top
1434 \[parenleftex] \e[parenleftex] u239C left parenthesis extension
1435 \[parenleftbt] \e[parenleftbt] u239D left parenthesis bottom
1436 \[parenrighttp] \e[parenrighttp] u239E right parenthesis top
1437 \[parenrightex] \e[parenrightex] u239F right parenthesis extension
1438 \[parenrightbt] \e[parenrightbt] u23A0 right parenthesis bottom
1439 .TE
1440 .
1441 .
1442 .br
1443 .if t .ne 4v
1444 .if n .ne 5v \" account for horizontal rule
1445 .\" ====================================================================
1446 .SS Arrows
1447 .\" ====================================================================
1448 .
1449 .TS
1450 L L L Lx.
1451 Output Input Unicode Notes
1452 _
1453 .T&
1454 L Lf(CR) L Lx.
1455 \[<-] \e[<\-] u2190 horizontal arrow left +
1456 \[->] \e[\->] u2192 horizontal arrow right +
1457 \[<>] \e[<>] u2194 bidirectional horizontal arrow
1458 \[da] \e[da] u2193 vertical arrow down +
1459 \[ua] \e[ua] u2191 vertical arrow up +
1460 \[va] \e[va] u2195 bidirectional vertical arrow
1461 \[lA] \e[lA] u21D0 horizontal double arrow left
1462 \[rA] \e[rA] u21D2 horizontal double arrow right
1463 \[hA] \e[hA] u21D4 bidirectional horizontal double arrow
1464 \[dA] \e[dA] u21D3 vertical double arrow down
1465 \[uA] \e[uA] u21D1 vertical double arrow up
1466 \[vA] \e[vA] u21D5 bidirectional vertical double arrow
1467 \[an] \e[an] u23AF horizontal arrow extension
1468 .TE
1469 .
1470 .
1471 .\" ====================================================================
1472 .SS "Rules and lines"
1473 .\" ====================================================================
1474 .
1475 On typesetting devices,
1476 the font-invariant glyphs
1477 (see subsection \[lq]Brackets\[rq] above)
1478 .BR \[rs][br] ,
1479 .BR \[rs][ul] ,
1480 and
1481 .B \[rs][rn]
1482 form corners when adjacent;
1483 they can be used to build boxes.
1484 .
1485 On terminal devices,
1486 they are mapped as shown in the table.
1487 .
1488 The Unicode-derived names of these three glyphs are approximations.
1489 .
1490 .
1491 .P
1492 The ordinary character
1493 .B _
1494 accesses the underscore glyph in a font;
1495 .\" unless one isn't available, but this seems to be only a theoretical
1496 .\" concern--what font doesn't support every ASCII codepoint these days?
1497 .BR \[rs][ul] ,
1498 by contrast,
1499 may be font-invariant on typesetting devices.
1500 .
1501 .
1502 .P
1503 The baseline rule
1504 .B \[rs][ru]
1505 is a font-invariant glyph,
1506 namely a rule of one-half em.
1507 .
1508 .
1509 .P
1510 In AT&T
1511 .IR troff , \" AT&T
1512 .B \[rs][rn]
1513 also served as a one\~en extension of the square root symbol.
1514 .
1515 .I groff
1516 favors
1517 .B \[rs][radicalex]
1518 for this purpose;
1519 see subsection \[lq]Mathematical symbols\[rq] below.
1520 .
1521 .
1522 .P
1523 .if t .ne 2v
1524 .if n .ne 3v \" account for horizontal rule
1525 .TS
1526 L L L Lx.
1527 Output Input Unicode Notes
1528 _
1529 .T&
1530 L Lf(CR) L Lx.
1531 | | u007C bar
1532 \[ba] \e[ba] u007C bar
1533 \[br] \e[br] u2502 box rule +
1534 \&_ \&_ u005F underscore, low line +
1535 \[ul] \e[ul] --- underrule +
1536 \[rn] \e[rn] u203E overline +
1537 \[ru] \e[ru] --- baseline rule +
1538 \[bb] \e[bb] u00A6 broken bar
1539 \[sl] / u002F slash, solidus +
1540 \[sl] \e[sl] u002F slash, solidus +
1541 \[rs] \e[rs] u005C reverse solidus
1542 .TE
1543 .
1544 .
1545 .br
1546 .if t .ne 4v
1547 .if n .ne 5v \" account for horizontal rule
1548 .\" ====================================================================
1549 .SS "Text markers"
1550 .\" ====================================================================
1551 .
