Commit | Line | Data |
---|---|---|
b8d4c8d0 GM |
1 | @c -*-texinfo-*- |
2 | @c This is part of the GNU Emacs Lisp Reference Manual. | |
3 | @c Copyright (C) 1990, 1991, 1992, 1993, 1994, 1998, 1999, 2001, 2002, 2003, | |
6ed161e1 | 4 | @c 2004, 2005, 2006, 2007, 2008, 2009 Free Software Foundation, Inc. |
b8d4c8d0 | 5 | @c See the file elisp.texi for copying conditions. |
6336d8c3 | 6 | @setfilename ../../info/debugging |
b8d4c8d0 GM |
7 | @node Debugging, Read and Print, Advising Functions, Top |
8 | @chapter Debugging Lisp Programs | |
9 | ||
10 | There are three ways to investigate a problem in an Emacs Lisp program, | |
11 | depending on what you are doing with the program when the problem appears. | |
12 | ||
13 | @itemize @bullet | |
14 | @item | |
15 | If the problem occurs when you run the program, you can use a Lisp | |
16 | debugger to investigate what is happening during execution. In addition | |
17 | to the ordinary debugger, Emacs comes with a source-level debugger, | |
18 | Edebug. This chapter describes both of them. | |
19 | ||
20 | @item | |
21 | If the problem is syntactic, so that Lisp cannot even read the program, | |
22 | you can use the Emacs facilities for editing Lisp to localize it. | |
23 | ||
24 | @item | |
25 | If the problem occurs when trying to compile the program with the byte | |
26 | compiler, you need to know how to examine the compiler's input buffer. | |
27 | @end itemize | |
28 | ||
29 | @menu | |
30 | * Debugger:: How the Emacs Lisp debugger is implemented. | |
31 | * Edebug:: A source-level Emacs Lisp debugger. | |
32 | * Syntax Errors:: How to find syntax errors. | |
33 | * Test Coverage:: Ensuring you have tested all branches in your code. | |
34 | * Compilation Errors:: How to find errors that show up in byte compilation. | |
35 | @end menu | |
36 | ||
37 | Another useful debugging tool is the dribble file. When a dribble | |
38 | file is open, Emacs copies all keyboard input characters to that file. | |
39 | Afterward, you can examine the file to find out what input was used. | |
40 | @xref{Terminal Input}. | |
41 | ||
42 | For debugging problems in terminal descriptions, the | |
43 | @code{open-termscript} function can be useful. @xref{Terminal Output}. | |
44 | ||
45 | @node Debugger | |
46 | @section The Lisp Debugger | |
47 | @cindex debugger for Emacs Lisp | |
48 | @cindex Lisp debugger | |
49 | @cindex break | |
50 | ||
51 | The ordinary @dfn{Lisp debugger} provides the ability to suspend | |
52 | evaluation of a form. While evaluation is suspended (a state that is | |
53 | commonly known as a @dfn{break}), you may examine the run time stack, | |
54 | examine the values of local or global variables, or change those values. | |
55 | Since a break is a recursive edit, all the usual editing facilities of | |
56 | Emacs are available; you can even run programs that will enter the | |
57 | debugger recursively. @xref{Recursive Editing}. | |
58 | ||
59 | @menu | |
60 | * Error Debugging:: Entering the debugger when an error happens. | |
61 | * Infinite Loops:: Stopping and debugging a program that doesn't exit. | |
62 | * Function Debugging:: Entering it when a certain function is called. | |
63 | * Explicit Debug:: Entering it at a certain point in the program. | |
64 | * Using Debugger:: What the debugger does; what you see while in it. | |
65 | * Debugger Commands:: Commands used while in the debugger. | |
66 | * Invoking the Debugger:: How to call the function @code{debug}. | |
67 | * Internals of Debugger:: Subroutines of the debugger, and global variables. | |
68 | @end menu | |
69 | ||
70 | @node Error Debugging | |
71 | @subsection Entering the Debugger on an Error | |
72 | @cindex error debugging | |
73 | @cindex debugging errors | |
74 | ||
75 | The most important time to enter the debugger is when a Lisp error | |
76 | happens. This allows you to investigate the immediate causes of the | |
77 | error. | |
78 | ||
79 | However, entry to the debugger is not a normal consequence of an | |
80 | error. Many commands frequently cause Lisp errors when invoked | |
81 | inappropriately (such as @kbd{C-f} at the end of the buffer), and during | |
82 | ordinary editing it would be very inconvenient to enter the debugger | |
83 | each time this happens. So if you want errors to enter the debugger, set | |
84 | the variable @code{debug-on-error} to non-@code{nil}. (The command | |
85 | @code{toggle-debug-on-error} provides an easy way to do this.) | |
86 | ||
87 | @defopt debug-on-error | |
88 | This variable determines whether the debugger is called when an error is | |
89 | signaled and not handled. If @code{debug-on-error} is @code{t}, all | |
90 | kinds of errors call the debugger (except those listed in | |
91 | @code{debug-ignored-errors}). If it is @code{nil}, none call the | |
21c58e4d GM |
92 | debugger. (Note that @code{eval-expression-debug-on-error} affects the |
93 | setting of this variable in some cases; see below.) | |
b8d4c8d0 GM |
94 | |
95 | The value can also be a list of error conditions that should call the | |
96 | debugger. For example, if you set it to the list | |
97 | @code{(void-variable)}, then only errors about a variable that has no | |
98 | value invoke the debugger. | |
99 | ||
100 | When this variable is non-@code{nil}, Emacs does not create an error | |
