Commit | Line | Data |
---|---|---|
b8d4c8d0 GM |
1 | @c -*-texinfo-*- |
2 | @c This is part of the GNU Emacs Lisp Reference Manual. | |
1e103a7c | 3 | @c Copyright (C) 1990-1995, 1998, 2001-2012 Free Software Foundation, Inc. |
b8d4c8d0 | 4 | @c See the file elisp.texi for copying conditions. |
b8d4c8d0 GM |
5 | @node Macros, Customization, Functions, Top |
6 | @chapter Macros | |
7 | @cindex macros | |
8 | ||
9 | @dfn{Macros} enable you to define new control constructs and other | |
10 | language features. A macro is defined much like a function, but instead | |
11 | of telling how to compute a value, it tells how to compute another Lisp | |
12 | expression which will in turn compute the value. We call this | |
13 | expression the @dfn{expansion} of the macro. | |
14 | ||
15 | Macros can do this because they operate on the unevaluated expressions | |
16 | for the arguments, not on the argument values as functions do. They can | |
17 | therefore construct an expansion containing these argument expressions | |
18 | or parts of them. | |
19 | ||
20 | If you are using a macro to do something an ordinary function could | |
21 | do, just for the sake of speed, consider using an inline function | |
22 | instead. @xref{Inline Functions}. | |
23 | ||
24 | @menu | |
25 | * Simple Macro:: A basic example. | |
26 | * Expansion:: How, when and why macros are expanded. | |
27 | * Compiling Macros:: How macros are expanded by the compiler. | |
28 | * Defining Macros:: How to write a macro definition. | |
b8d4c8d0 GM |
29 | * Problems with Macros:: Don't evaluate the macro arguments too many times. |
30 | Don't hide the user's variables. | |
31 | * Indenting Macros:: Specifying how to indent macro calls. | |
32 | @end menu | |
33 | ||
34 | @node Simple Macro | |
35 | @section A Simple Example of a Macro | |
36 | ||
37 | Suppose we would like to define a Lisp construct to increment a | |
38 | variable value, much like the @code{++} operator in C. We would like to | |
39 | write @code{(inc x)} and have the effect of @code{(setq x (1+ x))}. | |
40 | Here's a macro definition that does the job: | |
41 | ||
42 | @findex inc | |
43 | @example | |
44 | @group | |
45 | (defmacro inc (var) | |
46 | (list 'setq var (list '1+ var))) | |
47 | @end group | |
48 | @end example | |
49 | ||
50 | When this is called with @code{(inc x)}, the argument @var{var} is the | |
51 | symbol @code{x}---@emph{not} the @emph{value} of @code{x}, as it would | |
52 | be in a function. The body of the macro uses this to construct the | |
53 | expansion, which is @code{(setq x (1+ x))}. Once the macro definition | |
54 | returns this expansion, Lisp proceeds to evaluate it, thus incrementing | |
55 | @code{x}. | |
56 | ||
57 | @node Expansion | |
58 | @section Expansion of a Macro Call | |
59 | @cindex expansion of macros | |
60 | @cindex macro call | |
61 | ||
62 | A macro call looks just like a function call in that it is a list which | |
63 | starts with the name of the macro. The rest of the elements of the list | |
64 | are the arguments of the macro. | |
65 | ||
66 | Evaluation of the macro call begins like evaluation of a function call | |
67 | except for one crucial difference: the macro arguments are the actual | |
68 | expressions appearing in the macro call. They are not evaluated before | |
69 | they are given to the macro definition. By contrast, the arguments of a | |
70 | function are results of evaluating the elements of the function call | |
71 | list. | |
72 | ||
73 | Having obtained the arguments, Lisp invokes the macro definition just | |
74 | as a function is invoked. The argument variables of the macro are bound | |
75 | to the argument values from the macro call, or to a list of them in the | |
