1EQN(1) General Commands Manual EQN(1)
2
6 eqn - format equations for troff or MathML
7
9 eqn [-rvCNR] [-d xy] [-T name] [-M dir] [-f F] [-s n] [-p n] [-m n]
10 [file ...]
11
13 This manual page describes the GNU version of eqn, which is part of the
14 groff document formatting system. eqn compiles descriptions of equa‐
15 tions embedded within troff input files into commands that are under‐
16 stood by troff. Normally, it should be invoked using the -e option of
17 groff. The syntax is quite compatible with Unix eqn. The output of
18 GNU eqn cannot be processed with Unix troff; it must be processed with
19 GNU troff. If no files are given on the command line, the standard in‐
20 put is read. A filename of - causes the standard input to be read.
21 eqn searches for the file eqnrc in the directories given with the -M
23 option first, then in /etc/groff/site-tmac, /etc/groff/site-tmac, and
24 finally in the standard macro directory /usr/share/groff/1.22.4/tmac.
25 If it exists, eqn processes it before the other input files. The -R
26 option prevents this.
27 GNU eqn does not provide the functionality of neqn: it does not support
29 low-resolution, typewriter-like devices (although it may work ade‐
30 quately for very simple input).
31
33 Whitespace is permitted between a command-line option and its argument.
34 -dxy Specify delimiters x and y for the left and right end, respec‐
36 tively, of in-line equations. Any delim statements in the
37 source file overrides this.
38 -C Recognize .EQ and .EN even when followed by a character other
40 than space or newline. Also, the statement ‘delim on’ is not
41 handled specially.
42 -N Don't allow newlines within delimiters. This option allows eqn
44 to recover better from missing closing delimiters.
45 -v Print the version number.
47 -r Only one size reduction.
49 -mn The minimum point-size is n. eqn does not reduce the size of
51 subscripts or superscripts to a smaller size than n.
52 -Tname The output is for device name. Normally, the only effect of
54 this is to define a macro name with a value of 1; eqnrc uses
55 this to provide definitions appropriate for the output device.
56 However, if the specified device is “MathML”, the output is
57 MathML markup rather than troff commands, and eqnrc is not
58 loaded at all. The default output device is ps.
59 -Mdir Search dir for eqnrc before the default directories.
61 -R Don't load eqnrc.
63 -fF This is equivalent to a gfont F command.
65 -sn This is equivalent to a gsize n command. This option is depre‐
67 cated. eqn normally sets equations at whatever the current
68 point size is when the equation is encountered.
69 -pn This says that subscripts and superscripts should be n points
71 smaller than the surrounding text. This option is deprecated.
72 Normally eqn sets subscripts and superscripts at 70% of the size
73 of the surrounding text.
74
76 Only the differences between GNU eqn and Unix eqn are described here.
77 GNU eqn emits Presentation MathML output when invoked with the
79 -T MathML option.
80 GNU eqn sets the input token "..." as three periods or low dots,
82 rather than the three centered dots of classic eqn. To get three cen‐
83 tered dots, write cdots or cdot cdot cdot.
84 Most of the new features of the GNU eqn input language are based on
86 TeX. There are some references to the differences between TeX and GNU
87 eqn below; these may safely be ignored if you do not know TeX.
88 Controlling delimiters
90 If not in compatibility mode, eqn recognizes
91 delim on
93 to restore the delimiters which have been previously disabled with a
95 call to ‘delim off’. If delimiters haven't been specified, the call
96 has no effect.
97 Automatic spacing
99 eqn gives each component of an equation a type, and adjusts the spacing
100 between components using that type. Possible types are described in
101 the table below.