101 | handler around process filter functions and sentinels. Therefore, | |
102 | errors in these functions also invoke the debugger. @xref{Processes}. | |
103 | @end defopt | |
104 | ||
105 | @defopt debug-ignored-errors | |
106 | This variable specifies certain kinds of errors that should not enter | |
107 | the debugger. Its value is a list of error condition symbols and/or | |
108 | regular expressions. If the error has any of those condition symbols, | |
109 | or if the error message matches any of the regular expressions, then | |
110 | that error does not enter the debugger, regardless of the value of | |
111 | @code{debug-on-error}. | |
112 | ||
113 | The normal value of this variable lists several errors that happen often | |
114 | during editing but rarely result from bugs in Lisp programs. However, | |
115 | ``rarely'' is not ``never''; if your program fails with an error that | |
116 | matches this list, you will need to change this list in order to debug | |
117 | the error. The easiest way is usually to set | |
118 | @code{debug-ignored-errors} to @code{nil}. | |
119 | @end defopt | |
120 | ||
121 | @defopt eval-expression-debug-on-error | |
122 | If this variable has a non-@code{nil} value, then | |
123 | @code{debug-on-error} is set to @code{t} when evaluating with the | |
124 | command @code{eval-expression}. If | |
125 | @code{eval-expression-debug-on-error} is @code{nil}, then the value of | |
126 | @code{debug-on-error} is not changed. @xref{Lisp Eval,, Evaluating | |
127 | Emacs-Lisp Expressions, emacs, The GNU Emacs Manual}. | |
128 | @end defopt | |
129 | ||
130 | @defopt debug-on-signal | |
131 | Normally, errors that are caught by @code{condition-case} never run the | |
132 | debugger, even if @code{debug-on-error} is non-@code{nil}. In other | |
133 | words, @code{condition-case} gets a chance to handle the error before | |
134 | the debugger gets a chance. | |
135 | ||
136 | If you set @code{debug-on-signal} to a non-@code{nil} value, then the | |
137 | debugger gets the first chance at every error; an error will invoke the | |
138 | debugger regardless of any @code{condition-case}, if it fits the | |
139 | criteria specified by the values of @code{debug-on-error} and | |
140 | @code{debug-ignored-errors}. | |
141 | ||
142 | @strong{Warning:} This variable is strong medicine! Various parts of | |
143 | Emacs handle errors in the normal course of affairs, and you may not | |
144 | even realize that errors happen there. If you set | |
145 | @code{debug-on-signal} to a non-@code{nil} value, those errors will | |
146 | enter the debugger. | |
147 | ||
148 | @strong{Warning:} @code{debug-on-signal} has no effect when | |
149 | @code{debug-on-error} is @code{nil}. | |
150 | @end defopt | |
151 | ||
152 | To debug an error that happens during loading of the init | |
153 | file, use the option @samp{--debug-init}. This binds | |
154 | @code{debug-on-error} to @code{t} while loading the init file, and | |
155 | bypasses the @code{condition-case} which normally catches errors in the | |
156 | init file. | |
157 | ||
158 | If your init file sets @code{debug-on-error}, the effect may | |
159 | not last past the end of loading the init file. (This is an undesirable | |
160 | byproduct of the code that implements the @samp{--debug-init} command | |
161 | line option.) The best way to make the init file set | |
162 | @code{debug-on-error} permanently is with @code{after-init-hook}, like | |
163 | this: | |
164 | ||
165 | @example | |
166 | (add-hook 'after-init-hook | |
167 | (lambda () (setq debug-on-error t))) | |
168 | @end example | |
169 | ||
170 | @node Infinite Loops | |
171 | @subsection Debugging Infinite Loops | |
172 | @cindex infinite loops | |
173 | @cindex loops, infinite | |
174 | @cindex quitting from infinite loop | |
175 | @cindex stopping an infinite loop | |
176 | ||
177 | When a program loops infinitely and fails to return, your first | |
178 | problem is to stop the loop. On most operating systems, you can do this | |
179 | with @kbd{C-g}, which causes a @dfn{quit}. | |
180 | ||
181 | Ordinary quitting gives no information about why the program was | |
182 | looping. To get more information, you can set the variable | |
183 | @code{debug-on-quit} to non-@code{nil}. Quitting with @kbd{C-g} is not | |
184 | considered an error, and @code{debug-on-error} has no effect on the | |
185 | handling of @kbd{C-g}. Likewise, @code{debug-on-quit} has no effect on | |
186 | errors. | |
187 | ||
188 | Once you have the debugger running in the middle of the infinite loop, | |
189 | you can proceed from the debugger using the stepping commands. If you | |
190 | step through the entire loop, you will probably get enough information | |
191 | to solve the problem. | |
192 | ||
193 | @defopt debug-on-quit | |
194 | This variable determines whether the debugger is called when @code{quit} | |
195 | is signaled and not handled. If @code{debug-on-quit} is non-@code{nil}, | |
196 | then the debugger is called whenever you quit (that is, type @kbd{C-g}). | |
197 | If @code{debug-on-quit} is @code{nil}, then the debugger is not called | |
198 | when you quit. @xref{Quitting}. | |
199 | @end defopt | |
200 | ||
201 | @node Function Debugging | |