76 | case of a @code{&rest} argument. And the macro body executes and | |
77 | returns its value just as a function body does. | |
78 | ||
03988c98 CY |
79 | The second crucial difference between macros and functions is that |
80 | the value returned by the macro body is an alternate Lisp expression, | |
81 | also known as the @dfn{expansion} of the macro. The Lisp interpreter | |
b8d4c8d0 GM |
82 | proceeds to evaluate the expansion as soon as it comes back from the |
83 | macro. | |
84 | ||
85 | Since the expansion is evaluated in the normal manner, it may contain | |
86 | calls to other macros. It may even be a call to the same macro, though | |
87 | this is unusual. | |
88 | ||
89 | You can see the expansion of a given macro call by calling | |
90 | @code{macroexpand}. | |
91 | ||
92 | @defun macroexpand form &optional environment | |
93 | @cindex macro expansion | |
94 | This function expands @var{form}, if it is a macro call. If the result | |
95 | is another macro call, it is expanded in turn, until something which is | |
96 | not a macro call results. That is the value returned by | |
97 | @code{macroexpand}. If @var{form} is not a macro call to begin with, it | |
98 | is returned as given. | |
99 | ||
100 | Note that @code{macroexpand} does not look at the subexpressions of | |
101 | @var{form} (although some macro definitions may do so). Even if they | |
102 | are macro calls themselves, @code{macroexpand} does not expand them. | |
103 | ||
104 | The function @code{macroexpand} does not expand calls to inline functions. | |
105 | Normally there is no need for that, since a call to an inline function is | |
106 | no harder to understand than a call to an ordinary function. | |
107 | ||
108 | If @var{environment} is provided, it specifies an alist of macro | |
109 | definitions that shadow the currently defined macros. Byte compilation | |
110 | uses this feature. | |
111 | ||
ddff3351 | 112 | @example |
b8d4c8d0 GM |
113 | @group |
114 | (defmacro inc (var) | |
115 | (list 'setq var (list '1+ var))) | |
116 | @result{} inc | |
117 | @end group | |
118 | ||
119 | @group | |
120 | (macroexpand '(inc r)) | |
121 | @result{} (setq r (1+ r)) | |
122 | @end group | |
123 | ||
124 | @group | |
125 | (defmacro inc2 (var1 var2) | |
126 | (list 'progn (list 'inc var1) (list 'inc var2))) | |
127 | @result{} inc2 | |
128 | @end group | |
129 | ||
130 | @group | |
131 | (macroexpand '(inc2 r s)) | |
132 | @result{} (progn (inc r) (inc s)) ; @r{@code{inc} not expanded here.} | |
133 | @end group | |
ddff3351 | 134 | @end example |
b8d4c8d0 GM |
135 | @end defun |
136 | ||
137 | ||
138 | @defun macroexpand-all form &optional environment | |
139 | @code{macroexpand-all} expands macros like @code{macroexpand}, but | |
140 | will look for and expand all macros in @var{form}, not just at the | |
141 | top-level. If no macros are expanded, the return value is @code{eq} | |
142 | to @var{form}. | |
143 | ||
144 | Repeating the example used for @code{macroexpand} above with | |
145 | @code{macroexpand-all}, we see that @code{macroexpand-all} @emph{does} | |
146 | expand the embedded calls to @code{inc}: | |
147 | ||
ddff3351 | 148 | @example |
b8d4c8d0 GM |
149 | (macroexpand-all '(inc2 r s)) |
150 | @result{} (progn (setq r (1+ r)) (setq s (1+ s))) | |
ddff3351 | 151 | @end example |
b8d4c8d0 GM |
152 | |
153 | @end defun | |
154 | ||
155 | @node Compiling Macros | |
156 | @section Macros and Byte Compilation | |
157 | @cindex byte-compiling macros | |
158 | ||
159 | You might ask why we take the trouble to compute an expansion for a | |
160 | macro and then evaluate the expansion. Why not have the macro body | |
161 | produce the desired results directly? The reason has to do with | |
162 | compilation. | |
163 | ||