102 ordinary an ordinary character such as ‘1’ or ‘x’
104 operator a large operator such as ‘Σ’
105 binary a binary operator such as ‘+’
106 relation a relation such as ‘=’
107 opening a opening bracket such as ‘(’
108 closing a closing bracket such as ‘)’
109 punctuation a punctuation character such as ‘,’
110 inner a subformula contained within brackets
111 suppress a type that suppresses automatic spacing adjustment
112 Components of an equation get a type in one of two ways.
114 type t e
116 This yields an equation component that contains e but that has
117 type t, where t is one of the types mentioned above. For exam‐
118 ple, times is defined as
119 type "binary" \(mu
121 The name of the type doesn't have to be quoted, but quoting pro‐
123 tects from macro expansion.
124 chartype t text
126 Unquoted groups of characters are split up into individual char‐
127 acters, and the type of each character is looked up; this
128 changes the type that is stored for each character; it says that
129 the characters in text from now on have type t. For example,
130 chartype "punctuation" .,;:
132 would make the characters ‘.,;:’ have type punctuation whenever
134 they subsequently appeared in an equation. The type t can also
135 be letter or digit; in these cases chartype changes the font
136 type of the characters. See subsection “Fonts” below.
137 New primitives
139 big e Enlarges the expression it modifies; intended to have semantics
140 like CSS ‘large’. In troff output, the point size is increased
141 by 5; in MathML output, the expression uses
142 <mstyle mathsize='big'>
144 e1 smallover e2
146 This is similar to over; smallover reduces the size of e1 and
147 e2; it also puts less vertical space between e1 or e2 and the
148 fraction bar. The over primitive corresponds to the TeX \over
149 primitive in display styles; smallover corresponds to \over in
150 non-display styles.
151 vcenter e
153 This vertically centers e about the math axis. The math axis is
154 the vertical position about which characters such as ‘+’ and ‘−’
155 are centered; also it is the vertical position used for the bar
156 of fractions. For example, sum is defined as
157 { type "operator" vcenter size +5 \(*S }
159 (Note that vcenter is silently ignored when generating MathML.)
161 e1 accent e2
163 This sets e2 as an accent over e1. e2 is assumed to be at the
164 correct height for a lowercase letter; e2 is moved down accord‐
165 ing to whether e1 is taller or shorter than a lowercase letter.
166 For example, hat is defined as
167 accent { "^" }
169 dotdot, dot, tilde, vec, and dyad are also defined using the ac‐
171 cent primitive.
172 e1 uaccent e2
174 This sets e2 as an accent under e1. e2 is assumed to be at the
175 correct height for a character without a descender; e2 is moved
176 down if e1 has a descender. utilde is pre-defined using uaccent
177 as a tilde accent below the baseline.
178 split "text"
180 This has the same effect as simply
181 text
183 but text is not subject to macro expansion because it is quoted;
185 text is split up and the spacing between individual characters
186 is adjusted.
187 nosplit text
189 This has the same effect as
190 "text"
192 but because text is not quoted it is subject to macro expansion;
194 text is not split up and the spacing between individual charac‐
195 ters is not adjusted.
196 e opprime
198 This is a variant of prime that acts as an operator on e. It
199 produces a different result from prime in a case such as A op‐
200 prime sub 1: with opprime the 1 is tucked under the prime as a
201 subscript to the A (as is conventional in mathematical typeset‐
202 ting), whereas with prime the 1 is a subscript to the prime
203 character. The precedence of opprime is the same as that of bar
204 and under, which is higher than that of everything except accent
205 and uaccent. In unquoted text a ' that is not the first charac‐
206 ter is treated like opprime.
207 special text e
209 This constructs a new object from e using a troff(1) macro named
210 text. When the macro is called, the string 0s contains the out‐
211 put for e, and the number registers 0w, 0h, 0d, 0skern, and
212 0skew contain the width, height, depth, subscript kern, and skew
213 of e. (The subscript kern of an object says how much a sub‐
214 script on that object should be tucked in; the skew of an object
215 says how far to the right of the center of the object an accent
216 over the object should be placed.) The macro must modify 0s so
217 that it outputs the desired result with its origin at the cur‐
218 rent point, and increase the current horizontal position by the
219 width of the object. The number registers must also be modified
220 so that they correspond to the result.