202 | @subsection Entering the Debugger on a Function Call | |
203 | @cindex function call debugging | |
204 | @cindex debugging specific functions | |
205 | ||
206 | To investigate a problem that happens in the middle of a program, one | |
207 | useful technique is to enter the debugger whenever a certain function is | |
208 | called. You can do this to the function in which the problem occurs, | |
209 | and then step through the function, or you can do this to a function | |
210 | called shortly before the problem, step quickly over the call to that | |
211 | function, and then step through its caller. | |
212 | ||
213 | @deffn Command debug-on-entry function-name | |
214 | This function requests @var{function-name} to invoke the debugger each | |
215 | time it is called. It works by inserting the form | |
216 | @code{(implement-debug-on-entry)} into the function definition as the | |
217 | first form. | |
218 | ||
219 | Any function or macro defined as Lisp code may be set to break on | |
220 | entry, regardless of whether it is interpreted code or compiled code. | |
221 | If the function is a command, it will enter the debugger when called | |
222 | from Lisp and when called interactively (after the reading of the | |
223 | arguments). You can also set debug-on-entry for primitive functions | |
224 | (i.e., those written in C) this way, but it only takes effect when the | |
225 | primitive is called from Lisp code. Debug-on-entry is not allowed for | |
226 | special forms. | |
227 | ||
228 | When @code{debug-on-entry} is called interactively, it prompts for | |
229 | @var{function-name} in the minibuffer. If the function is already set | |
230 | up to invoke the debugger on entry, @code{debug-on-entry} does nothing. | |
231 | @code{debug-on-entry} always returns @var{function-name}. | |
232 | ||
233 | @strong{Warning:} if you redefine a function after using | |
234 | @code{debug-on-entry} on it, the code to enter the debugger is | |
235 | discarded by the redefinition. In effect, redefining the function | |
236 | cancels the break-on-entry feature for that function. | |
237 | ||
238 | Here's an example to illustrate use of this function: | |
239 | ||
240 | @example | |
241 | @group | |
242 | (defun fact (n) | |
243 | (if (zerop n) 1 | |
244 | (* n (fact (1- n))))) | |
245 | @result{} fact | |
246 | @end group | |
247 | @group | |
248 | (debug-on-entry 'fact) | |
249 | @result{} fact | |
250 | @end group | |
251 | @group | |
252 | (fact 3) | |
253 | @end group | |
254 | ||
255 | @group | |
256 | ------ Buffer: *Backtrace* ------ | |
257 | Debugger entered--entering a function: | |
258 | * fact(3) | |
259 | eval((fact 3)) | |
260 | eval-last-sexp-1(nil) | |
261 | eval-last-sexp(nil) | |
262 | call-interactively(eval-last-sexp) | |
263 | ------ Buffer: *Backtrace* ------ | |
264 | @end group | |
265 | ||
266 | @group | |
267 | (symbol-function 'fact) | |
268 | @result{} (lambda (n) | |
269 | (debug (quote debug)) | |
270 | (if (zerop n) 1 (* n (fact (1- n))))) | |
271 | @end group | |
272 | @end example | |
273 | @end deffn | |
274 | ||
275 | @deffn Command cancel-debug-on-entry &optional function-name | |
276 | This function undoes the effect of @code{debug-on-entry} on | |
277 | @var{function-name}. When called interactively, it prompts for | |
278 | @var{function-name} in the minibuffer. If @var{function-name} is | |
279 | omitted or @code{nil}, it cancels break-on-entry for all functions. | |
280 | Calling @code{cancel-debug-on-entry} does nothing to a function which is | |
281 | not currently set up to break on entry. | |
282 | @end deffn | |
283 | ||
284 | @node Explicit Debug | |
285 | @subsection Explicit Entry to the Debugger | |
286 | ||
287 | You can cause the debugger to be called at a certain point in your | |
288 | program by writing the expression @code{(debug)} at that point. To do | |
289 | this, visit the source file, insert the text @samp{(debug)} at the | |
290 | proper place, and type @kbd{C-M-x} (@code{eval-defun}, a Lisp mode key | |
291 | binding). @strong{Warning:} if you do this for temporary debugging | |
292 | purposes, be sure to undo this insertion before you save the file! | |
293 | ||
294 | The place where you insert @samp{(debug)} must be a place where an | |
295 | additional form can be evaluated and its value ignored. (If the value | |
296 | of @code{(debug)} isn't ignored, it will alter the execution of the | |
297 | program!) The most common suitable places are inside a @code{progn} or | |
298 | an implicit @code{progn} (@pxref{Sequencing}). | |
299 | ||
300 | @node Using Debugger | |
301 | @subsection Using the Debugger | |
302 | ||
303 | When the debugger is entered, it displays the previously selected | |
304 | buffer in one window and a buffer named @samp{*Backtrace*} in another | |
305 | window. The backtrace buffer contains one line for each level of Lisp | |
306 | function execution currently going on. At the beginning of this buffer | |
307 | is a message describing the reason that the debugger was invoked (such | |
308 | as the error message and associated data, if it was invoked due to an | |
309 | error). | |
310 | ||
311 | The backtrace buffer is read-only and uses a special major mode, | |
312 | Debugger mode, in which letters are defined as debugger commands. The | |