1552 .TS
1553 L L L Lx.
1554 Output Input Unicode Notes
1555 _
1556 .T&
1557 L Lf(CR) L Lx.
1558 \[ci] \e[ci] u25CB circle +
1559 \[bu] \e[bu] u2022 bullet +
1560 \[dg] \e[dg] u2020 dagger +
1561 \[dd] \e[dd] u2021 double dagger +
1562 \[lz] \e[lz] u25CA lozenge, diamond
1563 \[sq] \e[sq] u25A1 square +
1564 \[ps] \e[ps] u00B6 pilcrow sign
1565 \[sc] \e[sc] u00A7 section sign +
1566 \[lh] \e[lh] u261C hand pointing left +
1567 \[rh] \e[rh] u261E hand pointing right +
1568 \[at] @ u0040 at sign
1569 \[at] \e[at] u0040 at sign
1570 \[sh] # u0023 number sign
1571 \[sh] \e[sh] u0023 number sign
1572 \[CR] \e[CR] u21B5 carriage return
1573 \[OK] \e[OK] u2713 check mark
1574 .TE
1575 .
1576 .\" ====================================================================
1577 .SS "Legal symbols"
1578 .\" ====================================================================
1579 .
1580 The Bell System logo is not supported in
1581 .IR groff .
1582 .
1583 .
1584 .P
1585 .if t .ne 2v
1586 .if n .ne 3v \" account for horizontal rule
1587 .TS
1588 L L L Lx.
1589 Output Input Unicode Notes
1590 _
1591 .T&
1592 L Lf(CR) L Lx.
1593 \[co] \e[co] u00A9 copyright sign +
1594 \[rg] \e[rg] u00AE registered sign +
1595 \[tm] \e[tm] u2122 trade mark sign
1596 \[bs] \e[bs] --- Bell System logo +
1597 .TE
1598 .
1599 .
1600 .br
1601 .if t .ne 4v
1602 .if n .ne 5v \" account for horizontal rule
1603 .\" ====================================================================
1604 .SS "Currency symbols"
1605 .\" ====================================================================
1606 .
1607 .TS
1608 L L L Lx.
1609 Output Input Unicode Notes
1610 _
1611 .T&
1612 L Lf(CR) L Lx.
1613 \[Do] $ u0024 dollar sign
1614 \[Do] \e[Do] u0024 dollar sign
1615 \[ct] \e[ct] u00A2 cent sign +
1616 \[eu] \e[eu] u20AC Euro sign
1617 \[Eu] \e[Eu] u20AC variant Euro sign
1618 \[Ye] \e[Ye] u00A5 yen sign
1619 \[Po] \e[Po] u00A3 pound sign
1620 \[Cs] \e[Cs] u00A4 currency sign
1621 .TE
1622 .
1623 .
1624 .br
1625 .if t .ne 4v
1626 .if n .ne 5v \" account for horizontal rule
1627 .\" ====================================================================
1628 .SS Units
1629 .\" ====================================================================
1630 .
1631 .TS
1632 L L L Lx.
1633 Output Input Unicode Notes
1634 _
1635 .T&
1636 L Lf(CR) L Lx.
1637 \[de] \e[de] u00B0 degree sign +
1638 \[%0] \e[%0] u2030 per thousand, per mille sign
1639 \[fm] \e[fm] u2032 arc minute sign, foot mark +
1640 \[sd] \e[sd] u2033 arc second sign
1641 \[mc] \e[mc] u00B5 micro sign
1642 \[Of] \e[Of] u00AA feminine ordinal indicator
1643 \[Om] \e[Om] u00BA masculine ordinal indicator
1644 .TE
1645 .
1646 .
1647 .\" ====================================================================
1648 .SS "Logical symbols"
1649 .\" ====================================================================
1650 .
1651 The variants of the not sign may differ in appearance or spacing
1652 depending on the device and font selected.
1653 .
1654 Unicode does not encode a discrete \[lq]bitwise or\[rq] sign:
1655 on typesetting devices,
1656 it is drawn shorter than the bar,
1657 about the same height as a capital letter.
1658 .
1659 Terminal devices unify
1660 .B \[rs][ba]
1661 and
1662 .BR \[rs][or] .
1663 .
1664 .
