summaryrefslogtreecommitdiff
path: root/macros/generic/multiexpand/multiexpand.dtx
blob: ad3b304ef037acf0b1fea87cd7fad7f354cb0159 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
% \iffalse meta-comment
%<*internal>
\iffalse
%</internal>
%<*readme>
Variations on the `\expandafter` TeX primitive
==============================================
* E-mail: blflatex@gmail.com
* Released under the LaTeX Project Public License v1.3c or later
  See http://www.latex-project.org/lppl.txt

The package defines `\multiexpand`, `\multiexpandafter`, `\MultiExpand`
and `\MultiExpandAfter`, replacing large chains of `\expandafter`.
These four commands take as an argument the number of expansion to be
done.  If eTeX is available, this number is evaluated using `\numexpr`.

Say we want to expand `\C` five times before `\A` and `\B` in `\A\B\C`.
The traditional approach would be to insert 31 `\expandafter` before
`\A` and the same number before `\B`.  With this package one can use
any of

    \expandafter\A\expandafter\B\romannumeral\multiexpand{5}\C

    \expandafter\A\romannumeral\multiexpandafter{5}\B\C

    \MultiExpandAfter{2}\A\MultiExpandAfter{5}\B\C

    \MultiExpandAfter{2}\A\MultiExpandAfter{2}\B\MultiExpand{5}\C

In one step of expansion (triggered by the `\expandafter`'s),
`\romannumeral\multiexpand{5}` expands the following token 5 times,
whereas `\romannumeral\multiexpandafter{5}` expands the token after
that 5 times.  The macros `\MultiExpandAfter` and `\MultiExpand`
take two steps of expansion, but do not require `\romannumeral`.
Another example is that in two steps of expansion,

    \MultiExpandAfter{2}\a\MultiExpandAfter{3}{%
      \MultiExpandAfter{9}\b\MultiExpandAfter{10}\c\d}

expands `\d` 10 times, then `\c` 7 times (9-2), then `\b` once (3-2).

The package can be built from the file `multiexpand.dtx' by running

    pdflatex multiexpand.dtx
    pdflatex multiexpand.dtx

%</readme>
%<*lvt>
 \input regression-test.tex\relax
 \START
 \LONGTYPEOUT{Author: Bruno Le Floch}
%<lvt002> \let\numexpr\relax
 \input multiexpand.sty\relax

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % Introduce commands that expand to each other, to count how many times
 % things were expanded.
 \OMIT
 \def\i{}
 \def\v{\i\i\i\i}
 \def\x{\i\i\i\i\v}
 \def\l{\i\i\i\i\v\x\x\x\x}
 \def\c{\i\i\i\i\v\x\x\x\x\l}
 \def\d{\i\i\i\i\v\x\x\x\x\l\c\c\c\c}
 \def\m{\i\i\i\i\v\x\x\x\x\l\c\c\c\c\d}
 \TIMO

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % We test |\MultiExpand|, which needs two expansions. In particular, we see
 % that the implementation is reasonably quick, and we get the $2345$-th
 % expansion of |\m\m\m\m|.
 % 4000-2345 = 1655 = mdclv
 \TEST{MultiExpand many times (vlcdm)}{%
   \toks0\expandafter\expandafter\expandafter{\MultiExpand{2345}\m\m\m\m}%
   \TYPE{\the\toks0.}%
 }

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % Test |\MultiExpandAfter|, two expansions to get the specified expansion.
 % The results indirectly show that expansions happen in the expected order.
 % 100 = c
 % 100-(27-2) = 75 = lxxv
 % 100-(14-2) = 88 = lxxxviii
 % 100-38 = 62 = lxii
 \TEST{MultiExpandAfter chain (c vxxl iiivxxxl iixl)}{%
   \toks0\expandafter\expandafter\expandafter{%
     \MultiExpandAfter{27}\c
     \MultiExpandAfter{14}\c
     \MultiExpandAfter{38}\c\c}%
   \TYPE{\the\toks0.}%
 }