313 | usual Emacs editing commands are available; thus, you can switch windows | |
314 | to examine the buffer that was being edited at the time of the error, | |
315 | switch buffers, visit files, or do any other sort of editing. However, | |
316 | the debugger is a recursive editing level (@pxref{Recursive Editing}) | |
317 | and it is wise to go back to the backtrace buffer and exit the debugger | |
318 | (with the @kbd{q} command) when you are finished with it. Exiting | |
319 | the debugger gets out of the recursive edit and kills the backtrace | |
320 | buffer. | |
321 | ||
322 | @cindex current stack frame | |
323 | The backtrace buffer shows you the functions that are executing and | |
324 | their argument values. It also allows you to specify a stack frame by | |
325 | moving point to the line describing that frame. (A stack frame is the | |
326 | place where the Lisp interpreter records information about a particular | |
327 | invocation of a function.) The frame whose line point is on is | |
328 | considered the @dfn{current frame}. Some of the debugger commands | |
329 | operate on the current frame. If a line starts with a star, that means | |
330 | that exiting that frame will call the debugger again. This is useful | |
331 | for examining the return value of a function. | |
332 | ||
333 | If a function name is underlined, that means the debugger knows | |
334 | where its source code is located. You can click @kbd{Mouse-2} on that | |
335 | name, or move to it and type @key{RET}, to visit the source code. | |
336 | ||
337 | The debugger itself must be run byte-compiled, since it makes | |
338 | assumptions about how many stack frames are used for the debugger | |
339 | itself. These assumptions are false if the debugger is running | |
340 | interpreted. | |
341 | ||
342 | @node Debugger Commands | |
343 | @subsection Debugger Commands | |
344 | @cindex debugger command list | |
345 | ||
346 | The debugger buffer (in Debugger mode) provides special commands in | |
347 | addition to the usual Emacs commands. The most important use of | |
348 | debugger commands is for stepping through code, so that you can see | |
349 | how control flows. The debugger can step through the control | |
350 | structures of an interpreted function, but cannot do so in a | |
351 | byte-compiled function. If you would like to step through a | |
352 | byte-compiled function, replace it with an interpreted definition of | |
353 | the same function. (To do this, visit the source for the function and | |
354 | type @kbd{C-M-x} on its definition.) You cannot use the Lisp debugger | |
355 | to step through a primitive function. | |
356 | ||
357 | Here is a list of Debugger mode commands: | |
358 | ||
359 | @table @kbd | |
360 | @item c | |
361 | Exit the debugger and continue execution. When continuing is possible, | |
362 | it resumes execution of the program as if the debugger had never been | |
363 | entered (aside from any side-effects that you caused by changing | |
364 | variable values or data structures while inside the debugger). | |
365 | ||
366 | Continuing is possible after entry to the debugger due to function entry | |
367 | or exit, explicit invocation, or quitting. You cannot continue if the | |
368 | debugger was entered because of an error. | |
369 | ||
370 | @item d | |
371 | Continue execution, but enter the debugger the next time any Lisp | |
372 | function is called. This allows you to step through the | |
373 | subexpressions of an expression, seeing what values the subexpressions | |
374 | compute, and what else they do. | |
375 | ||
376 | The stack frame made for the function call which enters the debugger in | |
377 | this way will be flagged automatically so that the debugger will be | |
378 | called again when the frame is exited. You can use the @kbd{u} command | |
379 | to cancel this flag. | |
380 | ||
381 | @item b | |
382 | Flag the current frame so that the debugger will be entered when the | |
383 | frame is exited. Frames flagged in this way are marked with stars | |
384 | in the backtrace buffer. | |
385 | ||
386 | @item u | |
387 | Don't enter the debugger when the current frame is exited. This | |
388 | cancels a @kbd{b} command on that frame. The visible effect is to | |
389 | remove the star from the line in the backtrace buffer. | |
390 | ||
391 | @item j | |
392 | Flag the current frame like @kbd{b}. Then continue execution like | |
393 | @kbd{c}, but temporarily disable break-on-entry for all functions that | |
394 | are set up to do so by @code{debug-on-entry}. | |
395 | ||
396 | @item e | |
397 | Read a Lisp expression in the minibuffer, evaluate it, and print the | |
398 | value in the echo area. The debugger alters certain important | |
399 | variables, and the current buffer, as part of its operation; @kbd{e} | |
400 | temporarily restores their values from outside the debugger, so you can | |
401 | examine and change them. This makes the debugger more transparent. By | |
402 | contrast, @kbd{M-:} does nothing special in the debugger; it shows you | |
403 | the variable values within the debugger. | |
404 | ||
405 | @item R | |
406 | Like @kbd{e}, but also save the result of evaluation in the | |
407 | buffer @samp{*Debugger-record*}. | |
408 | ||
409 | @item q | |
410 | Terminate the program being debugged; return to top-level Emacs | |
411 | command execution. | |
412 | ||
413 | If the debugger was entered due to a @kbd{C-g} but you really want | |
414 | to quit, and not debug, use the @kbd{q} command. | |
415 | ||
416 | @item r | |
417 | Return a value from the debugger. The value is computed by reading an | |
418 | expression with the minibuffer and evaluating it. | |
419 | ||
420 | The @kbd{r} command is useful when the debugger was invoked due to exit | |
421 | from a Lisp call frame (as requested with @kbd{b} or by entering the | |
422 | frame with @kbd{d}); then the value specified in the @kbd{r} command is | |
423 | used as the value of that frame. It is also useful if you call | |
424 | @code{debug} and use its return value. Otherwise, @kbd{r} has the same | |
425 | effect as @kbd{c}, and the specified return value does not matter. | |
426 | ||
427 | You can't use @kbd{r} when the debugger was entered due to an error. | |
428 | ||
429 | @item l | |
430 | Display a list of functions that will invoke the debugger when called. | |
431 | This is a list of functions that are set to break on entry by means of | |
432 | @code{debug-on-entry}. @strong{Warning:} if you redefine such a | |
433 | function and thus cancel the effect of @code{debug-on-entry}, it may | |
434 | erroneously show up in this list. | |
435 | @end table | |
436 | ||
437 | @node Invoking the Debugger | |
438 | @subsection Invoking the Debugger | |
439 | ||
440 | Here we describe in full detail the function @code{debug} that is used | |
441 | to invoke the debugger. | |
442 | ||
443 | @defun debug &rest debugger-args | |
444 | This function enters the debugger. It switches buffers to a buffer | |
445 | named @samp{*Backtrace*} (or @samp{*Backtrace*<2>} if it is the second | |
446 | recursive entry to the debugger, etc.), and fills it with information | |
447 | about the stack of Lisp function calls. It then enters a recursive | |
448 | edit, showing the backtrace buffer in Debugger mode. | |
449 | ||
450 | The Debugger mode @kbd{c}, @kbd{d}, @kbd{j}, and @kbd{r} commands exit | |
451 | the recursive edit; then @code{debug} switches back to the previous | |
452 | buffer and returns to whatever called @code{debug}. This is the only | |
453 | way the function @code{debug} can return to its caller. | |
454 | ||
455 | The use of the @var{debugger-args} is that @code{debug} displays the | |
456 | rest of its arguments at the top of the @samp{*Backtrace*} buffer, so | |
457 | that the user can see them. Except as described below, this is the | |
458 | @emph{only} way these arguments are used. | |
459 | ||
460 | However, certain values for first argument to @code{debug} have a | |
461 | special significance. (Normally, these values are used only by the | |
462 | internals of Emacs, and not by programmers calling @code{debug}.) Here | |
463 | is a table of these special values: | |
464 | ||
465 | @table @code | |
466 | @item lambda | |
467 | @cindex @code{lambda} in debug | |
468 | A first argument of @code{lambda} means @code{debug} was called | |
469 | because of entry to a function when @code{debug-on-next-call} was | |
470 | non-@code{nil}. The debugger displays @samp{Debugger | |
471 | entered--entering a function:} as a line of text at the top of the | |
472 | buffer. | |
473 | ||
474 | @item debug | |
475 | @code{debug} as first argument means @code{debug} was called because | |
476 | of entry to a function that was set to debug on entry. The debugger | |
477 | displays the string @samp{Debugger entered--entering a function:}, | |
478 | just as in the @code{lambda} case. It also marks the stack frame for | |
479 | that function so that it will invoke the debugger when exited. | |
480 | ||
481 | @item t | |
482 | When the first argument is @code{t}, this indicates a call to | |
483 | @code{debug} due to evaluation of a function call form when | |
484 | @code{debug-on-next-call} is non-@code{nil}. The debugger displays | |
485 | @samp{Debugger entered--beginning evaluation of function call form:} | |
486 | as the top line in the buffer. | |
487 | ||
488 | @item exit | |
489 | When the first argument is @code{exit}, it indicates the exit of a | |
490 | stack frame previously marked to invoke the debugger on exit. The | |
491 | second argument given to @code{debug} in this case is the value being | |
492 | returned from the frame. The debugger displays @samp{Debugger | |
493 | entered--returning value:} in the top line of the buffer, followed by | |
494 | the value being returned. | |
495 | ||
496 | @item error | |
497 | @cindex @code{error} in debug | |
498 | When the first argument is @code{error}, the debugger indicates that | |
499 | it is being entered because an error or @code{quit} was signaled and | |
500 | not handled, by displaying @samp{Debugger entered--Lisp error:} | |
501 | followed by the error signaled and any arguments to @code{signal}. | |
502 | For example, | |
503 | ||
504 | @example | |
505 | @group | |
506 | (let ((debug-on-error t)) | |
507 | (/ 1 0)) | |
508 | @end group | |
509 | ||
510 | @group | |
511 | ------ Buffer: *Backtrace* ------ | |
512 | Debugger entered--Lisp error: (arith-error) | |
513 | /(1 0) | |
514 | ... | |
515 | ------ Buffer: *Backtrace* ------ | |
516 | @end group | |
517 | @end example | |
518 | ||
519 | If an error was signaled, presumably the variable | |
520 | @code{debug-on-error} is non-@code{nil}. If @code{quit} was signaled, | |
521 | then presumably the variable @code{debug-on-quit} is non-@code{nil}. | |
522 | ||
523 | @item nil | |
524 | Use @code{nil} as the first of the @var{debugger-args} when you want | |
525 | to enter the debugger explicitly. The rest of the @var{debugger-args} | |
526 | are printed on the top line of the buffer. You can use this feature to | |
527 | display messages---for example, to remind yourself of the conditions | |
528 | under which @code{debug} is called. | |
529 | @end table | |
530 | @end defun | |
531 | ||
532 | @node Internals of Debugger | |
533 | @subsection Internals of the Debugger | |
534 | ||
535 | This section describes functions and variables used internally by the | |
536 | debugger. | |
537 | ||
538 | @defvar debugger | |
539 | The value of this variable is the function to call to invoke the | |
540 | debugger. Its value must be a function of any number of arguments, or, | |
541 | more typically, the name of a function. This function should invoke | |
542 | some kind of debugger. The default value of the variable is | |
543 | @code{debug}. | |
544 | ||
545 | The first argument that Lisp hands to the function indicates why it | |
546 | was called. The convention for arguments is detailed in the description | |
547 | of @code{debug} (@pxref{Invoking the Debugger}). | |
548 | @end defvar | |
549 | ||
550 | @deffn Command backtrace | |
551 | @cindex run time stack | |
552 | @cindex call stack | |
553 | This function prints a trace of Lisp function calls currently active. | |
554 | This is the function used by @code{debug} to fill up the | |
555 | @samp{*Backtrace*} buffer. It is written in C, since it must have access | |
556 | to the stack to determine which function calls are active. The return | |
557 | value is always @code{nil}. | |
558 | ||
559 | In the following example, a Lisp expression calls @code{backtrace} | |
560 | explicitly. This prints the backtrace to the stream | |
561 | @code{standard-output}, which, in this case, is the buffer | |
562 | @samp{backtrace-output}. | |
563 | ||
564 | Each line of the backtrace represents one function call. The line shows | |
565 | the values of the function's arguments if they are all known; if they | |
566 | are still being computed, the line says so. The arguments of special | |
567 | forms are elided. | |
568 | ||
569 | @smallexample | |
570 | @group | |
571 | (with-output-to-temp-buffer "backtrace-output" | |
572 | (let ((var 1)) | |
573 | (save-excursion | |
574 | (setq var (eval '(progn | |
575 | (1+ var) | |
576 | (list 'testing (backtrace)))))))) | |
577 | ||
578 | @result{} (testing nil) | |
579 | @end group | |
580 | ||
581 | @group | |
582 | ----------- Buffer: backtrace-output ------------ | |
583 | backtrace() | |
584 | (list ...computing arguments...) | |
585 | @end group | |
586 | (progn ...) | |
587 | eval((progn (1+ var) (list (quote testing) (backtrace)))) | |
588 | (setq ...) | |
589 | (save-excursion ...) | |
590 | (let ...) | |
591 | (with-output-to-temp-buffer ...) | |
592 | eval((with-output-to-temp-buffer ...)) | |
593 | eval-last-sexp-1(nil) | |
594 | @group | |
595 | eval-last-sexp(nil) | |
596 | call-interactively(eval-last-sexp) | |
597 | ----------- Buffer: backtrace-output ------------ | |
598 | @end group | |
599 | @end smallexample | |
600 | @end deffn | |
601 | ||
602 | @ignore @c Not worth mentioning | |
603 | @defopt stack-trace-on-error | |
604 | @cindex stack trace | |
605 | This variable controls whether Lisp automatically displays a | |
606 | backtrace buffer after every error that is not handled. A quit signal | |
607 | counts as an error for this variable. If it is non-@code{nil} then a | |
608 | backtrace is shown in a pop-up buffer named @samp{*Backtrace*} on every | |
609 | error. If it is @code{nil}, then a backtrace is not shown. | |
610 | ||
611 | When a backtrace is shown, that buffer is not selected. If either | |
612 | @code{debug-on-quit} or @code{debug-on-error} is also non-@code{nil}, then | |
613 | a backtrace is shown in one buffer, and the debugger is popped up in | |
614 | another buffer with its own backtrace. | |
615 | ||
616 | We consider this feature to be obsolete and superseded by the debugger | |
617 | itself. | |
618 | @end defopt | |
619 | @end ignore | |
620 | ||
621 | @defvar debug-on-next-call | |
622 | @cindex @code{eval}, and debugging | |
623 | @cindex @code{apply}, and debugging | |
624 | @cindex @code{funcall}, and debugging | |
625 | If this variable is non-@code{nil}, it says to call the debugger before | |
626 | the next @code{eval}, @code{apply} or @code{funcall}. Entering the | |
627 | debugger sets @code{debug-on-next-call} to @code{nil}. | |
628 | ||
629 | The @kbd{d} command in the debugger works by setting this variable. | |
630 | @end defvar | |
631 | ||
632 | @defun backtrace-debug level flag | |
633 | This function sets the debug-on-exit flag of the stack frame @var{level} | |
634 | levels down the stack, giving it the value @var{flag}. If @var{flag} is | |
635 | non-@code{nil}, this will cause the debugger to be entered when that | |
636 | frame later exits. Even a nonlocal exit through that frame will enter | |
637 | the debugger. | |
638 | ||
639 | This function is used only by the debugger. | |
640 | @end defun | |
641 | ||
642 | @defvar command-debug-status | |
643 | This variable records the debugging status of the current interactive | |
644 | command. Each time a command is called interactively, this variable is | |
645 | bound to @code{nil}. The debugger can set this variable to leave | |
646 | information for future debugger invocations during the same command | |
647 | invocation. | |
648 | ||
649 | The advantage of using this variable rather than an ordinary global | |
650 | variable is that the data will never carry over to a subsequent command | |
651 | invocation. | |
652 | @end defvar | |
653 | ||
654 | @defun backtrace-frame frame-number | |
655 | The function @code{backtrace-frame} is intended for use in Lisp | |
656 | debuggers. It returns information about what computation is happening | |
657 | in the stack frame @var{frame-number} levels down. | |
658 | ||
659 | If that frame has not evaluated the arguments yet, or is a special | |
660 | form, the value is @code{(nil @var{function} @var{arg-forms}@dots{})}. | |
661 | ||
662 | If that frame has evaluated its arguments and called its function | |
663 | already, the return value is @code{(t @var{function} | |
664 | @var{arg-values}@dots{})}. | |
665 | ||
666 | In the return value, @var{function} is whatever was supplied as the | |
667 | @sc{car} of the evaluated list, or a @code{lambda} expression in the | |
668 | case of a macro call. If the function has a @code{&rest} argument, that | |
669 | is represented as the tail of the list @var{arg-values}. | |
670 | ||
671 | If @var{frame-number} is out of range, @code{backtrace-frame} returns | |
672 | @code{nil}. | |
673 | @end defun | |
674 | ||
675 | @include edebug.texi | |
676 | ||
677 | @node Syntax Errors | |
678 | @section Debugging Invalid Lisp Syntax | |
679 | @cindex debugging invalid Lisp syntax | |
680 | ||
681 | The Lisp reader reports invalid syntax, but cannot say where the real | |
682 | problem is. For example, the error ``End of file during parsing'' in | |
683 | evaluating an expression indicates an excess of open parentheses (or | |
684 | square brackets). The reader detects this imbalance at the end of the | |
685 | file, but it cannot figure out where the close parenthesis should have | |
686 | been. Likewise, ``Invalid read syntax: ")"'' indicates an excess close | |
687 | parenthesis or missing open parenthesis, but does not say where the | |
688 | missing parenthesis belongs. How, then, to find what to change? | |
689 | ||
690 | If the problem is not simply an imbalance of parentheses, a useful | |
691 | technique is to try @kbd{C-M-e} at the beginning of each defun, and see | |
692 | if it goes to the place where that defun appears to end. If it does | |
693 | not, there is a problem in that defun. | |
694 | ||
695 | @cindex unbalanced parentheses | |
696 | @cindex parenthesis mismatch, debugging | |
697 | However, unmatched parentheses are the most common syntax errors in | |
698 | Lisp, and we can give further advice for those cases. (In addition, | |
699 | just moving point through the code with Show Paren mode enabled might | |
700 | find the mismatch.) | |
701 | ||
702 | @menu | |
703 | * Excess Open:: How to find a spurious open paren or missing close. | |
704 | * Excess Close:: How to find a spurious close paren or missing open. | |
705 | @end menu | |
706 | ||
707 | @node Excess Open | |
708 | @subsection Excess Open Parentheses | |
709 | ||
710 | The first step is to find the defun that is unbalanced. If there is | |
711 | an excess open parenthesis, the way to do this is to go to the end of | |
712 | the file and type @kbd{C-u C-M-u}. This will move you to the | |
713 | beginning of the first defun that is unbalanced. | |
714 | ||
715 | The next step is to determine precisely what is wrong. There is no | |
716 | way to be sure of this except by studying the program, but often the | |
717 | existing indentation is a clue to where the parentheses should have | |
718 | been. The easiest way to use this clue is to reindent with @kbd{C-M-q} | |
719 | and see what moves. @strong{But don't do this yet!} Keep reading, | |
720 | first. | |
721 | ||
722 | Before you do this, make sure the defun has enough close parentheses. | |
723 | Otherwise, @kbd{C-M-q} will get an error, or will reindent all the rest | |
724 | of the file until the end. So move to the end of the defun and insert a | |
725 | close parenthesis there. Don't use @kbd{C-M-e} to move there, since | |
726 | that too will fail to work until the defun is balanced. | |
727 | ||
728 | Now you can go to the beginning of the defun and type @kbd{C-M-q}. | |
729 | Usually all the lines from a certain point to the end of the function | |
730 | will shift to the right. There is probably a missing close parenthesis, | |
731 | or a superfluous open parenthesis, near that point. (However, don't | |
732 | assume this is true; study the code to make sure.) Once you have found | |
733 | the discrepancy, undo the @kbd{C-M-q} with @kbd{C-_}, since the old | |
734 | indentation is probably appropriate to the intended parentheses. | |
735 | ||
736 | After you think you have fixed the problem, use @kbd{C-M-q} again. If | |