1665 .P
1666 .if t .ne 2v
1667 .if n .ne 3v \" account for horizontal rule
1668 .TS
1669 L L L Lx.
1670 Output Input Unicode Notes
1671 _
1672 .T&
1673 L Lf(CR) L Lx.
1674 \[AN] \e[AN] u2227 logical and
1675 \[OR] \e[OR] u2228 logical or
1676 \[no] \e[no] u00AC logical not + ***
1677 \[tno] \e[tno] u00AC text variant of \f[B]\e[no]\f[]
1678 \[te] \e[te] u2203 there exists
1679 \[fa] \e[fa] u2200 for all
1680 \[st] \e[st] u220B such that
1681 \[3d] \e[3d] u2234 therefore
1682 \[tf] \e[tf] u2234 therefore
1683 | | u007C bar +
1684 \[or] \e[or] u007C bitwise or +
1685 .TE
1686 .
1687 .
1688 .\" ====================================================================
1689 .SS "Mathematical symbols"
1690 .\" ====================================================================
1691 .
1692 .B \[rs][Fn]
1693 also appears in subsection \[lq]Supplementary Latin letters\[rq] above.
1694 .
1695 The
1696 plus-minus,
1697 multiplication,
1698 and division signs,
1699 .BR \[rs][+\-] ,
1700 .BR \[rs][mu] ,
1701 and
1702 .BR \[rs][di] ,
1703 are normally drawn from the special font,
1704 but have text font variants.
1705 .
1706 The plus,
1707 minus,
1708 and equals signs
1709 are normally drawn from text fonts,
1710 but have special font variants.
1711 .
1712 Variants may differ in appearance or spacing depending on the device and
1713 font selected.
1714 .
1715 .
1716 .P
1717 In AT&T
1718 .IR troff ,
1719 .B \[rs](rn
1720 (\[lq]root en extender\[rq])
1721 served as the horizontal extension of the radical
1722 (square root)
1723 sign,
1724 .BR \[rs](sr ,
1725 and was drawn at the maximum height of the typeface's bounding box,
1726 enabling it to double as an overline
1727 (see subsection \[lq]Rules and lines\[rq] above).
1728 .
1729 A contemporary font's radical sign might not ascend to such an extreme.
1730 .
1731 In
1732 .IR groff ,
1733 you can instead use
1734 .B \[rs][radicalex]
1735 to continue the radical sign
1736 .BR \[rs][sr] ;
1737 these special characters are intended for use with text fonts.
1738 .
1739 .B \[rs][sqrt]
1740 and
1741 .B \[rs][sqrtex]
1742 are their unstyled counterparts.
1743 .
1744 .
1745 .P
1746 .if t .ne 2v
1747 .if n .ne 3v \" account for horizontal rule
1748 .TS
1749 L L L Lx.
1750 Output Input Unicode Notes
1751 _
1752 .T&
1753 L Lf(CR) L Lx.
1754 \[12] \e[12] u00BD one half symbol +
1755 \[14] \e[14] u00BC one quarter symbol +
1756 \[34] \e[34] u00BE three quarters symbol +
1757 \[18] \e[18] u215B one eighth symbol
1758 \[38] \e[38] u215C three eighths symbol
1759 \[58] \e[58] u215D five eighths symbol
1760 \[78] \e[78] u215E seven eighths symbol
1761 \[S1] \e[S1] u00B9 superscript one
1762 \[S2] \e[S2] u00B2 superscript two
1763 \[S3] \e[S3] u00B3 superscript three
1764
1765 + + u002B plus
1766 \[pl] \e[pl] u002B special variant of plus + ***
1767 \- \e[\-] u002D minus
1768 \[mi] \e[mi] u2212 special variant of minus + ***
1769 \[-+] \e[\-+] u2213 minus-plus
1770 \[+-] \e[+\-] u00B1 plus-minus + ***
1771 \[t+-] \e[t+\-] u00B1 text variant of \f[B]\e[+\-]\f[]
1772 \[md] \e[md] u22C5 multiplication dot
1773 \[mu] \e[mu] u00D7 multiplication sign + ***
1774 \[tmu] \e[tmu] u00D7 text variant of \f[B]\e[mu]\f[]
1775 \[c*] \e[c*] u2297 circled times