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % Test the order of expansion.  The example was given in the original
 % question that led to writing this package.  For the record, this
 % package (at some point) used 32 |\expandafter| to do the expansion
 % in the right order, which was better than other approaches; however
 % this counting ignores the large overheads due to other primitives.
 \OMIT
 \def\a#1.{#1}
 \def\b#1:{#1.}
 \def\c#1,{#1:}
 \def\d{Hello world!,}
 \TIMO
 %
 \TEST{MultiExpandAfter order}{%
   \toks0\MultiExpandAfter{3}{%
     \MultiExpandAfter{3}\a
     \MultiExpandAfter{3}\b
     \MultiExpandAfter{1}\c
     \d}%
   \TYPE{\the\toks0}%
 }

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % The other example from the same question (expanding seven tokens once
 % each in the reverse order) takes 74 \expandafter.
 \OMIT
 \def\b.{.b}    \def\c.{.c}    \def\d.{.d}
 \def\e.{.e}    \def\f.{.f}    \def\g.{.g}
 \TIMO
 %
 \TEST{MultiExpandAfter order (2)}{%
   \toks0\MultiExpandAfter{2}{%
     \MultiExpandAfter3\a    \MultiExpandAfter3\b
     \MultiExpandAfter3\c    \MultiExpandAfter3\d
     \MultiExpandAfter3\e    \MultiExpandAfter1\f\g.}%
   \TYPE{\the\toks0}%
 }

 \END
%</lvt>
%<*tlg>
This is a generated file for the l3build validation system.
Don't change this file in any respect.
Author: Bruno Le Floch
(multiexpand.sty)
============================================================
TEST 1: MultiExpand many times (vlcdm)
============================================================
\v \l \c \d \m .
============================================================
============================================================
TEST 2: MultiExpandAfter chain (c vxxl iiivxxxl iixl)
============================================================
\c \v \x \x \l \i \i \i \v \x \x \x \l \i \i \x \l .
============================================================
============================================================
TEST 3: MultiExpandAfter order
============================================================
Hello world!
============================================================
============================================================
TEST 4: MultiExpandAfter order (2)
============================================================
\a .bcdefg
============================================================
%</tlg>
%<*internal>
\fi
\def\nameofplainTeX{plain}
\ifx\fmtname\nameofplainTeX\else
  \expandafter\begingroup
\fi
%</internal>
%<*install>
\input docstrip.tex
\keepsilent
\askforoverwritefalse
\preamble
----------------------------------------------------------------
multiexpand --- trigger multiple expansions in one expansion step.
E-mail: blflatex@gmail.com
Released under the LaTeX Project Public License v1.3c or later
See http://www.latex-project.org/lppl.txt
----------------------------------------------------------------

\endpreamble
\postamble

Copyright (C) 2011-2017 by Bruno Le Floch <blflatex@gmail.com>

This work may be distributed and/or modified under the
conditions of the LaTeX Project Public License (LPPL), either
version 1.3c of this license or (at your option) any later
version.  The latest version of this license is in the file:

http://www.latex-project.org/lppl.txt

This work is "maintained" (as per LPPL maintenance status) by
Bruno Le Floch.

This work consists of the file  multiexpand.dtx
and the derived files           multiexpand.ins,
                                multiexpand.pdf and
                                multiexpand.sty.

\endpostamble
\usedir{tex/latex/multiexpand}
\generate{
  \file{\jobname.sty}{\from{\jobname.dtx}{package}}
}
%</install>
%<install>\endbatchfile
%<*internal>
\usedir{source/latex/multiexpand}
\generate{
  \file{\jobname.ins}{\from{\jobname.dtx}{install}}
  \file{\jobname 001.lvt}{\from{\jobname.dtx}{lvt}}
  \file{\jobname 002.lvt}{\from{\jobname.dtx}{lvt,lvt002}}
}
\nopreamble\nopostamble
\generate{
  \file{\jobname 001.tlg}{\from{\jobname.dtx}{tlg}}
  \file{\jobname 002.tlg}{\from{\jobname.dtx}{tlg}}
}
\usedir{doc/latex/multiexpand}
\generate{
  \file{README.md}{\from{\jobname.dtx}{readme}}
}

