lread.c 145 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, 1997-2017 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"
Paul Eggert's avatar
Paul Eggert committed
26
#include <stdlib.h>
Jim Blandy's avatar
Jim Blandy committed
27 28 29
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/file.h>
30
#include <errno.h>
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"
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 "systime.h"
43
#include "termhooks.h"
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
44
#include "blockinput.h"
45
#include <c-ctype.h>
Jim Blandy's avatar
Jim Blandy committed
46

47 48
#ifdef MSDOS
#include "msdos.h"
Eli Zaretskii's avatar
Eli Zaretskii committed
49 50 51 52
#if __DJGPP__ == 2 && __DJGPP_MINOR__ < 5
# define INFINITY  __builtin_inf()
# define NAN       __builtin_nan("")
#endif
53 54
#endif

55 56 57 58
#ifdef HAVE_NS
#include "nsterm.h"
#endif

Andreas Schwab's avatar
Andreas Schwab committed
59
#include <unistd.h>
Jim Blandy's avatar
Jim Blandy committed
60

Richard M. Stallman's avatar
Richard M. Stallman committed
61 62 63 64
#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

Dave Love's avatar
Dave Love committed
65
#include <fcntl.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
66

67
#ifdef HAVE_FSEEKO
68 69 70 71 72 73 74
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

Ken Raeburn's avatar
Ken Raeburn committed
75 76 77 78
#ifndef HAVE_GETC_UNLOCKED
#define getc_unlocked getc
#endif

79 80 81 82
/* 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.  */
83
static Lisp_Object read_objects;
84

85
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
86 87
static FILE *instream;

88
/* For use within read-from-string (this reader is non-reentrant!!)  */
89 90 91
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
92

93
/* Number of characters read in the current call to Fread or
94
   Fread_from_string.  */
95
static EMACS_INT readchar_count;
96

97
/* This contains the last string skipped with #@.  */
98 99
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
100
static ptrdiff_t saved_doc_string_size;
101
/* Length of actual data in saved_doc_string.  */
102
static ptrdiff_t saved_doc_string_length;
103
/* This is the file position that string came from.  */
104
static file_offset saved_doc_string_position;
105

106 107 108 109 110
/* 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.  */
111
static ptrdiff_t prev_saved_doc_string_size;
112
/* Length of actual data in prev_saved_doc_string.  */
113
static ptrdiff_t prev_saved_doc_string_length;
114
/* This is the file position that string came from.  */
115
static file_offset prev_saved_doc_string_position;
116

117
/* True means inside a new-style backquote
118
   with no surrounding parentheses.
119
   Fread initializes this to false, so we need not specbind it
120
   or worry about what happens to it when there is an error.  */
121
static bool new_backquote_flag;
122 123 124 125

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

Lute Kamstra's avatar
Lute Kamstra committed
126
static Lisp_Object Vloads_in_progress;
127

128 129
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
130

131
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
132 133
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
134

Kenichi Handa's avatar
Kenichi Handa committed
135 136 137 138 139 140
/* 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.  */

141 142 143
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
144

Jim Blandy's avatar
Jim Blandy committed
145 146
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
147 148
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
149
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
150

151
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
152 153
#define UNREAD(c) unreadchar (readcharfun, c)

154 155 156
/* 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
157
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
158 159
   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
160
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
161 162
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
163
static int
164
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
165 166
{
  Lisp_Object tem;
167
  register int c;
168
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
169 170
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
171
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
172

173 174 175
  if (multibyte)
    *multibyte = 0;

176
  readchar_count++;
177

178
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
179
    {
180
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
181

182
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
183

184 185 186
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

187 188
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
189

Tom Tromey's avatar
Tom Tromey committed
190
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
191
	{
192
	  /* Fetch the character code from the buffer.  */
193 194
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
195
	  c = STRING_CHAR (p);
196 197
	  if (multibyte)
	    *multibyte = 1;
198 199 200 201
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
202
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
203
	    c = BYTE8_TO_CHAR (c);
204
	  pt_byte++;
205
	}
206
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
207