1776 \[c+] \e[c+] u2295 circled plus
1777 \[di] \e[di] u00F7 division sign, obelus + ***
1778 \[tdi] \e[tdi] u00F7 text variant of \f[B]\e[di]\f[]
1779 \[f/] \e[f/] u2044 fraction slash
1780 * * u002A asterisk
1781 \[**] \e[**] u2217 mathematical asterisk +
1782
1783 \[<=] \e[<=] u2264 less than or equal to +
1784 \[>=] \e[>=] u2265 greater than or equal to +
1785 \[<<] \e[<<] u226A much less than
1786 \[>>] \e[>>] u226B much greater than
1787 \&= \&= u003D equals
1788 \[eq] \e[eq] u003D special variant of equals + ***
1789 \[!=] \e[!=] u003D_0338 not equals +
1790 \[==] \e[==] u2261 equivalent +
1791 \[ne] \e[ne] u2261_0338 not equivalent
1792 \[=~] \e[=\[ti]] u2245 approximately equal to
1793 \[|=] \e[|=] u2243 asymptotically equal to +
1794 \[ti] \e[ti] u007E tilde +
1795 \[ap] \e[ap] u223C similar to, tilde operator +
1796 \[~~] \e[\[ti]\[ti]] u2248 almost equal to
1797 \[~=] \e[\[ti]=] u2248 almost equal to
1798 \[pt] \e[pt] u221D proportional to +
1799
1800 \[es] \e[es] u2205 empty set +
1801 \[mo] \e[mo] u2208 element of a set +
1802 \[nm] \e[nm] u2208_0338 not element of set
1803 \[sb] \e[sb] u2282 proper subset +
1804 \[nb] \e[nb] u2282_0338 not subset
1805 \[sp] \e[sp] u2283 proper superset +
1806 \[nc] \e[nc] u2283_0338 not superset
1807 \[ib] \e[ib] u2286 subset or equal +
1808 \[ip] \e[ip] u2287 superset or equal +
1809 \[ca] \e[ca] u2229 intersection, cap +
1810 \[cu] \e[cu] u222A union, cup +
1811
1812 \[/_] \e[/_] u2220 angle
1813 \[pp] \e[pp] u22A5 perpendicular
1814 \[is] \e[is] u222B integral +
1815 \[integral] \e[integral] u222B integral ***
1816 \[sum] \e[sum] u2211 summation ***
1817 \[product] \e[product] u220F product ***
1818 \[coproduct] \e[coproduct] u2210 coproduct ***
1819 \[gr] \e[gr] u2207 gradient +
1820 \[sr] \e[sr] u221A radical sign, square root +
1821 \[rn] \e[rn] u203E overline +
1822 \[radicalex] \e[radicalex] --- radical extension
1823 \[sqrt] \e[sqrt] u221A radical sign, square root ***
1824 \[sqrtex] \e[sqrtex] --- radical extension ***
1825
1826 \[lc] \e[lc] u2308 left ceiling +
1827 \[rc] \e[rc] u2309 right ceiling +
1828 \[lf] \e[lf] u230A left floor +
1829 \[rf] \e[rf] u230B right floor +
1830
1831 \[if] \e[if] u221E infinity +
1832 \[Ah] \e[Ah] u2135 aleph symbol
1833 \[Fn] \e[Fn] u0192 lowercase f with hook, function
1834 \[Im] \e[Im] u2111 blackletter I, imaginary part
1835 \[Re] \e[Re] u211C blackletter R, real part
1836 \[wp] \e[wp] u2118 Weierstrass p
1837 \[pd] \e[pd] u2202 partial differential
1838 \[-h] \e[\-h] u210F h bar
1839 \[hbar] \e[hbar] u210F h bar
1840 .TE
1841 .
1842 .
1843 .\" ====================================================================
1844 .SS "Greek glyphs"
1845 .\" ====================================================================
1846 .
1847 These glyphs are intended for technical use,
1848 not for typesetting Greek language text;
1849 normally,
1850 the uppercase letters have upright shape,
1851 and the lowercase ones are slanted.
1852 For the latter,
1853 we include in parentheses under \[lq]Unicode\[rq]
1854 more appropriate code points
1855 from the Mathematical Alphanumeric Symbols block
1856 of the Supplementary Multilingual Plane.
1857 .
1858 .