221 For example, suppose you wanted a construct that ‘cancels’ an
223 expression by drawing a diagonal line through it.
224 .EQ
226 define cancel 'special Ca'
227 .EN
228 .de Ca
229 . ds 0s \
230 \Z'\\*(0s'\
231 \v'\\n(0du'\
232 \D'l \\n(0wu -\\n(0hu-\\n(0du'\
233 \v'\\n(0hu'
234 ..
235 Then you could cancel an expression e with cancel { e }
237 Here's a more complicated construct that draws a box round an
239 expression:
240 .EQ
242 define box 'special Bx'
243 .EN
244 .de Bx
245 . ds 0s \
246 \Z'\h'1n'\\*(0s'\
247 \Z'\
248 \v'\\n(0du+1n'\
249 \D'l \\n(0wu+2n 0'\
250 \D'l 0 -\\n(0hu-\\n(0du-2n'\
251 \D'l -\\n(0wu-2n 0'\
252 \D'l 0 \\n(0hu+\\n(0du+2n'\
253 '\
254 \h'\\n(0wu+2n'
255 . nr 0w +2n
256 . nr 0d +1n
257 . nr 0h +1n
258 ..
259 space n
261 A positive value of the integer n (in hundredths of an em) sets
262 the vertical spacing before the equation, a negative value sets
263 the spacing after the equation, replacing the default values.
264 This primitive provides an interface to groff's \x escape (but
265 with opposite sign).
266 This keyword has no effect if the equation is part of a pic pic‐
268 ture.
269 Extended primitives
271 col n { ... }
272 ccol n { ... }
273 lcol n { ... }
274 rcol n { ... }
275 pile n { ... }
276 cpile n { ... }
277 lpile n { ... }
278 rpile n { ... }
279 The integer value n (in hundredths of an em) increases the ver‐
280 tical spacing between rows, using groff's \x escape (the value
281 has no effect in MathML mode). Negative values are possible but
282 have no effect. If there is more than a single value given in a
283 matrix, the biggest one is used.
284 Customization
286 When eqn is generating troff markup, the appearance of equations is
287 controlled by a large number of parameters. They have no effect when
288 generating MathML mode, which pushes typesetting and fine motions down‐
289 stream to a MathML rendering engine. These parameters can be set using
290 the set command.
291 set p n
293 This sets parameter p to value n; n is an integer. For example,
294 set x_height 45
296 says that eqn should assume an x height of 0.45 ems.
298 Possible parameters are as follows. Values are in units of hun‐
300 dredths of an em unless otherwise stated. These descriptions
301 are intended to be expository rather than definitive.
302 minimum_size
304 eqn doesn't set anything at a smaller point-size than
305 this. The value is in points.
306 fat_offset
308 The fat primitive emboldens an equation by overprinting
309 two copies of the equation horizontally offset by this
310 amount. This parameter is not used in MathML mode; in‐
311 stead, fat text uses
312 <mstyle mathvariant='double-struck'>
314 over_hang
316 A fraction bar is longer by twice this amount than the
317 maximum of the widths of the numerator and denominator;
318 in other words, it overhangs the numerator and denomina‐
319 tor by at least this amount.
320 accent_width
322 When bar or under is applied to a single character, the
323 line is this long. Normally, bar or under produces a
324 line whose length is the width of the object to which it
325 applies; in the case of a single character, this tends to
326 produce a line that looks too long.
327 delimiter_factor
329 Extensible delimiters produced with the left and right
330 primitives have a combined height and depth of at least
331 this many thousandths of twice the maximum amount by
332 which the sub-equation that the delimiters enclose ex‐
333 tends away from the axis.