737 | the old indentation actually fit the intended nesting of parentheses, | |
738 | and you have put back those parentheses, @kbd{C-M-q} should not change | |
739 | anything. | |
740 | ||
741 | @node Excess Close | |
742 | @subsection Excess Close Parentheses | |
743 | ||
744 | To deal with an excess close parenthesis, first go to the beginning | |
745 | of the file, then type @kbd{C-u -1 C-M-u} to find the end of the first | |
746 | unbalanced defun. | |
747 | ||
748 | Then find the actual matching close parenthesis by typing @kbd{C-M-f} | |
749 | at the beginning of that defun. This will leave you somewhere short of | |
750 | the place where the defun ought to end. It is possible that you will | |
751 | find a spurious close parenthesis in that vicinity. | |
752 | ||
753 | If you don't see a problem at that point, the next thing to do is to | |
754 | type @kbd{C-M-q} at the beginning of the defun. A range of lines will | |
755 | probably shift left; if so, the missing open parenthesis or spurious | |
756 | close parenthesis is probably near the first of those lines. (However, | |
757 | don't assume this is true; study the code to make sure.) Once you have | |
758 | found the discrepancy, undo the @kbd{C-M-q} with @kbd{C-_}, since the | |
759 | old indentation is probably appropriate to the intended parentheses. | |
760 | ||
761 | After you think you have fixed the problem, use @kbd{C-M-q} again. If | |
762 | the old indentation actually fits the intended nesting of parentheses, | |
763 | and you have put back those parentheses, @kbd{C-M-q} should not change | |
764 | anything. | |
765 | ||
766 | @node Test Coverage | |
767 | @section Test Coverage | |
768 | @cindex coverage testing | |
769 | ||
770 | @findex testcover-start | |
771 | @findex testcover-mark-all | |
772 | @findex testcover-next-mark | |
773 | You can do coverage testing for a file of Lisp code by loading the | |
774 | @code{testcover} library and using the command @kbd{M-x | |
775 | testcover-start @key{RET} @var{file} @key{RET}} to instrument the | |
776 | code. Then test your code by calling it one or more times. Then use | |
777 | the command @kbd{M-x testcover-mark-all} to display colored highlights | |
778 | on the code to show where coverage is insufficient. The command | |
779 | @kbd{M-x testcover-next-mark} will move point forward to the next | |
780 | highlighted spot. | |
781 | ||
782 | Normally, a red highlight indicates the form was never completely | |
783 | evaluated; a brown highlight means it always evaluated to the same | |
784 | value (meaning there has been little testing of what is done with the | |
785 | result). However, the red highlight is skipped for forms that can't | |
786 | possibly complete their evaluation, such as @code{error}. The brown | |
787 | highlight is skipped for forms that are expected to always evaluate to | |
788 | the same value, such as @code{(setq x 14)}. | |
789 | ||
790 | For difficult cases, you can add do-nothing macros to your code to | |
791 | give advice to the test coverage tool. | |
792 | ||
793 | @defmac 1value form | |
794 | Evaluate @var{form} and return its value, but inform coverage testing | |
795 | that @var{form}'s value should always be the same. | |
796 | @end defmac | |
797 | ||
798 | @defmac noreturn form | |
799 | Evaluate @var{form}, informing coverage testing that @var{form} should | |
800 | never return. If it ever does return, you get a run-time error. | |
801 | @end defmac | |
802 | ||
803 | Edebug also has a coverage testing feature (@pxref{Coverage | |
804 | Testing}). These features partly duplicate each other, and it would | |
805 | be cleaner to combine them. | |
806 | ||
807 | @node Compilation Errors | |
808 | @section Debugging Problems in Compilation | |
809 | @cindex debugging byte compilation problems | |
810 | ||
811 | When an error happens during byte compilation, it is normally due to | |
812 | invalid syntax in the program you are compiling. The compiler prints a | |
813 | suitable error message in the @samp{*Compile-Log*} buffer, and then | |
814 | stops. The message may state a function name in which the error was | |
815 | found, or it may not. Either way, here is how to find out where in the | |
816 | file the error occurred. | |
817 | ||
818 | What you should do is switch to the buffer @w{@samp{ *Compiler Input*}}. | |
819 | (Note that the buffer name starts with a space, so it does not show | |
820 | up in @kbd{M-x list-buffers}.) This buffer contains the program being | |
821 | compiled, and point shows how far the byte compiler was able to read. | |
822 | ||
823 | If the error was due to invalid Lisp syntax, point shows exactly where | |
824 | the invalid syntax was @emph{detected}. The cause of the error is not | |
825 | necessarily near by! Use the techniques in the previous section to find | |
826 | the error. | |
827 | ||
828 | If the error was detected while compiling a form that had been read | |
829 | successfully, then point is located at the end of the form. In this | |
830 | case, this technique can't localize the error precisely, but can still | |
831 | show you which function to check. | |
832 | ||
833 | @ignore | |
834 | arch-tag: ddc57378-b0e6-4195-b7b6-43f8777395a7 | |
835 | @end ignore |