\ifx\fmtname\nameofplainTeX
  \expandafter\endbatchfile
\else
  \expandafter\endgroup
\fi
%</internal>
%<*driver>
\documentclass[12pt]{ltxdoc}
\usepackage[T1]{fontenc}
\usepackage{hologo}
\usepackage{multiexpand}
\usepackage[margin=4cm]{geometry}
\usepackage{hyperref}
\RecordChanges
\begin{document}
  \DocInput{multiexpand.dtx}
\end{document}
%</driver>
% \fi
%
% \CheckSum{126}
%
% \def\fileversion{v1.5} \def\filedate{2017/11/29}
%\title{\texttt{multiexpand}\\
%  Trigger multiple expansions \\ in one expansion step\thanks{This file
%    describes version \fileversion, last revised \filedate.}}
%\author{Bruno Le Floch\thanks{E-mail: blflatex@gmail.com}
%  \thanks{I have gathered ideas from various posts in the \texttt{\{TeX\}}
%    community at \url{http://tex.stackexchange.com}. Thanks to their authors.}}
%\date{Released \filedate}
%
%\maketitle
%
% \tableofcontents
% \changes{v1.0}{2011/02/15}{First version with documentation}
% \changes{v1.1}{2013/01/08}{Version submitted to CTAN}
% \changes{v1.2}{2013/01/09}{Use fewer expandafter for large arguments}
% \changes{v1.2}{2013/01/09}{Change ME prefix to multiexpand}
% \changes{v1.4}{2015/09/20}{Clarify that eTeX is not required}
% \changes{v1.5}{2017/11/29}{Updates due to l3build changes}
%
% \section{Two user commands}
%
% \begin{itemize}
% \item For $n>0$, expanding |\MultiExpand{|$n$|}\macro| twice gives
%   the $n$-th expansion of |\macro|.
% \item For $n>0$, expanding |\MultiExpandAfter{|$n$|}\macroA\macroB|
%   twice expands |\macroB| $n$ times before expanding |\macroA|.
% \end{itemize}
% Note that neither functions work for $n = 0$.
%
% These can typically be combined as
% \begin{verbatim}
% \MultiExpand{7}%
% \MultiExpandAfter{4}\a\MultiExpandAfter{7}\b%
% \MultiExpandAfter{3}\c\d
% \end{verbatim}
% which would expand |\d| $3$~times, then |\c| $5$~times ($2$~of the $7$~times
% were used to expand |\MultiExpandAfter{3}|), then |\b| twice ($4-2$), and
% finally |\a| five times ($7-2$). Note that all this happens in precisely
% \emph{two} steps of expansion.
%
% In some cases, one needs to achieve the same effect in \emph{one} step
% only.  For this, we use the first expansion of |\MultiExpand|, which
% is |\romannumeral| |\multiexpand|, or of |\MultiExpandAfter|, which is
% |\romannumeral| |\multiexpandafter|.  In detail, expanding |\romannumeral|
% |\multiexpand{|$n$|}| once expands the following token $n$ times, and
% similarly for |\romannumeral| |\multiexpandafter{|$n$|}|.
%
% These are especially useful when we want to expand several times a
% very specific token which is buried behind many others. For instance,
% expanding the following code once
% \begin{verbatim}
% \expandafter\macroA\expandafter\macroB
% \romannumeral\multiexpandafter{4}\macroC\macroD
% \end{verbatim}
% will expand |\macroD| $4$ times before the three other macros.
%
% Note: as we mentionned, this breaks for $n=0$. But in this case, consider
% using |\expandafter\empty|, or a variant thereof.
%
% \section{Implementation}
%\StopEventually{\PrintChanges\PrintIndex}
%    \begin{macrocode}
%<*package>
%    \end{macrocode}
%
% We work inside a group, to change the catcode of |@|. So we will only
% do |\gdef|s.  Note that this code can be read several times with no
% issue; no need to bother to check whether it was already read or not.
%    \begin{macrocode}
\begingroup
\catcode `\@=11
%    \end{macrocode}