164 | When a macro call appears in a Lisp program being compiled, the Lisp | |
165 | compiler calls the macro definition just as the interpreter would, and | |
166 | receives an expansion. But instead of evaluating this expansion, it | |
167 | compiles the expansion as if it had appeared directly in the program. | |
168 | As a result, the compiled code produces the value and side effects | |
169 | intended for the macro, but executes at full compiled speed. This would | |
170 | not work if the macro body computed the value and side effects | |
171 | itself---they would be computed at compile time, which is not useful. | |
172 | ||
173 | In order for compilation of macro calls to work, the macros must | |
174 | already be defined in Lisp when the calls to them are compiled. The | |
175 | compiler has a special feature to help you do this: if a file being | |
176 | compiled contains a @code{defmacro} form, the macro is defined | |
ebfbce67 CY |
177 | temporarily for the rest of the compilation of that file. |
178 | ||
179 | Byte-compiling a file also executes any @code{require} calls at | |
180 | top-level in the file, so you can ensure that necessary macro | |
181 | definitions are available during compilation by requiring the files | |
182 | that define them (@pxref{Named Features}). To avoid loading the macro | |
183 | definition files when someone @emph{runs} the compiled program, write | |
b8d4c8d0 GM |
184 | @code{eval-when-compile} around the @code{require} calls (@pxref{Eval |
185 | During Compile}). | |
186 | ||
187 | @node Defining Macros | |
188 | @section Defining Macros | |
189 | ||
190 | A Lisp macro is a list whose @sc{car} is @code{macro}. Its @sc{cdr} should | |
191 | be a function; expansion of the macro works by applying the function | |
192 | (with @code{apply}) to the list of unevaluated argument-expressions | |
193 | from the macro call. | |
194 | ||
195 | It is possible to use an anonymous Lisp macro just like an anonymous | |
196 | function, but this is never done, because it does not make sense to pass | |
197 | an anonymous macro to functionals such as @code{mapcar}. In practice, | |
198 | all Lisp macros have names, and they are usually defined with the | |
199 | special form @code{defmacro}. | |
200 | ||
201 | @defspec defmacro name argument-list body-forms@dots{} | |
202 | @code{defmacro} defines the symbol @var{name} as a macro that looks | |
203 | like this: | |
204 | ||
205 | @example | |
206 | (macro lambda @var{argument-list} . @var{body-forms}) | |
207 | @end example | |
208 | ||
209 | (Note that the @sc{cdr} of this list is a function---a lambda expression.) | |
210 | This macro object is stored in the function cell of @var{name}. The | |
211 | value returned by evaluating the @code{defmacro} form is @var{name}, but | |
212 | usually we ignore this value. | |
213 | ||
214 | The shape and meaning of @var{argument-list} is the same as in a | |
215 | function, and the keywords @code{&rest} and @code{&optional} may be used | |
216 | (@pxref{Argument List}). Macros may have a documentation string, but | |
217 | any @code{interactive} declaration is ignored since macros cannot be | |
218 | called interactively. | |
219 | @end defspec | |
220 | ||
03988c98 CY |
221 | Macros often need to construct large list structures from a mixture |
222 | of constants and nonconstant parts. To make this easier, use the | |
223 | @samp{`} syntax (@pxref{Backquote}). For example: | |
224 | ||
225 | @example | |
226 | @example | |
227 | @group | |
228 | (defmacro t-becomes-nil (variable) | |
229 | `(if (eq ,variable t) | |
230 | (setq ,variable nil))) | |
231 | @end group | |
232 | ||
233 | @group | |
234 | (t-becomes-nil foo) | |
235 | @equiv{} (if (eq foo t) (setq foo nil)) | |
236 | @end group | |
237 | @end example | |