1859 .P
1860 .if t .ne 2v
1861 .if n .ne 3v \" account for horizontal rule
1862 .TS
1863 L L L Lx.
1864 Output Input Unicode Notes
1865 _
1866 .T&
1867 L Lf(CR) L Lx.
1868 \[*A] \e[*A] u0391 uppercase alpha +
1869 \[*B] \e[*B] u0392 uppercase beta +
1870 \[*G] \e[*G] u0393 uppercase gamma +
1871 \[*D] \e[*D] u0394 uppercase delta +
1872 \[*E] \e[*E] u0395 uppercase epsilon +
1873 \[*Z] \e[*Z] u0396 uppercase zeta +
1874 \[*Y] \e[*Y] u0397 uppercase eta +
1875 \[*H] \e[*H] u0398 uppercase theta +
1876 \[*I] \e[*I] u0399 uppercase iota +
1877 \[*K] \e[*K] u039A uppercase kappa +
1878 \[*L] \e[*L] u039B uppercase lambda +
1879 \[*M] \e[*M] u039C uppercase mu +
1880 \[*N] \e[*N] u039D uppercase nu +
1881 \[*C] \e[*C] u039E uppercase xi +
1882 \[*O] \e[*O] u039F uppercase omicron +
1883 \[*P] \e[*P] u03A0 uppercase pi +
1884 \[*R] \e[*R] u03A1 uppercase rho +
1885 \[*S] \e[*S] u03A3 uppercase sigma +
1886 \[*T] \e[*T] u03A4 uppercase tau +
1887 \[*U] \e[*U] u03A5 uppercase upsilon +
1888 \[*F] \e[*F] u03A6 uppercase phi +
1889 \[*X] \e[*X] u03A7 uppercase chi +
1890 \[*Q] \e[*Q] u03A8 uppercase psi +
1891 \[*W] \e[*W] u03A9 uppercase omega +
1892
1893 \[*a] \e[*a] u03B1 (u1D6FC) lowercase alpha +
1894 \[*b] \e[*b] u03B2 (u1D6FD) lowercase beta +
1895 \[*g] \e[*g] u03B3 (u1D6FE) lowercase gamma +
1896 \[*d] \e[*d] u03B4 (u1D6FF) lowercase delta +
1897 \[*e] \e[*e] u03B5 (u1D700) lowercase epsilon +
1898 \[*z] \e[*z] u03B6 (u1D701) lowercase zeta +
1899 \[*y] \e[*y] u03B7 (u1D702) lowercase eta +
1900 \[*h] \e[*h] u03B8 (u1D703) lowercase theta +
1901 \[*i] \e[*i] u03B9 (u1D704) lowercase iota +
1902 \[*k] \e[*k] u03BA (u1D705) lowercase kappa +
1903 \[*l] \e[*l] u03BB (u1D706) lowercase lambda +
1904 \[*m] \e[*m] u03BC (u1D707) lowercase mu +
1905 \[*n] \e[*n] u03BD (u1D708) lowercase nu +
1906 \[*c] \e[*c] u03BE (u1D709) lowercase xi +
1907 \[*o] \e[*o] u03BF (u1D70A) lowercase omicron +
1908 \[*p] \e[*p] u03C0 (u1D70B) lowercase pi +
1909 \[*r] \e[*r] u03C1 (u1D70C) lowercase rho +
1910 \[*s] \e[*s] u03C3 (u1D70E) lowercase sigma +
1911 \[*t] \e[*t] u03C4 (u1D70F) lowercase tau +
1912 \[*u] \e[*u] u03C5 (u1D710) lowercase upsilon +
1913 \[*f] \e[*f] u03D5 (u1D717) lowercase phi +
1914 \[*x] \e[*x] u03C7 (u1D712) lowercase chi +
1915 \[*q] \e[*q] u03C8 (u1D713) lowercase psi +
1916 \[*w] \e[*w] u03C9 (u1D714) lowercase omega +
1917
1918 \[+e] \e[+e] u03F5 (u1D716) variant epsilon (lunate)
1919 \[+h] \e[+h] u03D1 (u1D717) variant theta (cursive form)
1920 \[+p] \e[+p] u03D6 (u1D717) variant pi (similar to omega)
1921 \[+f] \e[+f] u03C6 (u1D71B) variant phi (curly shape)
1922 \[ts] \e[ts] u03C2 (u1D70D) terminal lowercase sigma +
1923 .TE
1924 .
1925 .
1926 .br
1927 .if t .ne 4v
1928 .if n .ne 5v \" account for horizontal rule
1929 .\" ====================================================================
1930 .SS "Playing card symbols"
1931 .\" ====================================================================
1932 .