208
      return c;
Jim Blandy's avatar
Jim Blandy committed
209
    }
210
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
211
    {
212
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
213

214
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
215

216 217
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
218

Tom Tromey's avatar
Tom Tromey committed
219
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
220
	{
221
	  /* Fetch the character code from the buffer.  */
222 223
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
224
	  c = STRING_CHAR (p);
225 226
	  if (multibyte)
	    *multibyte = 1;
227 228 229 230
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
231
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
232
	    c = BYTE8_TO_CHAR (c);
233
	  bytepos++;
234 235
	}

236 237 238 239
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
240
    }
241 242

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
243 244 245 246
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
247

Jim Blandy's avatar
Jim Blandy committed
248
  if (EQ (readcharfun, Qget_file_char))
249
    {
Kenichi Handa's avatar
Kenichi Handa committed
250 251
      readbyte = readbyte_from_file;
      goto read_multibyte;
252
    }
Jim Blandy's avatar
Jim Blandy committed
253

254
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
255
    {
256
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
257
	c = -1;
258 259 260 261 262 263 264 265
      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);
	}
266
      else
267 268 269 270 271
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
272 273 274
      return c;
    }

275
  if (CONSP (readcharfun) && STRINGP (XCAR (readcharfun)))
Kenichi Handa's avatar
Kenichi Handa committed
276 277 278 279 280 281 282 283 284 285 286
    {
      /* 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
287

Kenichi Handa's avatar
Kenichi Handa committed
288 289 290 291 292 293
  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
294 295 296

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
297
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
298 299
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
300 301 302 303 304 305 306 307 308

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
309
  if (c < 0)
310 311 312
    return c;
  if (multibyte)
    *multibyte = 1;
313
  if (ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
    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;
    }
331
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
332 333
}

Stefan Monnier's avatar
Stefan Monnier committed
334 335 336 337
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

338 339 340
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
341
  if (FROM_FILE_P (readcharfun))
342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360
    {
      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');
    }
}

361 362 363 364 365 366 367 368 369 370 371 372 373
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
374 375 376 377
/* 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
378
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
379
{
380
  readchar_count--;
381 382 383 384
  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.  */
    ;
385
  else if (BUFFERP (readcharfun))
386
    {
387
      struct buffer *b = XBUFFER (readcharfun);
388 389
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
390

Tom Tromey's avatar
Tom Tromey committed
391
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
392
	BUF_DEC_POS (b, bytepos);
393
      else
Kenichi Handa's avatar
Kenichi Handa committed
394
	bytepos--;
395

396
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
397
    }
398
  else if (MARKERP (readcharfun))
399
    {
400
      struct buffer *b = XMARKER (readcharfun)->buffer;
401
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
402

Kenichi Handa's avatar
Kenichi Handa committed
403
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
404
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
405
	BUF_DEC_POS (b, bytepos);
406
      else
Kenichi Handa's avatar
Kenichi Handa committed
407
	bytepos--;
408

Kenichi Handa's avatar
Kenichi Handa committed
409
      XMARKER (readcharfun)->bytepos = bytepos;
410
    }
411
  else if (STRINGP (readcharfun))
412 413 414 415 416
    {
      read_from_string_index--;
      read_from_string_index_byte
	= string_char_to_byte (readcharfun, read_from_string_index);
    }
417
  else if (CONSP (readcharfun) && STRINGP (XCAR (readcharfun)))
Kenichi Handa's avatar
Kenichi Handa committed
418 419 420
    {
      unread_char = c;
    }
421
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
422 423 424
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
425
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
426
    {
427
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
428
    }
Jim Blandy's avatar
Jim Blandy committed
429 430 431 432
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
433
static int
434
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
435 436 437 438 439 440
{
  return read_bytecode_char (c >= 0);
}


static int
441
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
442 443 444
{
  if (c >= 0)
    {
445
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
446
      ungetc (c, instream);
447
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
448 449 450
      return 0;
    }