%
% \subsection{Common to the \hologo{eTeX} and non-\hologo{eTeX} cases}
%
% For the ``lazy'', who do not want to use |\romannumeral|, we provide
% |\MultiExpand| and |\MultiExpandAfter|, simple shorthands.  A drawback
% is that they require two steps of expansion rather than only one.
%    \begin{macrocode}
\gdef \MultiExpand {\romannumeral \multiexpand }
\gdef \MultiExpandAfter {\romannumeral \multiexpandafter }
%    \end{macrocode}
%
% \subsection{Without \hologo{eTeX}'s \cs{numexpr}}
% \changes{v1.3}{2015/03/03}{Support TeX with no numexpr}
%
% No need for the usual |\begingroup\expandafter\endgroup| to prevent
% |\numexpr| from being set to |\relax|, because we are already in a
% group.
%    \begin{macrocode}
\expandafter \ifx \csname numexpr\endcsname \relax
%    \end{macrocode}
%
% A helper.
%    \begin{macrocode}
\long \gdef \multiexpand@gobble #1{}
%    \end{macrocode}
%
% The user commands |\multiexpand| and |\multiexpandafter|, to be used
% after |\romannumeral|.  They only differ a little bit.
%    \begin{macrocode}
\gdef \multiexpand {\multiexpand@aux \multiexpand@ }
\gdef \multiexpandafter {\multiexpand@aux \multiexpand@after }
%    \end{macrocode}
% The user commands receives a number, and to accept various forms of
% numbers we hit it with |\number|.  If it is non-positive, stop the
% |\romannumeral| expansion with |0| and a space.  Otherwise, reverse
% the number, to make it easy to subtract~$1$.
%    \begin{macrocode}
\long \gdef \multiexpand@aux #1#2%
  {\expandafter \multiexpand@test \number #2;#1}
\gdef \multiexpand@test #1;#2%
  {%
    \ifnum #1>0
      \multiexpand@reverse #1{?\multiexpand@reverse@end }?;;#2%
    \fi
    0 %
  }
%    \end{macrocode}
% The macro |\multiexpand@reverse| puts characters from the number one
% by one (as~|#1|) after the semicolon, to reverse the number.  After
% the last digit, |#1| is |{?\multiexpand@reverse@end}|.  The question
% mark is removed by |\multiexpand@gobble|, and the |reverse@end|
% macro cleans up.  In particular, one should not forget to close the
% conditional using~|#5|, which is the trailing~|\fi|.  At this stage,
% |#4| is the function that distinguishes |\multiexpand| from
% |\multiexpand@after|, and |#3| is the reversed
% number.  \begin{macrocode}
\gdef \multiexpand@reverse #1#2;%
  {\multiexpand@gobble #1\multiexpand@reverse #2;#1}
\gdef \multiexpand@reverse@end #1;?#2#3;#4#50
  {#5\multiexpand@iterate #41#3;}
%    \end{macrocode}
% The macro |\multiexpand@iterate| applies a \meta{function} a certain
% number of times to what follows in the input stream.  It expects to
% receive \meta{function} \meta{nines} |1|\meta{reversed number}|;|.
% The argument \meta{nines}, made entirely of the digit~|9|, is used to
% compute carries when subtracting~$1$, and is initally empty.  As a
% concrete example, after |\multiexpand{302}| the successive
% calls to |\multiexpand@iterate| would go as follows.
% \begin{verbatim}
%   \multiexpand@iterate \multiexpand@  1203;
%   \multiexpand@iterate \multiexpand@  1103;
%   \multiexpand@iterate \multiexpand@  1003;
%   \multiexpand@iterate \multiexpand@ 9 103;
%   \multiexpand@iterate \multiexpand@ 99 13;
%   \multiexpand@iterate \multiexpand@  1992;
%   \multiexpand@iterate \multiexpand@  1892;
%   \multiexpand@iterate \multiexpand@  1792;
% \end{verbatim}
% Note in particular how carries are done in several steps.  The details