1933 .TS
1934 L L L Lx.
1935 Output Input Unicode Notes
1936 _
1937 .T&
1938 L Lf(CR) L Lx.
1939 \[CL] \e[CL] u2663 solid club suit
1940 \[SP] \e[SP] u2660 solid spade suit
1941 \[HE] \e[HE] u2665 solid heart suit
1942 \[DI] \e[DI] u2666 solid diamond suit
1943 .TE
1944 .
1945 .
1946 .br
1947 .ne 4v
1948 .\" ====================================================================
1949 .SH History
1950 .\" ====================================================================
1951 .
1952 A consideration of the typefaces originally available to AT&T
1953 .I nroff \" AT&T
1954 and
1955 .I troff \" AT&T
1956 illuminates many conventions that one might regard as idiosyncratic
1957 fifty years afterward.
1958 .
1959 (See section \[lq]History\[rq] of
1960 .MR roff 7
1961 for more context.)
1962 .
1963 The face used by the Teletype Model\~37 terminals of the Murray Hill
1964 Unix Room was based on ASCII,
1965 but assigned multiple meanings to several code points,
1966 as suggested by that standard.
1967 .
1968 Decimal 34
1969 .RB ( \[dq] )
1970 served as a dieresis accent and neutral double quotation mark;
1971 decimal 39
1972 .RB ( \[aq] )
1973 as an acute accent,
1974 apostrophe,
1975 and closing (right) single quotation mark;
1976 decimal 45
1977 .RB ( \[-] )
1978 as a hyphen and a minus sign;
1979 decimal 94
1980 .RB ( \[ha] )
1981 as a circumflex accent and caret;
1982 decimal 96
1983 .RB ( \[ga] )
1984 as a grave accent and opening (left) single quotation mark;
1985 and decimal 126
1986 .RB ( \[ti] )
1987 as a tilde accent and
1988 (with a half-line motion)
1989 swung dash.
1990 .
1991 The Model\~37 bore an optional extended character set offering upright
1992 Greek letters and several mathematical symbols;
1993 these were documented as early as the
1994 .IR kbd (VII)
1995 man page of the
1996 (First Edition)
1997 .I Unix Programmer's Manual.
1998 .
1999 .
2000 .br
2001 .ne 2v
2002 .P
2003 At the time Graphic Systems delivered the C/A/T phototypesetter to AT&T,
2004 the ASCII character set was not considered a standard basis for a glyph
2005 repertoire by traditional typographers.
2006 .
2007 In the stock Times roman,
2008 italic,
2009 and bold styles available,
2010 several ASCII characters were not present at all,
2011 nor was most of the Teletype's extended character set.
2012 .
2013 AT&T commissioned a \[lq]special\[rq] font to retain their accustomed
2014 repertoire.
2015 .
2016 .
2017 .br
2018 .ne 2v
2019 .P
2020 A representation of the coverage of the C/A/T's text fonts follows.
2021 .
2022 The glyph resembling an underscore is a baseline rule,
2023 and that resembling a vertical line is a box rule.
2024 .
2025 In italics,
2026 the box rule was not slanted.
2027 .
2028 We also observe that the hyphen and minus sign were already
2029 \[lq]de-unified\[rq] by the fonts provided;
2030 a decision whither to map an input \[lq]\-\[rq] therefore had to be
2031 taken.
2032 .
2033 .
2034 .br
2035 .if t .ne 5v
2036 .if t .ne 7v \" account for box border
2037 .P
2038 .TS
2039 center box;
2040 Lf(R).
2041 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
2042 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
2043 0 1 2 3 4 5 6 7 8 9 \[fi] \[fl] \[Fi] \[Fl]
2044 ! $ % & ( ) \[oq] \[cq] * + \- . , / : ; = ? [ ] \[br]
2045 \[bu] \[sq] \[em] \[hy] \[ru] \[14] \[12] \[34] \
2046 \[de] \[dg] \[fm] \[ct] \[rg] \[co]
2047 .TE
2048 .
2049 .
2050 .P
2051 The special font supplied the missing ASCII and Teletype extended
2052 glyphs,
2053 among several others.
2054 .
2055 The plus,
2056 minus,
2057 and equals signs appeared in the special font despite availability in
2058 text fonts \[lq]to insulate the appearance of equations from the choice
2059 of standard [read: text] fonts\[rq]\[em]a priority since
2060 .I troff \" AT&T
2061 was turned to the task of mathematical typesetting as soon as it was
2062 developed.
2063 .
2064 .