238 | @end example | |
239 | ||
240 | The body of a macro definition can include a @code{declare} form, | |
b8d4c8d0 GM |
241 | which can specify how @key{TAB} should indent macro calls, and how to |
242 | step through them for Edebug. | |
243 | ||
244 | @defmac declare @var{specs}@dots{} | |
245 | @anchor{Definition of declare} | |
246 | A @code{declare} form is used in a macro definition to specify various | |
543757a8 | 247 | additional information about it. The following specifications are |
b8d4c8d0 GM |
248 | currently supported: |
249 | ||
250 | @table @code | |
251 | @item (debug @var{edebug-form-spec}) | |
252 | Specify how to step through macro calls for Edebug. | |
253 | @xref{Instrumenting Macro Calls}. | |
254 | ||
255 | @item (indent @var{indent-spec}) | |
256 | Specify how to indent calls to this macro. @xref{Indenting Macros}, | |
257 | for more details. | |
d027f83d LMI |
258 | |
259 | @item (doc-string @var{number}) | |
8376d7c2 CY |
260 | Specify which element of the macro is the documentation string, if |
261 | any. | |
b8d4c8d0 GM |
262 | @end table |
263 | ||
264 | A @code{declare} form only has its special effect in the body of a | |
265 | @code{defmacro} form if it immediately follows the documentation | |
266 | string, if present, or the argument list otherwise. (Strictly | |
267 | speaking, @emph{several} @code{declare} forms can follow the | |
268 | documentation string or argument list, but since a @code{declare} form | |
269 | can have several @var{specs}, they can always be combined into a | |
270 | single form.) When used at other places in a @code{defmacro} form, or | |
271 | outside a @code{defmacro} form, @code{declare} just returns @code{nil} | |
272 | without evaluating any @var{specs}. | |
273 | @end defmac | |
274 | ||
275 | No macro absolutely needs a @code{declare} form, because that form | |
276 | has no effect on how the macro expands, on what the macro means in the | |
69b0acb9 | 277 | program. It only affects the secondary features listed above. |
b8d4c8d0 | 278 | |
b8d4c8d0 GM |
279 | @node Problems with Macros |
280 | @section Common Problems Using Macros | |
281 | ||
03988c98 CY |
282 | Macro expansion can have counterintuitive consequences. This |
283 | section describes some important consequences that can lead to | |
b8d4c8d0 GM |
284 | trouble, and rules to follow to avoid trouble. |
285 | ||
286 | @menu | |
287 | * Wrong Time:: Do the work in the expansion, not in the macro. | |
288 | * Argument Evaluation:: The expansion should evaluate each macro arg once. | |
289 | * Surprising Local Vars:: Local variable bindings in the expansion | |
290 | require special care. | |
291 | * Eval During Expansion:: Don't evaluate them; put them in the expansion. | |
292 | * Repeated Expansion:: Avoid depending on how many times expansion is done. | |
293 | @end menu | |
294 | ||
295 | @node Wrong Time | |
296 | @subsection Wrong Time | |
297 | ||
298 | The most common problem in writing macros is doing some of the | |
299 | real work prematurely---while expanding the macro, rather than in the | |
300 | expansion itself. For instance, one real package had this macro | |
301 | definition: | |
302 | ||
303 | @example | |
304 | (defmacro my-set-buffer-multibyte (arg) | |
305 | (if (fboundp 'set-buffer-multibyte) | |
306 | (set-buffer-multibyte arg))) | |
307 | @end example | |
308 | ||
309 | With this erroneous macro definition, the program worked fine when | |
310 | interpreted but failed when compiled. This macro definition called | |
311 | @code{set-buffer-multibyte} during compilation, which was wrong, and | |
312 | then did nothing when the compiled package was run. The definition | |
313 | that the programmer really wanted was this: | |
314 | ||
315 | @example | |
