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 "intervals.h"
Kenichi Handa's avatar
Kenichi Handa committed
35
#include "character.h"
36
#include "buffer.h"
Karl Heuer's avatar
Karl Heuer committed
37
#include "charset.h"
Kenichi Handa's avatar
Kenichi Handa committed
38
#include "coding.h"
39
#include <epaths.h>
Jim Blandy's avatar
Jim Blandy committed
40
#include "commands.h"
Jim Blandy's avatar
Jim Blandy committed
41
#include "keyboard.h"
42
#include "frame.h"
43
#include "termhooks.h"
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
44
#include "blockinput.h"
Jim Blandy's avatar
Jim Blandy committed
45

46 47 48 49
#ifdef MSDOS
#include "msdos.h"
#endif

50 51 52 53
#ifdef HAVE_NS
#include "nsterm.h"
#endif

Andreas Schwab's avatar
Andreas Schwab committed
54
#include <unistd.h>
Jim Blandy's avatar
Jim Blandy committed
55

Richard M. Stallman's avatar
Richard M. Stallman committed
56 57 58 59
#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

Dave Love's avatar
Dave Love committed
60
#include <fcntl.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
61

62
#ifdef HAVE_FSEEKO
63 64 65 66 67 68 69
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

70 71 72 73
/* 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.  */
74
static Lisp_Object read_objects;
75

76
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
77 78
static FILE *instream;

79
/* For use within read-from-string (this reader is non-reentrant!!)  */
80 81 82
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
83

84
/* Number of characters read in the current call to Fread or
85
   Fread_from_string.  */
86
static EMACS_INT readchar_count;
87

88
/* This contains the last string skipped with #@.  */
89 90
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
91
static ptrdiff_t saved_doc_string_size;
92
/* Length of actual data in saved_doc_string.  */
93
static ptrdiff_t saved_doc_string_length;
94
/* This is the file position that string came from.  */
95
static file_offset saved_doc_string_position;
96

97 98 99 100 101
/* 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.  */
102
static ptrdiff_t prev_saved_doc_string_size;
103
/* Length of actual data in prev_saved_doc_string.  */
104
static ptrdiff_t prev_saved_doc_string_length;
105
/* This is the file position that string came from.  */
106
static file_offset prev_saved_doc_string_position;
107

108
/* True means inside a new-style backquote
109
   with no surrounding parentheses.
110
   Fread initializes this to false, so we need not specbind it
111
   or worry about what happens to it when there is an error.  */
112
static bool new_backquote_flag;
113 114 115 116

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

Lute Kamstra's avatar
Lute Kamstra committed
117
static Lisp_Object Vloads_in_progress;
118

119 120
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
121

122
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
123 124
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
125

Kenichi Handa's avatar
Kenichi Handa committed
126 127 128 129 130 131
/* 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.  */

132 133 134
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
135

Jim Blandy's avatar
Jim Blandy committed
136 137
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
138 139
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
140
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
141

142
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
143 144
#define UNREAD(c) unreadchar (readcharfun, c)

145 146 147
/* 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
148
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
149 150
   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
151
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
152 153
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
154
static int
155
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
156 157
{
  Lisp_Object tem;
158
  register int c;
159
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
160 161
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
162
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
163

164 165 166
  if (multibyte)
    *multibyte = 0;

167
  readchar_count++;
168

169
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
170
    {
171
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
172

173
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
174

175 176 177
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

178 179
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
180

Tom Tromey's avatar
Tom Tromey committed
181
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
182
	{
183
	  /* Fetch the character code from the buffer.  */
184 185
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
186
	  c = STRING_CHAR (p);
187 188
	  if (multibyte)
	    *multibyte = 1;
189 190 191 192
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
193
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
194
	    c = BYTE8_TO_CHAR (c);
195
	  pt_byte++;
196
	}
197
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
198