2065 .P
2066 We note that AT&T took the opportunity to de-unify the apostrophe/right
2067 single quotation mark from the acute accent
2068 (a choice ISO later duplicated in its 8859 series of standards).
2069 .
2070 A slash intended to be mirror-symmetric with the backslash was also
2071 included,
2072 as was the Bell System logo;
2073 we do not attempt to depict the latter.
2074 .
2075 .
2076 .br
2077 .if t .ne 5v
2078 .if t .ne 7v \" account for box border
2079 .P
2080 .TS
2081 center box;
2082 Lf(I),Lf(R).
2083 \[*a] \[*b] \[*g] \[*d] \[*e] \[*z] \[*y] \[*h] \[*i] \[*k] \[*l] \
2084 \[*m] \[*n] \[*c] \[*o] \[*p] \[*r] \[*s] \[ts] \[*t] \[*u] \[*f] \
2085 \[*x] \[*q] \[*w]
2086 \[*G] \[*D] \[*H] \[*L] \[*C] \[*P] \[*S] \[*U] \[*F] \[*Q] \[*W]
2087 \[dq] \[aa] \[rs] \[ha] \[ul] \[ga] \[ti] \[sl] < > { } # @ \
2088 \[pl] \[mi] \[eq] \[**]
2089 .\" We use \[radicalex] instead of \[rn] for more reliable simulation of
2090 .\" the typeface shown in Table I of CSTR #54 (1976); see subsection
2091 .\" "Mathematical symbols" above.
2092 \[>=] \[<=] \[==] \[~=] \[ap] \[!=] \
2093 \[ua] \[da] \[<-] \[->] \[mu] \[di] \[+-] \
2094 \[if] \[pd] \[gr] \[no] \[is] \[pt] \[sr] \[radicalex] \
2095 \[cu] \[ca] \[sb] \[sp] \[ib] \[ip] \[es] \[mo]
2096 \[sc] \[dd] \[lh] \[rh] \[or] \[ci] \
2097 \[lt] \[lb] \[rt] \[rb] \[lk] \[rk] \[bv] \[lf] \[rf] \[lc] \[rc]
2098 .TE
2099 .
2100 .
2101 .P
2102 One ASCII character as rendered by the Model 37 was apparently
2103 abandoned.
2104 .
2105 That device printed decimal 124 (\[or]) as a broken vertical line,
2106 like Unicode U+00A6 (\[bb]).
2107 .
2108 No equivalent was available on the C/A/T;
2109 the box rule
2110 .BR \[rs][br] ,
2111 brace vertical extension
2112 .\" CSTR #54 (1976 edition) called this the "bold vertical", probably
2113 .\" because it was thicker than the box rule and matched the thickness
2114 .\" of the bracket pieces \(lt, \(lb, \(rt, \(rb, \(lk, \(rk, and so on.
2115 .\" Saying "bold" could be misleading because it appeared only in the
2116 .\" special font, not a bold text font.
2117 .BR \[rs][bv] ,
2118 and \[lq]or\[rq] operator
2119 .B \[rs][or]
2120 were used as contextually appropriate.
2121 .
2122 .
2123 .P
2124 .\" In the Holt, Reinhart, Winston edition of the _Unix Programmer's
2125 .\" Manual_, Revised and Expanded Version, Volume 2 (1983), the square
2126 .\" \(sq in Times bold is _not_ shown as filled on page 226.
2127 .\"
2128 .\" ...but in the AT&T USG Unix 4.0 manual (ca. 1981), typeset on the
2129 .\" Autologic APS-5, the Times bold \(sq _is_ filled.
2130 .\"
2131 .\" https://www.tuhs.org/Archive/Documentation/Manuals/Unix_4.0/
2132 .\" Volume_1/00_Annotated_Table_of_Contents.pdf
2133 .\" Volume_1/C.1.2_NROFF_TROFF_Users_Manual.pdf
2134 .\" -- GBR
2135 Devices supported by AT&T device-independent
2136 .I troff
2137 exhibited some differences in glyph detail.
2138 .
2139 For example,
2140 on the Autologic APS-5 phototypesetter,
2141 the square
2142 .B \[rs](sq
2143 became filled in the Times bold face.
2144 .
2145 .
2146 .\" ====================================================================
2147 .SH Files
2148 .\" ====================================================================
2149 .
2150 The files below are loaded automatically by the default
2151 .IR troffrc .
2152 .
2153 .