% are left as an exercise to the reader.  The most common case is when
% |#2| is empty and |#3| is a non-zero digit.  Then |\number| is
% expanded, triggering |\ifcase| which shifts |#3| by one unit, and |#1|
% takes care of expanding the tokens are required by |\multiexpand| or
% |\multiexpandafter|.  If |#3| is~$0$, then |\multiexpand@zero| is
% called, closing the conditional with |#1|, and iterating, this time
% with a non-empty \meta{nines}, which are the argument |#2| of a new
% call to |\multiexpand@iterate|.  Those \meta{nines} are put back into
% the number by |\multiexpand@iterate|, unless the next significant
% digit is also $0$, in which case |\multiexpand@zero| is called again,
% until finding a non-zero digit; at each step, one more $9$ is added to
% the \meta{nines}.  If all digits are zero, we reach |;| this way, and
% end, after cleaning up.
%    \begin{macrocode}
\gdef \multiexpand@iterate #1#21#3%
  {%
    \ifx ;#3\multiexpand@end \fi
    \ifx 0#3\multiexpand@zero \fi
    \expandafter \multiexpand@iterate
    \expandafter #1%
    \number 1#2%
      \ifcase #3 \or 0\or 1\or 2\or 3\or 4\or 5\or 6\or 7\or 8\fi
      #1%
  }
\gdef \multiexpand@zero #1#2\number 1#3\ifcase #4\fi #5%
  {#1\multiexpand@iterate #59#31}
\gdef \multiexpand@end #1#2\ifcase #3\fi #4{#10 }
%    \end{macrocode}
% Finally, the two different expansion commands.
%    \begin{macrocode}
\gdef \multiexpand@ #1;{#1\expandafter ;}
\gdef \multiexpand@after #1;{#1\expandafter ;\expandafter }
%    \end{macrocode}
%
% \subsection{With \hologo{eTeX}}
%
%    \begin{macrocode}
\else
%    \end{macrocode}
%
% With \hologo{eTeX}, everything is much easier, since the engine knows
% how to subtract~$1$.
%
% The main looping macros expect their arguments as an integer followed
% by a semicolon.  As long as the argument is at least~$2$, decrement
% it, and expand what follows.  Once the argument is~$1$ (or less: the
% macros are not meant to handle that case), call |\multiexpand@end| to clean up
% and stop looping.
%    \begin{macrocode}
\gdef \multiexpand@ #1;%
  {%
    \ifnum #1<2 \multiexpand@end \fi
    \expandafter \multiexpand@
    \the \numexpr #1-1\expandafter ;%
  }
\gdef \multiexpand@after #1;%
  {%
    \ifnum #1<2 \multiexpand@end \fi
    \expandafter \multiexpand@after
    \the \numexpr #1-1\expandafter ;\expandafter
  }
%    \end{macrocode}
% The looping macros are used within an overarching |\romannumeral|
% expansion, which we end with a |0| and a space, as well as the
% appropriate |\expandafter|.  Here, |#1| is |\fi| which needs to remain
% to close the conditional, |#2| is |\expandafter|, and there is a
% trailing |\expandafter| in the case of |\multiexpand@after|.
%    \begin{macrocode}
\gdef \multiexpand@end #1#2#3;{#10#2 }
%    \end{macrocode}
% Finally, user commands, used as |\romannumeral| |\multiexpand(after)|.
% Those evaluate their argument, and pass it to |\multiexpand@(after)|.
% The argument might contain |\par| tokens (who knows)
%    \begin{macrocode}
\long \gdef \multiexpand #1%
  {\expandafter \multiexpand@ \the \numexpr #1;}
\long \gdef \multiexpandafter #1%
  {\expandafter \multiexpand@after \the \numexpr #1;}
%    \end{macrocode}
%
%    \begin{macrocode}
\fi
%    \end{macrocode}
%
% Close the group.
%    \begin{macrocode}
\endgroup
%    \end{macrocode}
%
%    \begin{macrocode}
%</package>
%    \end{macrocode}
%\Finale
\endinput