316 | (defmacro my-set-buffer-multibyte (arg) | |
317 | (if (fboundp 'set-buffer-multibyte) | |
318 | `(set-buffer-multibyte ,arg))) | |
319 | @end example | |
320 | ||
321 | @noindent | |
322 | This macro expands, if appropriate, into a call to | |
323 | @code{set-buffer-multibyte} that will be executed when the compiled | |
324 | program is actually run. | |
325 | ||
326 | @node Argument Evaluation | |
327 | @subsection Evaluating Macro Arguments Repeatedly | |
328 | ||
329 | When defining a macro you must pay attention to the number of times | |
330 | the arguments will be evaluated when the expansion is executed. The | |
03988c98 CY |
331 | following macro (used to facilitate iteration) illustrates the |
332 | problem. This macro allows us to write a ``for'' loop construct. | |
b8d4c8d0 GM |
333 | |
334 | @findex for | |
ddff3351 | 335 | @example |
b8d4c8d0 GM |
336 | @group |
337 | (defmacro for (var from init to final do &rest body) | |
338 | "Execute a simple \"for\" loop. | |
339 | For example, (for i from 1 to 10 do (print i))." | |
340 | (list 'let (list (list var init)) | |
341 | (cons 'while (cons (list '<= var final) | |
342 | (append body (list (list 'inc var))))))) | |
343 | @end group | |
344 | @result{} for | |
345 | ||
346 | @group | |
347 | (for i from 1 to 3 do | |
348 | (setq square (* i i)) | |
349 | (princ (format "\n%d %d" i square))) | |
350 | @expansion{} | |
351 | @end group | |
352 | @group | |
353 | (let ((i 1)) | |
354 | (while (<= i 3) | |
355 | (setq square (* i i)) | |
356 | (princ (format "\n%d %d" i square)) | |
357 | (inc i))) | |
358 | @end group | |
359 | @group | |
360 | ||
361 | @print{}1 1 | |
362 | @print{}2 4 | |
363 | @print{}3 9 | |
364 | @result{} nil | |
365 | @end group | |
ddff3351 | 366 | @end example |
b8d4c8d0 GM |
367 | |
368 | @noindent | |
369 | The arguments @code{from}, @code{to}, and @code{do} in this macro are | |
370 | ``syntactic sugar''; they are entirely ignored. The idea is that you | |
371 | will write noise words (such as @code{from}, @code{to}, and @code{do}) | |
372 | in those positions in the macro call. | |
373 | ||
374 | Here's an equivalent definition simplified through use of backquote: | |
375 | ||
ddff3351 | 376 | @example |
b8d4c8d0 GM |
377 | @group |
378 | (defmacro for (var from init to final do &rest body) | |
379 | "Execute a simple \"for\" loop. | |
380 | For example, (for i from 1 to 10 do (print i))." | |
381 | `(let ((,var ,init)) | |
382 | (while (<= ,var ,final) | |
383 | ,@@body | |
384 | (inc ,var)))) | |
385 | @end group | |
ddff3351 | 386 | @end example |
b8d4c8d0 GM |
387 | |
388 | Both forms of this definition (with backquote and without) suffer from | |
389 | the defect that @var{final} is evaluated on every iteration. If | |
390 | @var{final} is a constant, this is not a problem. If it is a more | |
391 | complex form, say @code{(long-complex-calculation x)}, this can slow | |
392 | down the execution significantly. If @var{final} has side effects, | |
393 | executing it more than once is probably incorrect. | |
394 | ||
395 | @cindex macro argument evaluation | |
396 | A well-designed macro definition takes steps to avoid this problem by | |
397 | producing an expansion that evaluates the argument expressions exactly | |
398 | once unless repeated evaluation is part of the intended purpose of the | |
399 | macro. Here is a correct expansion for the @code{for} macro: | |
400 | ||
ddff3351 | 401 | @example |
b8d4c8d0 GM |
402 | @group |
403 | (let ((i 1) | |
404 | (max 3)) | |
405 | (while (<= i max) | |
406 | (setq square (* i i)) | |
407 | (princ (format "%d %d" i square)) | |
408 | (inc i))) | |
409 | @end group | |
ddff3351 | 410 | @end example |
b8d4c8d0 GM |
411 | |
412 | Here is a macro definition that creates this expansion: | |
413 | ||
ddff3351 | 414 | @example |
b8d4c8d0 GM |
415 | @group |
416 | (defmacro for (var from init to final do &rest body) | |
417 | "Execute a simple for loop: (for i from 1 to 10 do (print i))." | |
418 | `(let ((,var ,init) | |
419 | (max ,final)) | |
420 | (while (<= ,var max) | |
421 | ,@@body | |
422 | (inc ,var)))) | |
423 | @end group | |
ddff3351 | 424 | @end example |
b8d4c8d0 GM |
425 | |
426 | Unfortunately, this fix introduces another problem, | |
427 | described in the following section. | |
428 | ||
429 | @node Surprising Local Vars | |
430 | @subsection Local Variables in Macro Expansions | |
431 | ||
432 | @ifnottex | |
433 | In the previous section, the definition of @code{for} was fixed as | |
434 | follows to make the expansion evaluate the macro arguments the proper | |
435 | number of times: | |
436 | ||
ddff3351 | 437 | @example |
b8d4c8d0 GM |
438 | @group |
439 | (defmacro for (var from init to final do &rest body) | |
440 | "Execute a simple for loop: (for i from 1 to 10 do (print i))." | |
441 | @end group | |
442 | @group | |
443 | `(let ((,var ,init) | |
444 | (max ,final)) | |
445 | (while (<= ,var max) | |
446 | ,@@body | |
447 | (inc ,var)))) | |
448 | @end group | |
ddff3351 | 449 | @end example |
b8d4c8d0 GM |
450 | @end ifnottex |
451 | ||
452 | The new definition of @code{for} has a new problem: it introduces a | |
453 | local variable named @code{max} which the user does not expect. This | |
454 | causes trouble in examples such as the following: | |
455 | ||
ddff3351 | 456 | @example |
b8d4c8d0 GM |
457 | @group |
458 | (let ((max 0)) | |
459 | (for x from 0 to 10 do | |
460 | (let ((this (frob x))) | |
461 | (if (< max this) | |
462 | (setq max this))))) | |
463 | @end group | |
ddff3351 | 464 | @end example |
b8d4c8d0 GM |
465 | |
466 | @noindent | |
467 | The references to @code{max} inside the body of the @code{for}, which | |
468 | are supposed to refer to the user's binding of @code{max}, really access | |
469 | the binding made by @code{for}. | |
470 | ||
471 | The way to correct this is to use an uninterned symbol instead of | |
472 | @code{max} (@pxref{Creating Symbols}). The uninterned symbol can be | |
473 | bound and referred to just like any other symbol, but since it is | |
474 | created by @code{for}, we know that it cannot already appear in the | |
475 | user's program. Since it is not interned, there is no way the user can | |
476 | put it into the program later. It will never appear anywhere except | |
477 | where put by @code{for}. Here is a definition of @code{for} that works | |
478 | this way: | |
479 | ||
ddff3351 | 480 | @example |
b8d4c8d0 GM |
481 | @group |
482 | (defmacro for (var from init to final do &rest body) | |
483 | "Execute a simple for loop: (for i from 1 to 10 do (print i))." | |
484 | (let ((tempvar (make-symbol "max"))) | |
485 | `(let ((,var ,init) | |
486 | (,tempvar ,final)) | |
487 | (while (<= ,var ,tempvar) | |
488 | ,@@body | |
489 | (inc ,var))))) | |
490 | @end group | |
ddff3351 | 491 | @end example |
b8d4c8d0 GM |
492 | |
493 | @noindent | |
494 | This creates an uninterned symbol named @code{max} and puts it in the | |
495 | expansion instead of the usual interned symbol @code{max} that appears | |
496 | in expressions ordinarily. | |
497 | ||
498 | @node Eval During Expansion | |
499 | @subsection Evaluating Macro Arguments in Expansion | |
500 | ||
501 | Another problem can happen if the macro definition itself | |
502 | evaluates any of the macro argument expressions, such as by calling | |
503 | @code{eval} (@pxref{Eval}). If the argument is supposed to refer to the | |
504 | user's variables, you may have trouble if the user happens to use a | |
505 | variable with the same name as one of the macro arguments. Inside the | |
506 | macro body, the macro argument binding is the most local binding of this | |
507 | variable, so any references inside the form being evaluated do refer to | |
508 | it. Here is an example: | |
509 | ||
510 | @example | |
511 | @group | |
512 | (defmacro foo (a) | |
513 | (list 'setq (eval a) t)) | |
514 | @result{} foo | |
515 | @end group | |
516 | @group | |
517 | (setq x 'b) | |
518 | (foo x) @expansion{} (setq b t) | |
519 | @result{} t ; @r{and @code{b} has been set.} | |
520 | ;; @r{but} | |
521 | (setq a 'c) | |
522 | (foo a) @expansion{} (setq a t) | |
523 | @result{} t ; @r{but this set @code{a}, not @code{c}.} | |
524 | ||
525 | @end group | |
526 | @end example | |
527 | ||
528 | It makes a difference whether the user's variable is named @code{a} or | |
529 | @code{x}, because @code{a} conflicts with the macro argument variable | |
530 | @code{a}. | |
531 | ||
532 | Another problem with calling @code{eval} in a macro definition is that | |
533 | it probably won't do what you intend in a compiled program. The | |
cc060ff7 | 534 | byte compiler runs macro definitions while compiling the program, when |
b8d4c8d0 GM |
535 | the program's own computations (which you might have wished to access |
536 | with @code{eval}) don't occur and its local variable bindings don't | |
537 | exist. | |
538 | ||
539 | To avoid these problems, @strong{don't evaluate an argument expression | |
540 | while computing the macro expansion}. Instead, substitute the | |
541 | expression into the macro expansion, so that its value will be computed | |
542 | as part of executing the expansion. This is how the other examples in | |
543 | this chapter work. | |
544 | ||
545 | @node Repeated Expansion | |
546 | @subsection How Many Times is the Macro Expanded? | |
547 | ||
548 | Occasionally problems result from the fact that a macro call is | |
549 | expanded each time it is evaluated in an interpreted function, but is | |
550 | expanded only once (during compilation) for a compiled function. If the | |
551 | macro definition has side effects, they will work differently depending | |
552 | on how many times the macro is expanded. | |
553 | ||
554 | Therefore, you should avoid side effects in computation of the | |
555 | macro expansion, unless you really know what you are doing. | |
556 | ||
557 | One special kind of side effect can't be avoided: constructing Lisp | |
558 | objects. Almost all macro expansions include constructed lists; that is | |
559 | the whole point of most macros. This is usually safe; there is just one | |
560 | case where you must be careful: when the object you construct is part of a | |
561 | quoted constant in the macro expansion. | |
562 | ||
563 | If the macro is expanded just once, in compilation, then the object is | |
564 | constructed just once, during compilation. But in interpreted | |
565 | execution, the macro is expanded each time the macro call runs, and this | |
566 | means a new object is constructed each time. | |
567 | ||
568 | In most clean Lisp code, this difference won't matter. It can matter | |
569 | only if you perform side-effects on the objects constructed by the macro | |
570 | definition. Thus, to avoid trouble, @strong{avoid side effects on | |
571 | objects constructed by macro definitions}. Here is an example of how | |
572 | such side effects can get you into trouble: | |
573 | ||
574 | @lisp | |
575 | @group | |
576 | (defmacro empty-object () | |
577 | (list 'quote (cons nil nil))) | |
578 | @end group | |
579 | ||
580 | @group | |
581 | (defun initialize (condition) | |
582 | (let ((object (empty-object))) | |
583 | (if condition | |
584 | (setcar object condition)) | |
585 | object)) | |
586 | @end group | |
587 | @end lisp | |