334 delimiter_shortfall
336 Extensible delimiters produced with the left and right
337 primitives have a combined height and depth not less than
338 the difference of twice the maximum amount by which the
339 sub-equation that the delimiters enclose extends away
340 from the axis and this amount.
341 null_delimiter_space
343 This much horizontal space is inserted on each side of a
344 fraction.
345 script_space
347 The width of subscripts and superscripts is increased by
348 this amount.
349 thin_space
351 This amount of space is automatically inserted after
352 punctuation characters.
353 medium_space
355 This amount of space is automatically inserted on either
356 side of binary operators.
357 thick_space
359 This amount of space is automatically inserted on either
360 side of relations.
361 x_height
363 The height of lowercase letters without ascenders such as
364 ‘x’.
365 axis_height
367 The height above the baseline of the center of characters
368 such as ‘+’ and ‘−’. It is important that this value is
369 correct for the font you are using.
370 default_rule_thickness
372 This should set to the thickness of the \(ru character,
373 or the thickness of horizontal lines produced with the \D
374 escape sequence.
375 num1 The over command shifts up the numerator by at least this
377 amount.
378 num2 The smallover command shifts up the numerator by at least
380 this amount.
381 denom1 The over command shifts down the denominator by at least
383 this amount.
384 denom2 The smallover command shifts down the denominator by at
386 least this amount.
387 sup1 Normally superscripts are shifted up by at least this
389 amount.
390 sup2 Superscripts within superscripts or upper limits or nu‐
392 merators of smallover fractions are shifted up by at
393 least this amount. This is usually less than sup1.
394 sup3 Superscripts within denominators or square roots or sub‐
396 scripts or lower limits are shifted up by at least this
397 amount. This is usually less than sup2.
398 sub1 Subscripts are normally shifted down by at least this
400 amount.
401 sub2 When there is both a subscript and a superscript, the
403 subscript is shifted down by at least this amount.
404 sup_drop
406 The baseline of a superscript is no more than this much
407 amount below the top of the object on which the super‐
408 script is set.
409 sub_drop
411 The baseline of a subscript is at least this much below
412 the bottom of the object on which the subscript is set.
413 big_op_spacing1
415 The baseline of an upper limit is at least this much
416 above the top of the object on which the limit is set.
417 big_op_spacing2
419 The baseline of a lower limit is at least this much below
420 the bottom of the object on which the limit is set.
421 big_op_spacing3
423 The bottom of an upper limit is at least this much above
424 the top of the object on which the limit is set.
425 big_op_spacing4
427 The top of a lower limit is at least this much below the
428 bottom of the object on which the limit is set.
429 big_op_spacing5
431 This much vertical space is added above and below limits.
432 baseline_sep
434 The baselines of the rows in a pile or matrix are nor‐
435 mally this far apart. In most cases this should be equal
436 to the sum of num1 and denom1.
437 shift_down
439 The midpoint between the top baseline and the bottom
440 baseline in a matrix or pile is shifted down by this much
441 from the axis. In most cases this should be equal to
442 axis_height.
443 column_sep
445 This much space is added between columns in a matrix.
446 matrix_side_sep
448 This much space is added at each side of a matrix.
449 draw_lines
451 If this is non-zero, lines are drawn using the \D escape
452 sequence, rather than with the \l escape sequence and the
453 \(ru character.
454 body_height
456 The amount by which the height of the equation exceeds
457 this is added as extra space before the line containing
458 the equation (using \x). The default value is 85.
459 body_depth
461 The amount by which the depth of the equation exceeds
462 this is added as extra space after the line containing
463 the equation (using \x). The default value is 35.
464 nroff If this is non-zero, then ndefine behaves like define and
466 tdefine is ignored, otherwise tdefine behaves like define
467 and ndefine is ignored. The default value is 0 (This is
468 typically changed to 1 by the eqnrc file for the ascii,
469 latin1, utf8, and cp1047 devices.)
470 A more precise description of the role of many of these parame‐
472 ters can be found in Appendix H of The TeXbook.
473 Macros
475 Macros can take arguments. In a macro body, $n where n is between 1
476 and 9, is replaced by the nth argument if the macro is called with ar‐
477 guments; if there are fewer than n arguments, it is replaced by noth‐
478 ing. A word containing a left parenthesis where the part of the word
479 before the left parenthesis has been defined using the define command
480 is recognized as a macro call with arguments; characters following the
481 left parenthesis up to a matching right parenthesis are treated as
482 comma-separated arguments; commas inside nested parentheses do not ter‐
483 minate an argument.
484 sdefine name X anything X
486 This is like the define command, but name is not recognized if
487 called with arguments.
488 include "file"
490 copy "file"
491 Include the contents of file (include and copy are synonyms).
492 Lines of file beginning with .EQ or .EN are ignored.
493 ifdef name X anything X
495 If name has been defined by define (or has been automatically
496 defined because name is the output device) process anything;
497 otherwise ignore anything. X can be any character not appearing
498 in anything.
499 undef name
501 Remove definition of name, making it undefined.
502 Besides the macros mentioned above, the following definitions are
504 available: Alpha, Beta, ..., Omega (this is the same as ALPHA, BETA,
505 ..., OMEGA), ldots (three dots on the base line), and dollar.
506 Fonts
508 eqn normally uses at least two fonts to set an equation: an italic font
509 for letters, and a roman font for everything else. The existing gfont
510 command changes the font that is used as the italic font. By default
511 this is I. The font that is used as the roman font can be changed us‐
512 ing the new grfont command.
513 grfont f
515 Set the roman font to f.
516 The italic primitive uses the current italic font set by gfont; the ro‐
518 man primitive uses the current roman font set by grfont. There is also
519 a new gbfont command, which changes the font used by the bold primi‐
520 tive. If you only use the roman, italic and bold primitives to changes
521 fonts within an equation, you can change all the fonts used by your
522 equations just by using gfont, grfont and gbfont commands.
523 You can control which characters are treated as letters (and therefore
525 set in italics) by using the chartype command described above. A type
526 of letter causes a character to be set in italic type. A type of digit
527 causes a character to be set in roman type.
528
530 /usr/share/groff/1.22.4/tmac/eqnrc
531 Initialization file.
532
534 MathML is designed on the assumption that it cannot know the exact
535 physical characteristics of the media and devices on which it will be
536 rendered. It does not support fine control of motions and sizes to the
537 same degree troff does. Thus:
538 * eqn parameters have no effect on the generated MathML.
540 * The special, up, down, fwd, and back operations cannot be imple‐
542 mented, and yield a MathML ‘<merror>’ message instead.
543 * The vcenter keyword is silently ignored, as centering on the
545 math axis is the MathML default.
546 * Characters that eqn over troff sets extra large – notably the
548 integral sign – may appear too small and need to have their
549 ‘<mstyle>’ wrappers adjusted by hand.
550 As in its troff mode, eqn in MathML mode leaves the .EQ and .EN delim‐
552 iters in place for displayed equations, but emits no explicit delim‐
553 iters around inline equations. They can, however, be recognized as
554 strings that begin with ‘<math>’ and end with ‘</math>’ and do not
555 cross line boundaries.
556 See section “Bugs” below for translation limits specific to eqn.
558
560 Inline equations are set at the point size that is current at the be‐
561 ginning of the input line.
562 In MathML mode, the mark and lineup features don't work. These could,
564 in theory, be implemented with ‘<maligngroup>’ elements.
565 In MathML mode, each digit of a numeric literal gets a separate ‘<mn>
567 </mn>’ pair, and decimal points are tagged with ‘<mo></mo>’. This is
568 allowed by the specification, but inefficient.
569
571 groff(1), troff(1), pic(1), groff_font(5), The TeXbook
572groff 1.22.4 21 July 2022 EQN(1)