2154 .TP
2155 .I /usr/\:\%share/\:\%groff/\:\%1.24.1/\:\%tmac/\:\%composite\:.tmac
2156 assigns alternate mappings for identifiers after the first in a
2157 composite special character escape sequence.
2158 .
2159 See subsection \[lq]Accents\[rq] above.
2160 .
2161 .
2162 .TP
2163 .I /usr/\:\%share/\:\%groff/\:\%1.24.1/\:\%tmac/\:\%fallbacks\:.tmac
2164 defines fallback mappings for Unicode code points such as the increment
2165 sign (U+2206) and upper- and lowercase Roman numerals.
2166 .
2167 .
2168 .\" ====================================================================
2169 .SH Authors
2170 .\" ====================================================================
2171 .
2172 This document was written by
2173 .MT jjc@\:jclark\:.com
2174 James Clark
2175 .ME ,
2176 with additions by
2177 .MT wl@\:gnu\:.org
2178 Werner Lemberg
2179 .ME
2180 and
2181 .MT groff\-bernd\:.warken\-72@\:web\:.de
2182 Bernd Warken
2183 .ME ,
2184 revised to use
2185 .MR \%tbl 1
2186 by
2187 .MT esr@\:thyrsus\:.com
2188 Eric S.\& Raymond
2189 .ME ,
2190 and largely rewritten by
2191 .MT g.branden\:.robinson@\:gmail\:.com
2192 G.\& Branden Robinson
2193 .ME .
2194 .
2195 .
2196 .\" ====================================================================
2197 .SH "See also"
2198 .\" ====================================================================
2199 .
2200 .IR "Groff: The GNU Implementation of troff" ,
2201 by Trent A.\& Fisher and Werner Lemberg,
2202 is the primary
2203 .I groff
2204 manual.
2205 .
2206 Section \[lq]Using Symbols\[rq] may be of particular note.
2207 .
2208 You can browse it interactively with \[lq]info \[aq](groff) Using
2209 Symbols\[aq]\[rq].
2210 .
2211 .
2212 .P
2213 \[lq]An extension to the
2214 .I troff
2215 character set for Europe\[rq],
2216 E.G.\& Keizer,
2217 K.J.\& Simonsen,
2218 J.\& Akkerhuis;
2219 EUUG Newsletter,
2220 Volume 9,
2221 No.\& 2,
2222 Summer 1989
2223 .
2224 .
2225 .P
2226 .UR http://\:www\:.unicode\:.org
2227 The Unicode Standard
2228 .UE
2229 .
2230 .
2231 .br
2232 .ne 2v
2233 .P
2234 .UR https://\:www\:.aivosto\:.com/\:articles/\:charsets\-7bit\:.html
2235 \[lq]7-bit Character Sets\[rq]
2236 .UE
2237 by Tuomas Salste documents the inherent ambiguity and configurable code
2238 points of the ASCII encoding standard.
2239 .
2240 .
2241 .P
2242 \[lq]Nroff/Troff User's Manual\[rq]
2243 by Joseph F.\& Ossanna,
2244 1976,
2245 AT&T Bell Laboratories Computing Science Technical Report No.\& 54,
2246 features two tables that throw light on the glyph repertoire available
2247 to \[lq]typesetter
2248 .IR roff \[rq]
2249 when it was first written.
2250 .
2251 Be careful of re-typeset versions of this document that can be found on
2252 the Internet.
2253 .
2254 Some do not accurately represent the original document:
2255 several glyphs are obviously missing.
2256 .
2257 More subtly,
2258 lowercase Greek letters are rendered upright,
2259 not slanted as they appeared in the C/A/T's special font and as expected
2260 by
2261 .I troff \" AT&T
2262 users.
2263 .
2264 .
2265 .P
2266 .MR groff_rfc1345 7
2267 describes an alternative set of special character glyph names,
2268 which extends and in some cases overrides the definitions listed above.
2269 .
2270 .
2271 .P
2272 .MR groff 1 ,
2273 .MR \%troff 1 ,
2274 .MR groff 7
2275 .
2276 .
2277 .\" Restore compatibility mode (for, e.g., Solaris 10/11).
2278 .cp \n[*groff_groff_char_7_man_C]
2279 .do rr *groff_groff_char_7_man_C
2280 .
2281 .
2282 .\" Local Variables:
2283 .\" fill-column: 72
2284 .\" mode: nroff
2285 .\" tab-width: 20
2286 .\" End:
2287 .\" vim: set filetype=groff tabstop=20 textwidth=72: