lread.c 136 KB
Newer Older
Jim Blandy's avatar
Jim Blandy committed
1
/* Lisp parsing and input streams.
2

Paul Eggert's avatar
Paul Eggert committed
3
Copyright (C) 1985-1989, 1993-1995, 1997-2015 Free Software Foundation,
4
Inc.
Jim Blandy's avatar
Jim Blandy committed
5 6 7

This file is part of GNU Emacs.

8
GNU Emacs is free software: you can redistribute it and/or modify
Jim Blandy's avatar
Jim Blandy committed
9
it under the terms of the GNU General Public License as published by
10 11
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Jim Blandy's avatar
Jim Blandy committed
12 13 14 15 16 17 18

GNU Emacs is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
19
along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
Jim Blandy's avatar
Jim Blandy committed
20

21 22
/* Tell globals.h to define tables needed by init_obarray.  */
#define DEFINE_SYMBOLS
Jim Blandy's avatar
Jim Blandy committed
23

Richard M. Stallman's avatar
Richard M. Stallman committed
24
#include <config.h>
25
#include "sysstdio.h"
Jim Blandy's avatar
Jim Blandy committed
26 27 28
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/file.h>
29
#include <errno.h>
30
#include <limits.h>	/* For CHAR_BIT.  */
31
#include <math.h>
32
#include <stat-time.h>
Jim Blandy's avatar
Jim Blandy committed
33
#include "lisp.h"
34
#include "dispextern.h"
35
#include "intervals.h"
Kenichi Handa's avatar
Kenichi Handa committed
36
#include "character.h"
37
#include "buffer.h"
Karl Heuer's avatar
Karl Heuer committed
38
#include "charset.h"
Kenichi Handa's avatar
Kenichi Handa committed
39
#include "coding.h"
40
#include <epaths.h>
Jim Blandy's avatar
Jim Blandy committed
41
#include "commands.h"
Jim Blandy's avatar
Jim Blandy committed
42
#include "keyboard.h"
43
#include "systime.h"
44
#include "termhooks.h"
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
45
#include "blockinput.h"
Jim Blandy's avatar
Jim Blandy committed
46

47 48 49 50
#ifdef MSDOS
#include "msdos.h"
#endif

51 52 53 54
#ifdef HAVE_NS
#include "nsterm.h"
#endif

Andreas Schwab's avatar
Andreas Schwab committed
55
#include <unistd.h>
Jim Blandy's avatar
Jim Blandy committed
56

Richard M. Stallman's avatar
Richard M. Stallman committed
57 58 59 60
#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

Dave Love's avatar
Dave Love committed
61
#include <fcntl.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
62

63
#ifdef HAVE_FSEEKO
64 65 66 67 68 69 70
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

71 72 73 74
/* The association list of objects read with the #n=object form.
   Each member of the list has the form (n . object), and is used to
   look up the object for the corresponding #n# construct.
   It must be set to nil before all top-level calls to read0.  */
75
static Lisp_Object read_objects;
76

77
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
78 79
static FILE *instream;

80
/* For use within read-from-string (this reader is non-reentrant!!)  */
81 82 83
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
84

85
/* Number of characters read in the current call to Fread or
86
   Fread_from_string.  */
87
static EMACS_INT readchar_count;
88

89
/* This contains the last string skipped with #@.  */
90 91
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
92
static ptrdiff_t saved_doc_string_size;
93
/* Length of actual data in saved_doc_string.  */
94
static ptrdiff_t saved_doc_string_length;
95
/* This is the file position that string came from.  */
96
static file_offset saved_doc_string_position;
97

98 99 100 101 102
/* This contains the previous string skipped with #@.
   We copy it from saved_doc_string when a new string
   is put in saved_doc_string.  */
static char *prev_saved_doc_string;
/* Length of buffer allocated in prev_saved_doc_string.  */
103
static ptrdiff_t prev_saved_doc_string_size;
104
/* Length of actual data in prev_saved_doc_string.  */
105
static ptrdiff_t prev_saved_doc_string_length;
106
/* This is the file position that string came from.  */
107
static file_offset prev_saved_doc_string_position;
108

109
/* True means inside a new-style backquote
110
   with no surrounding parentheses.
111
   Fread initializes this to false, so we need not specbind it
112
   or worry about what happens to it when there is an error.  */
113
static bool new_backquote_flag;
114 115 116 117

/* A list of file names for files being loaded in Fload.  Used to
   check for recursive loads.  */

Lute Kamstra's avatar
Lute Kamstra committed
118
static Lisp_Object Vloads_in_progress;
119

120 121
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
122

123
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
124 125
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
126

Kenichi Handa's avatar
Kenichi Handa committed
127 128 129 130 131 132
/* Functions that read one byte from the current source READCHARFUN
   or unreads one byte.  If the integer argument C is -1, it returns
   one read byte, or -1 when there's no more byte in the source.  If C
   is 0 or positive, it unreads C, and the return value is not
   interesting.  */

133 134 135
static int readbyte_for_lambda (int, Lisp_Object);
static int readbyte_from_file (int, Lisp_Object);
static int readbyte_from_string (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
136

Jim Blandy's avatar
Jim Blandy committed
137 138
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
139 140
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
141
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
142

143
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
144 145
#define UNREAD(c) unreadchar (readcharfun, c)

146 147 148
/* Same as READCHAR but set *MULTIBYTE to the multibyteness of the source.  */
#define READCHAR_REPORT_MULTIBYTE(multibyte) readchar (readcharfun, multibyte)

Kenichi Handa's avatar
Kenichi Handa committed
149
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
150 151
   Qlambda, or a cons, we use this to keep an unread character because
   a file stream can't handle multibyte-char unreading.  The value -1
152
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
153 154
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
155
static int
156
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
157 158
{
  Lisp_Object tem;
159
  register int c;
160
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
161 162
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
163
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
164

165 166 167
  if (multibyte)
    *multibyte = 0;

168
  readchar_count++;
169

170
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
171
    {
172
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
173

174
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
175

176 177 178
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

179 180
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
181

Tom Tromey's avatar
Tom Tromey committed
182
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
183
	{
184
	  /* Fetch the character code from the buffer.  */
185 186
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
187
	  c = STRING_CHAR (p);
188 189
	  if (multibyte)
	    *multibyte = 1;
190 191 192 193
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
194
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
195
	    c = BYTE8_TO_CHAR (c);
196
	  pt_byte++;
197
	}
198
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
199

200
      return c;
Jim Blandy's avatar
Jim Blandy committed
201
    }
202
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
203
    {
204
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
205

206
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
207

208 209
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
210

Tom Tromey's avatar
Tom Tromey committed
211
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
212
	{
213
	  /* Fetch the character code from the buffer.  */
214 215
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
216
	  c = STRING_CHAR (p);
217 218
	  if (multibyte)
	    *multibyte = 1;
219 220 221 222
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
223
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
224
	    c = BYTE8_TO_CHAR (c);
225
	  bytepos++;
226 227
	}

228 229 230 231
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
232
    }
233 234

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
235 236 237 238
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
239

Jim Blandy's avatar
Jim Blandy committed
240
  if (EQ (readcharfun, Qget_file_char))
241
    {
Kenichi Handa's avatar
Kenichi Handa committed
242 243
      readbyte = readbyte_from_file;
      goto read_multibyte;
244
    }
Jim Blandy's avatar
Jim Blandy committed
245

246
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
247
    {
248
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
249
	c = -1;
250 251 252 253 254 255 256 257
      else if (STRING_MULTIBYTE (readcharfun))
	{
	  if (multibyte)
	    *multibyte = 1;
	  FETCH_STRING_CHAR_ADVANCE_NO_CHECK (c, readcharfun,
					      read_from_string_index,
					      read_from_string_index_byte);
	}
258
      else
259 260 261 262 263
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
264 265 266
      return c;
    }

Kenichi Handa's avatar
Kenichi Handa committed
267 268 269 270 271 272 273 274 275 276 277 278
  if (CONSP (readcharfun))
    {
      /* This is the case that read_vector is reading from a unibyte
	 string that contains a byte sequence previously skipped
	 because of #@NUMBER.  The car part of readcharfun is that
	 string, and the cdr part is a value of readcharfun given to
	 read_vector.  */
      readbyte = readbyte_from_string;
      if (EQ (XCDR (readcharfun), Qget_emacs_mule_file_char))
	emacs_mule_encoding = 1;
      goto read_multibyte;
    }
Kenichi Handa's avatar
Kenichi Handa committed
279

Kenichi Handa's avatar
Kenichi Handa committed
280 281 282 283 284 285
  if (EQ (readcharfun, Qget_emacs_mule_file_char))
    {
      readbyte = readbyte_from_file;
      emacs_mule_encoding = 1;
      goto read_multibyte;
    }
Jim Blandy's avatar
Jim Blandy committed
286 287 288

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
289
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
290 291
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
292 293 294 295 296 297 298 299 300

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
301
  if (c < 0)
302 303 304
    return c;
  if (multibyte)
    *multibyte = 1;
305
  if (ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322
    return c;
  if (emacs_mule_encoding)
    return read_emacs_mule_char (c, readbyte, readcharfun);
  i = 0;
  buf[i++] = c;
  len = BYTES_BY_CHAR_HEAD (c);
  while (i < len)
    {
      c = (*readbyte) (-1, readcharfun);
      if (c < 0 || ! TRAILING_CODE_P (c))
	{
	  while (--i > 1)
	    (*readbyte) (buf[i], readcharfun);
	  return BYTE8_TO_CHAR (buf[0]);
	}
      buf[i++] = c;
    }
323
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
324 325
}

Stefan Monnier's avatar
Stefan Monnier committed
326 327 328 329
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

330 331 332
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
333
  if (FROM_FILE_P (readcharfun))
334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
    {
      block_input ();		/* FIXME: Not sure if it's needed.  */
      fseek (instream, n, SEEK_CUR);
      unblock_input ();
    }
  else
    { /* We're not reading directly from a file.  In that case, it's difficult
	 to reliably count bytes, since these are usually meant for the file's
	 encoding, whereas we're now typically in the internal encoding.
	 But luckily, skip_dyn_bytes is used to skip over a single
	 dynamic-docstring (or dynamic byte-code) which is always quoted such
	 that \037 is the final char.  */
      int c;
      do {
	c = READCHAR;
      } while (c >= 0 && c != '\037');
    }
}

353 354 355 356 357 358 359 360 361 362 363 364 365
static void
skip_dyn_eof (Lisp_Object readcharfun)
{
  if (FROM_FILE_P (readcharfun))
    {
      block_input ();		/* FIXME: Not sure if it's needed.  */
      fseek (instream, 0, SEEK_END);
      unblock_input ();
    }
  else
    while (READCHAR >= 0);
}

Jim Blandy's avatar
Jim Blandy committed
366 367 368 369
/* Unread the character C in the way appropriate for the stream READCHARFUN.
   If the stream is a user function, call it with the char as argument.  */

static void
370
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
371
{
372
  readchar_count--;
373 374 375 376
  if (c == -1)
    /* Don't back up the pointer if we're unreading the end-of-input mark,
       since readchar didn't advance it when we read it.  */
    ;
377
  else if (BUFFERP (readcharfun))
378
    {
379
      struct buffer *b = XBUFFER (readcharfun);
380 381
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
382

Tom Tromey's avatar
Tom Tromey committed
383
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
384
	BUF_DEC_POS (b, bytepos);
385
      else
Kenichi Handa's avatar
Kenichi Handa committed
386
	bytepos--;
387

388
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
389
    }
390
  else if (MARKERP (readcharfun))
391
    {
392
      struct buffer *b = XMARKER (readcharfun)->buffer;
393
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
394

Kenichi Handa's avatar
Kenichi Handa committed
395
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
396
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
397
	BUF_DEC_POS (b, bytepos);
398
      else
Kenichi Handa's avatar
Kenichi Handa committed
399
	bytepos--;
400

Kenichi Handa's avatar
Kenichi Handa committed
401
      XMARKER (readcharfun)->bytepos = bytepos;
402
    }
403
  else if (STRINGP (readcharfun))
404 405 406 407 408
    {
      read_from_string_index--;
      read_from_string_index_byte
	= string_char_to_byte (readcharfun, read_from_string_index);
    }
Kenichi Handa's avatar
Kenichi Handa committed
409 410 411 412
  else if (CONSP (readcharfun))
    {
      unread_char = c;
    }
413
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
414 415 416
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
417
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
418
    {
419
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
420
    }
Jim Blandy's avatar
Jim Blandy committed
421 422 423 424
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
425
static int
426
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
427 428 429 430 431 432
{
  return read_bytecode_char (c >= 0);
}


static int
433
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
434 435 436
{
  if (c >= 0)
    {
437
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
438
      ungetc (c, instream);
439
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
440 441 442
      return 0;
    }

443
  block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
444
  c = getc (instream);
Miles Bader's avatar
Miles Bader committed
445

446
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
447 448
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
449
      unblock_input ();
Miles Bader's avatar
Miles Bader committed
450
      QUIT;
451
      block_input ();
Miles Bader's avatar
Miles Bader committed
452
      clearerr (instream);
Miles Bader's avatar
Miles Bader committed
453
      c = getc (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
454
    }
Miles Bader's avatar
Miles Bader committed
455

456
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
457

Kenichi Handa's avatar
Kenichi Handa committed
458 459 460 461
  return (c == EOF ? -1 : c);
}

static int
462
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
463 464 465 466 467 468 469 470 471
{
  Lisp_Object string = XCAR (readcharfun);

  if (c >= 0)
    {
      read_from_string_index--;
      read_from_string_index_byte
	= string_char_to_byte (string, read_from_string_index);
    }
Kenichi Handa's avatar
Kenichi Handa committed
472

Kenichi Handa's avatar
Kenichi Handa committed
473 474 475 476 477 478 479 480 481 482 483 484 485 486 487
  if (read_from_string_index >= read_from_string_limit)
    c = -1;
  else
    FETCH_STRING_CHAR_ADVANCE (c, string,
			       read_from_string_index,
			       read_from_string_index_byte);
  return c;
}


/* Read one non-ASCII character from INSTREAM.  The character is
   encoded in `emacs-mule' and the first byte is already read in
   C.  */

static int
488
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516
{
  /* Emacs-mule coding uses at most 4-byte for one character.  */
  unsigned char buf[4];
  int len = emacs_mule_bytes[c];
  struct charset *charset;
  int i;
  unsigned code;

  if (len == 1)
    /* C is not a valid leading-code of `emacs-mule'.  */
    return BYTE8_TO_CHAR (c);

  i = 0;
  buf[i++] = c;
  while (i < len)
    {
      c = (*readbyte) (-1, readcharfun);
      if (c < 0xA0)
	{
	  while (--i > 1)
	    (*readbyte) (buf[i], readcharfun);
	  return BYTE8_TO_CHAR (buf[0]);
	}
      buf[i++] = c;
    }

  if (len == 2)
    {
517
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
518 519 520 521 522 523 524
      code = buf[1] & 0x7F;
    }
  else if (len == 3)
    {
      if (buf[0] == EMACS_MULE_LEADING_CODE_PRIVATE_11
	  || buf[0] == EMACS_MULE_LEADING_CODE_PRIVATE_12)
	{
525
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
526 527 528 529
	  code = buf[2] & 0x7F;
	}
      else
	{
530
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
531 532 533 534 535
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
536
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
537
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
538 539 540 541
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
542
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
543 544 545 546
  return c;
}


547 548 549
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
550
static Lisp_Object read1 (Lisp_Object, int *, bool);
551

552 553
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
554

555 556 557 558 559
static Lisp_Object substitute_object_recurse (Lisp_Object, Lisp_Object,
                                              Lisp_Object);
static void substitute_object_in_subtree (Lisp_Object,
                                          Lisp_Object);
static void substitute_in_interval (INTERVAL, Lisp_Object);
560

Jim Blandy's avatar
Jim Blandy committed
561

562
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
563

564 565
/* Read input events until we get one that's acceptable for our purposes.

566
   If NO_SWITCH_FRAME, switch-frame events are stashed
567 568 569
   until we get a character we like, and then stuffed into
   unread_switch_frame.

570
   If ASCII_REQUIRED, check function key events to see
571 572 573
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

574 575 576
   If ERROR_NONASCII, signal an error if the input we
   get isn't an ASCII character with modifiers.  If it's false but
   ASCII_REQUIRED is true, just re-read until we get an ASCII
577 578
   character.

579
   If INPUT_METHOD, invoke the current input method
580 581
   if the character warrants that.

582
   If SECONDS is a number, wait that many seconds for input, and
583
   return Qnil if no input arrives within that time.  */
584

Andreas Schwab's avatar
Andreas Schwab committed
585
static Lisp_Object
586 587
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
588
{
589
  Lisp_Object val, delayed_switch_frame;
590
  struct timespec end_time;
591

592
#ifdef HAVE_WINDOW_SYSTEM
593 594
  if (display_hourglass_p)
    cancel_hourglass ();
595
#endif
596

597
  delayed_switch_frame = Qnil;
598

599 600 601
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
602
      double duration = extract_float (seconds);
603 604
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
605 606
    }

607
  /* Read until we get an acceptable event.  */
608
 retry:
609
  do
610
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
611
		     NUMBERP (seconds) ? &end_time : NULL);
612
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
613

614
  if (BUFFERP (val))
615 616
    goto retry;

617
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
618
     character.  This is better than signaling an error just because
619 620 621 622 623
     the last characters were typed to a separate minibuffer frame,
     for example.  Eventually, some code which can deal with
     switch-frame events will read it and process it.  */
  if (no_switch_frame
      && EVENT_HAS_PARAMETERS (val)
624
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
625 626 627 628 629
    {
      delayed_switch_frame = val;
      goto retry;
    }

630
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
631 632
    {
      /* Convert certain symbols to their ASCII equivalents.  */
633
      if (SYMBOLP (val))
634
	{
635
	  Lisp_Object tem, tem1;
636 637 638 639 640 641 642
	  tem = Fget (val, Qevent_symbol_element_mask);
	  if (!NILP (tem))
	    {
	      tem1 = Fget (Fcar (tem), Qascii_character);
	      /* Merge this symbol's modifier bits
		 with the ASCII equivalent of its basic code.  */
	      if (!NILP (tem1))
643
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
644 645
	    }
	}
646

647
      /* If we don't have a character now, deal with it appropriately.  */
648
      if (!INTEGERP (val))
649 650 651
	{
	  if (error_nonascii)
	    {
652
	      Vunread_command_events = list1 (val);
653 654 655 656 657 658 659 660 661 662
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

  if (! NILP (delayed_switch_frame))
    unread_switch_frame = delayed_switch_frame;

663 664
#if 0

665
#ifdef HAVE_WINDOW_SYSTEM
666 667
  if (display_hourglass_p)
    start_hourglass ();
668
#endif
669 670 671

#endif

672 673 674
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
675
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
676 677
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
678 679 680
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

681 682
If the user generates an event which is not a character (i.e. a mouse
click or function key event), `read-char' signals an error.  As an
683 684
exception, switch-frame events are put off until non-character events
can be read.
685 686 687 688 689 690
If you want to read non-character events, or ignore them, call
`read-event' or `read-char-exclusive' instead.

If the optional argument PROMPT is non-nil, display that as a prompt.
If the optional argument INHERIT-INPUT-METHOD is non-nil and some
input method is turned on in the current buffer, that input method
691 692 693 694 695
is used for reading a character.
If the optional argument SECONDS is non-nil, it should be a number
specifying the maximum number of seconds to wait for input.  If no
input arrives in that time, return nil.  SECONDS may be a
floating-point value.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
696
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
697
{
698 699
  Lisp_Object val;

700 701
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
702
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
703 704 705