199
      return c;
Jim Blandy's avatar
Jim Blandy committed
200
    }
201
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
202
    {
203
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
204

205
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
206

207 208
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
209

Tom Tromey's avatar
Tom Tromey committed
210
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
211
	{
212
	  /* Fetch the character code from the buffer.  */
213 214
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
215
	  c = STRING_CHAR (p);
216 217
	  if (multibyte)
	    *multibyte = 1;
218 219 220 221
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
222
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
223
	    c = BYTE8_TO_CHAR (c);
224
	  bytepos++;
225 226
	}

227 228 229 230
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
231
    }
232 233

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
234 235 236 237
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
238

Jim Blandy's avatar
Jim Blandy committed
239
  if (EQ (readcharfun, Qget_file_char))
240
    {
Kenichi Handa's avatar
Kenichi Handa committed
241 242
      readbyte = readbyte_from_file;
      goto read_multibyte;
243
    }
Jim Blandy's avatar
Jim Blandy committed
244

245
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
246
    {
247
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
248
	c = -1;
249 250 251 252 253 254 255 256
      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);
	}
257
      else
258 259 260 261 262
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
263 264 265
      return c;
    }

Kenichi Handa's avatar
Kenichi Handa committed
266 267 268 269 270 271 272 273 274 275 276 277
  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
278

Kenichi Handa's avatar
Kenichi Handa committed
279 280 281 282 283 284
  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
285 286 287

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
288
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
289 290
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
291 292 293 294 295 296 297 298 299

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
300
  if (c < 0)
301 302 303
    return c;
  if (multibyte)
    *multibyte = 1;
304
  if (ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
    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;
    }
322
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
323 324
}

Stefan Monnier's avatar
Stefan Monnier committed
325 326 327 328
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

329 330 331
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
332
  if (FROM_FILE_P (readcharfun))
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
    {
      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');
    }
}

352 353 354 355 356 357 358 359 360 361 362 363 364
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
365 366 367 368
/* 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
369
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
370
{
371
  readchar_count--;
372 373 374 375
  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.  */
    ;
376
  else if (BUFFERP (readcharfun))
377
    {
378
      struct buffer *b = XBUFFER (readcharfun);
379 380
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
381

Tom Tromey's avatar
Tom Tromey committed
382
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
383
	BUF_DEC_POS (b, bytepos);
384
      else
Kenichi Handa's avatar
Kenichi Handa committed
385
	bytepos--;
386

387
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
388
    }
389
  else if (MARKERP (readcharfun))
390
    {
391
      struct buffer *b = XMARKER (readcharfun)->buffer;
392
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
393

Kenichi Handa's avatar
Kenichi Handa committed
394
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
395
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
396
	BUF_DEC_POS (b, bytepos);
397
      else
Kenichi Handa's avatar
Kenichi Handa committed
398
	bytepos--;
399

Kenichi Handa's avatar
Kenichi Handa committed
400
      XMARKER (readcharfun)->bytepos = bytepos;
401
    }
402
  else if (STRINGP (readcharfun))
403 404 405 406 407
    {
      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
408 409 410 411
  else if (CONSP (readcharfun))
    {
      unread_char = c;
    }
412
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
413 414 415
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
416
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
417
    {
418
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
419
    }
Jim Blandy's avatar
Jim Blandy committed
420 421 422 423
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
424
static int
425
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
426 427 428 429 430 431
{
  return read_bytecode_char (c >= 0);
}


static int
432
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
433 434 435
{
  if (c >= 0)
    {
436
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
437
      ungetc (c, instream);
438
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
439 440 441
      return 0;
    }

442
  block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
443
  c = getc (instream);
Miles Bader's avatar
Miles Bader committed
444

445
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
446 447
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
448
      unblock_input ();
Miles Bader's avatar
Miles Bader committed
449
      QUIT;
450
      block_input ();
Miles Bader's avatar
Miles Bader committed
451
      clearerr (instream);
Miles Bader's avatar
Miles Bader committed
452
      c = getc (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
453
    }
Miles Bader's avatar
Miles Bader committed
454

455
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
456

Kenichi Handa's avatar
Kenichi Handa committed
457 458 459 460
  return (c == EOF ? -1 : c);
}

static int
461
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
462 463 464 465 466 467 468 469 470
{
  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
471

Kenichi Handa's avatar
Kenichi Handa committed
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486
  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
487
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
{
  /* 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)
    {
516
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
517 518 519 520 521 522 523
      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)
	{
524
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
525 526 527 528
	  code = buf[2] & 0x7F;
	}
      else
	{
529
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
530 531 532 533 534
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
535
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
536
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
537 538 539 540
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
541
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
542 543 544 545
  return c;
}


546 547 548
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
549
static Lisp_Object read1 (Lisp_Object, int *, bool);
550

551 552
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
553

554 555 556 557 558
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);
559

Jim Blandy's avatar
Jim Blandy committed
560

561
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
562

563 564
/* Read input events until we get one that's acceptable for our purposes.

565
   If NO_SWITCH_FRAME, switch-frame events are stashed
566 567 568
   until we get a character we like, and then stuffed into
   unread_switch_frame.

569
   If ASCII_REQUIRED, check function key events to see
570 571 572
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

573 574 575
   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
576 577
   character.

578
   If INPUT_METHOD, invoke the current input method
579 580
   if the character warrants that.

581
   If SECONDS is a number, wait that many seconds for input, and
582
   return Qnil if no input arrives within that time.  */
583

Andreas Schwab's avatar
Andreas Schwab committed
584
static Lisp_Object
585 586
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
587
{
588
  Lisp_Object val, delayed_switch_frame;
589
  struct timespec end_time;
590

591
#ifdef HAVE_WINDOW_SYSTEM
592 593
  if (display_hourglass_p)
    cancel_hourglass ();
594
#endif
595

596
  delayed_switch_frame = Qnil;
597

598 599 600
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
601
      double duration = extract_float (seconds);
602 603
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
604 605
    }

606
  /* Read until we get an acceptable event.  */
607
 retry:
608
  do
609
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
610
		     NUMBERP (seconds) ? &end_time : NULL);
611
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
612

613
  if (BUFFERP (val))
614 615
    goto retry;

616
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
617
     character.  This is better than signaling an error just because
618 619 620 621 622
     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)
623
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
624 625 626 627 628
    {
      delayed_switch_frame = val;
      goto retry;
    }

629
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
630 631
    {
      /* Convert certain symbols to their ASCII equivalents.  */
632
      if (SYMBOLP (val))
633
	{
634
	  Lisp_Object tem, tem1;
635 636 637 638 639 640 641
	  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))
642
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
643 644
	    }
	}
645

646
      /* If we don't have a character now, deal with it appropriately.  */
647
      if (!INTEGERP (val))
648 649 650
	{
	  if (error_nonascii)
	    {
651
	      Vunread_command_events = list1 (val);
652 653 654 655 656 657 658 659 660 661
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

662 663
#if 0

664
#ifdef HAVE_WINDOW_SYSTEM
665 666
  if (display_hourglass_p)
    start_hourglass ();
667
#endif
668 669 670

#endif

671 672 673
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
674
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
675 676
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
677 678 679
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

680 681
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
682 683
exception, switch-frame events are put off until non-character events
can be read.
684 685 686 687 688 689
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
690 691 692 693 694
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
695
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
696
{
697 698
  Lisp_Object val;

699 700
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
701
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
702 703 704

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
705 706
}

Paul Eggert's avatar
Paul Eggert committed
707
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
708 709 710 711
       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
712 713 714 715 716
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
717
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
718
{
719 720
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
721
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
722 723
}

724
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
725 726
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
727 728
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
729 730 731 732

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
733 734 735 736 737
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
738
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
739
{
740 741
  Lisp_Object val;

742 743
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
744

745
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
746 747 748

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
749 750 751
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
752
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
753
  (void)
Jim Blandy's avatar
Jim Blandy committed
754 755
{
  register Lisp_Object val;
756
  block_input ();
757
  XSETINT (val, getc (instream));
758
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
759 760
  return val;
}
761 762


763 764 765 766


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
767 768
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
769

770
static bool
Stefan Monnier's avatar
Stefan Monnier committed
771
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
772 773
{
  int ch = READCHAR;
774 775 776 777 778 779 780 781 782 783 784 785 786

  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
787 788
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
789 790
    }

791 792 793 794 795 796 797 798 799
  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.  */
    {
800
      bool rv = 0;
801
      enum {
Paul Eggert's avatar
Paul Eggert committed
802
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
803
      } beg_end_state = NOMINAL;
804
      bool in_file_vars = 0;
805

806 807 808 809 810 811 812 813 814 815
#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;					\
816 817 818 819 820 821 822 823 824 825 826 827
    }

      /* 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)
	{
828
	  char var[100], val[100];
829
	  unsigned i;
830 831 832 833 834 835 836

	  ch = READCHAR;

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

837
	  i = 0;
838
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
839
	    {
840 841
	      if (i < sizeof var - 1)
		var[i++] = ch;
842 843 844
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
845

846
	  /* Stop scanning if no colon was found before end marker.  */
847
	  if (!in_file_vars || ch == '\n' || ch == EOF)
848 849
	    break;

850 851 852
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
853 854 855 856 857 858 859 860 861

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

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

862
	      i = 0;
863 864
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
865 866
		  if (i < sizeof val - 1)
		    val[i++] = ch;
867 868 869 870
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
871
		/* The value was terminated by an end-marker, which remove.  */
872 873 874 875
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';
876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891

	      if (strcmp (var, "lexical-binding") == 0)
		/* This is it...  */
		{
		  rv = (strcmp (val, "nil") != 0);
		  break;
		}
	    }
	}

      while (ch != '\n' && ch != EOF)
	ch = READCHAR;

      return rv;
    }
}
892

Kenichi Handa's avatar
Kenichi Handa committed
893
/* Value is a version number of byte compiled code if the file
Miles Bader's avatar
Miles Bader committed
894
   associated with file descriptor FD is a compiled Lisp file that's
Kenichi Handa's avatar
Kenichi Handa committed
895 896 897
   safe to load.  Only files compiled with Emacs are safe to load.
   Files compiled with XEmacs can lead to a crash in Fbyte_code
   because of an incompatible change in the byte compiler.  */
898 899

static int
900
safe_to_load_version (int fd)
901 902 903
{
  char buf[512];
  int nbytes, i;
Kenichi Handa's avatar
Kenichi Handa committed
904
  int version = 1;
905 906 907

  /* Read the first few bytes from the file, and look for a line
     specifying the byte compiler version used.  */
908
  nbytes = emacs_read (fd, buf, sizeof buf);
909 910 911
  if (nbytes > 0)
    {
      /* Skip to the next newline, skipping over the initial `ELC'
Dave Love's avatar
Dave Love committed
912
	 with NUL bytes following it, but note the version.  */
913
      for (i = 0; i < nbytes && buf[i] != '\n'; ++i)
Dave Love's avatar
Dave Love committed
914
	if (i == 4)
Kenichi Handa's avatar
Kenichi Handa committed
915
	  version = buf[i];
916

917
      if (i >= nbytes
918
	  || fast_c_string_match_ignore_case (Vbytecomp_version_regexp,
919
					      buf + i, nbytes - i) < 0)
920
	version = 0;
921 922 923
    }

  lseek (fd, 0, SEEK_SET);
924
  return version;
925 926 927
}


928 929 930
/* Callback for record_unwind_protect.  Restore the old load list OLD,
   after loading a file successfully.  */</