588 | ||
589 | @noindent | |
590 | If @code{initialize} is interpreted, a new list @code{(nil)} is | |
591 | constructed each time @code{initialize} is called. Thus, no side effect | |
592 | survives between calls. If @code{initialize} is compiled, then the | |
593 | macro @code{empty-object} is expanded during compilation, producing a | |
594 | single ``constant'' @code{(nil)} that is reused and altered each time | |
595 | @code{initialize} is called. | |
596 | ||
597 | One way to avoid pathological cases like this is to think of | |
598 | @code{empty-object} as a funny kind of constant, not as a memory | |
599 | allocation construct. You wouldn't use @code{setcar} on a constant such | |
600 | as @code{'(nil)}, so naturally you won't use it on @code{(empty-object)} | |
601 | either. | |
602 | ||
603 | @node Indenting Macros | |
604 | @section Indenting Macros | |
605 | ||
03988c98 | 606 | Within a macro definition, you can use the @code{declare} form |
a2715669 | 607 | (@pxref{Defining Macros}) to specify how @key{TAB} should indent |
34706efa | 608 | calls to the macro. An indentation specification is written like this: |
b8d4c8d0 GM |
609 | |
610 | @example | |
611 | (declare (indent @var{indent-spec})) | |
612 | @end example | |
613 | ||
614 | @noindent | |
615 | Here are the possibilities for @var{indent-spec}: | |
616 | ||
617 | @table @asis | |
618 | @item @code{nil} | |
619 | This is the same as no property---use the standard indentation pattern. | |
620 | @item @code{defun} | |
621 | Handle this function like a @samp{def} construct: treat the second | |
622 | line as the start of a @dfn{body}. | |
623 | @item an integer, @var{number} | |
624 | The first @var{number} arguments of the function are | |
625 | @dfn{distinguished} arguments; the rest are considered the body | |
626 | of the expression. A line in the expression is indented according to | |
627 | whether the first argument on it is distinguished or not. If the | |
628 | argument is part of the body, the line is indented @code{lisp-body-indent} | |
629 | more columns than the open-parenthesis starting the containing | |
630 | expression. If the argument is distinguished and is either the first | |
631 | or second argument, it is indented @emph{twice} that many extra columns. | |
632 | If the argument is distinguished and not the first or second argument, | |
633 | the line uses the standard pattern. | |
634 | @item a symbol, @var{symbol} | |
635 | @var{symbol} should be a function name; that function is called to | |
636 | calculate the indentation of a line within this expression. The | |
637 | function receives two arguments: | |
03988c98 | 638 | |
b8d4c8d0 GM |
639 | @table @asis |
640 | @item @var{state} | |
641 | The value returned by @code{parse-partial-sexp} (a Lisp primitive for | |
642 | indentation and nesting computation) when it parses up to the | |
643 | beginning of this line. | |
644 | @item @var{pos} | |
645 | The position at which the line being indented begins. | |
646 | @end table | |
03988c98 | 647 | |
b8d4c8d0 GM |
648 | @noindent |
649 | It should return either a number, which is the number of columns of | |
650 | indentation for that line, or a list whose car is such a number. The | |
651 | difference between returning a number and returning a list is that a | |
652 | number says that all following lines at the same nesting level should | |
653 | be indented just like this one; a list says that following lines might | |
654 | call for different indentations. This makes a difference when the | |
655 | indentation is being computed by @kbd{C-M-q}; if the value is a | |
656 | number, @kbd{C-M-q} need not recalculate indentation for the following | |
657 | lines until the end of the list. | |
658 | @end table |