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
706 707
}

Paul Eggert's avatar
Paul Eggert committed
708
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
709 710 711 712
       doc: /* Read an event object from the input stream.
If the optional argument PROMPT is non-nil, display that as a prompt.
If the optional argument INHERIT-INPUT-METHOD is non-nil and some
input method is turned on in the current buffer, that input method
713 714 715 716 717
is used for reading a character.
If the optional argument SECONDS is non-nil, it should be a number
specifying the maximum number of seconds to wait for input.  If no
input arrives in that time, return nil.  SECONDS may be a
floating-point value.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
718
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
719
{
720 721
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
722
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
723 724
}

725
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
726 727
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
728 729
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
730 731 732 733

If the optional argument PROMPT is non-nil, display that as a prompt.
If the optional argument INHERIT-INPUT-METHOD is non-nil and some
input method is turned on in the current buffer, that input method
734 735 736 737 738
is used for reading a character.
If the optional argument SECONDS is non-nil, it should be a number
specifying the maximum number of seconds to wait for input.  If no
input arrives in that time, return nil.  SECONDS may be a
floating-point value.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
739
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
740
{
741 742
  Lisp_Object val;

743 744
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
745

746
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
747 748 749

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
750 751 752
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
753
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
754
  (void)
Jim Blandy's avatar
Jim Blandy committed
755 756
{
  register Lisp_Object val;
757
  block_input ();
758
  XSETINT (val, getc (instream));
759
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
760 761
  return val;
}
762 763


764 765 766 767


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
768 769
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
770

771
static bool
Stefan Monnier's avatar
Stefan Monnier committed
772
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
773 774
{
  int ch = READCHAR;
775 776 777 778 779 780 781 782 783 784 785 786 787

  if (ch == '#')
    {
      ch = READCHAR;
      if (ch != '!')
        {
          UNREAD (ch);
          UNREAD ('#');
          return 0;
        }
      while (ch != '\n' && ch != EOF)
        ch = READCHAR;
      if (ch == '\n') ch = READCHAR;
Glenn Morris's avatar
Comment  
Glenn Morris committed
788 789
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
790 791
    }

792 793 794 795 796 797 798 799 800
  if (ch != ';')
    /* The first line isn't a comment, just give up.  */
    {
      UNREAD (ch);
      return 0;
    }
  else
    /* Look for an appropriate file-variable in the first line.  */
    {
801
      bool rv = 0;
802
      enum {
Paul Eggert's avatar
Paul Eggert committed
803
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
804
      } beg_end_state = NOMINAL;
805
      bool in_file_vars = 0;
806

807 808 809 810 811 812 813 814 815 816
#define UPDATE_BEG_END_STATE(ch)				\
  if (beg_end_state == NOMINAL)					\
    beg_end_state = (ch == '-' ? AFTER_FIRST_DASH : NOMINAL);	\
  else if (beg_end_state == AFTER_FIRST_DASH)			\
    beg_end_state = (ch == '*' ? AFTER_ASTERIX : NOMINAL);	\
  else if (beg_end_state == AFTER_ASTERIX)			\
    {								\
      if (ch == '-')						\
	in_file_vars = !in_file_vars;				\
      beg_end_state = NOMINAL;					\
817 818 819 820 821 822 823 824 825 826 827 828
    }

      /* Skip until we get to the file vars, if any.  */
      do
	{
	  ch = READCHAR;
	  UPDATE_BEG_END_STATE (ch);
	}
      while (!in_file_vars && ch != '\n' && ch != EOF);

      while (in_file_vars)
	{
829
	  char var[100], val[100];
830
	  unsigned i;
831 832 833 834 835 836 837

	  ch = READCHAR;

	  /* Read a variable name.  */
	  while (ch == ' ' || ch == '\t')
	    ch = READCHAR;

838
	  i = 0;
839
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
840
	    {
841 842
	      if (i < sizeof var - 1)
		var[i++] = ch;
843 844 845
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
846

847
	  /* Stop scanning if no colon was found before end marker.  */
848
	  if (!in_file_vars || ch == '\n' || ch == EOF)
849 850
	    break;

851 852 853
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
854 855 856 857 858 859 860 861 862

	  if (ch == ':')
	    {
	      /* Read a variable value.  */
	      ch = READCHAR;

	      while (ch == ' ' || ch == '\t')
		ch = READCHAR;