451
  block_input ();
Ken Raeburn's avatar
Ken Raeburn committed
452
  c = getc_unlocked (instream);
Miles Bader's avatar
Miles Bader committed
453

454
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
455 456
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
457
      unblock_input ();
Paul Eggert's avatar
Paul Eggert committed
458
      maybe_quit ();
459
      block_input ();
Miles Bader's avatar
Miles Bader committed
460
      clearerr (instream);
Ken Raeburn's avatar
Ken Raeburn committed
461
      c = getc_unlocked (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
462
    }
Miles Bader's avatar
Miles Bader committed
463

464
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
465

Kenichi Handa's avatar
Kenichi Handa committed
466 467 468 469
  return (c == EOF ? -1 : c);
}

static int
470
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
471 472 473 474 475 476 477 478 479
{
  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
480

Kenichi Handa's avatar
Kenichi Handa committed
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
  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
496
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524
{
  /* 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)
    {
525
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
526 527 528 529 530 531 532
      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)
	{
533
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
534 535 536 537
	  code = buf[2] & 0x7F;
	}
      else
	{
538
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
539 540 541 542 543
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
544
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
545
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
546 547 548 549
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
550
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
551 552 553 554
  return c;
}


555 556 557
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
558
static Lisp_Object read1 (Lisp_Object, int *, bool);
559

560 561
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
562

563 564 565
static Lisp_Object substitute_object_recurse (Lisp_Object, Lisp_Object,
                                              Lisp_Object);
static void substitute_in_interval (INTERVAL, Lisp_Object);
566

Jim Blandy's avatar
Jim Blandy committed
567

568
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
569

570 571
/* Read input events until we get one that's acceptable for our purposes.

572
   If NO_SWITCH_FRAME, switch-frame events are stashed
573 574 575
   until we get a character we like, and then stuffed into
   unread_switch_frame.

576
   If ASCII_REQUIRED, check function key events to see
577 578 579
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

580 581 582
   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
583 584
   character.

585
   If INPUT_METHOD, invoke the current input method
586 587
   if the character warrants that.

588
   If SECONDS is a number, wait that many seconds for input, and
589
   return Qnil if no input arrives within that time.  */
590

Andreas Schwab's avatar
Andreas Schwab committed
591
static Lisp_Object
592 593
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
594
{
595
  Lisp_Object val, delayed_switch_frame;
596
  struct timespec end_time;
597

598
#ifdef HAVE_WINDOW_SYSTEM
599 600
  if (display_hourglass_p)
    cancel_hourglass ();
601
#endif
602

603
  delayed_switch_frame = Qnil;
604

605 606 607
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
608
      double duration = XFLOATINT (seconds);
609 610
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
611 612
    }

613
  /* Read until we get an acceptable event.  */
614
 retry:
615
  do
616
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
617
		     NUMBERP (seconds) ? &end_time : NULL);
618
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
619

620
  if (BUFFERP (val))
621 622
    goto retry;

623
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
624
     character.  This is better than signaling an error just because
625 626 627 628 629
     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)
630
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
631 632 633 634 635
    {
      delayed_switch_frame = val;
      goto retry;
    }

636
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
637 638
    {
      /* Convert certain symbols to their ASCII equivalents.  */
639
      if (SYMBOLP (val))
640
	{
641
	  Lisp_Object tem, tem1;
642 643 644 645 646 647 648
	  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))
649
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
650 651
	    }
	}
652

653
      /* If we don't have a character now, deal with it appropriately.  */
654
      if (!INTEGERP (val))
655 656 657
	{
	  if (error_nonascii)
	    {
658
	      Vunread_command_events = list1 (val);
659 660 661 662 663 664 665 666 667 668
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

669 670
#if 0

671
#ifdef HAVE_WINDOW_SYSTEM
672 673
  if (display_hourglass_p)
    start_hourglass ();
674
#endif
675 676 677

#endif

678 679 680
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
681
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
682 683
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
684 685 686
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

687 688
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
689 690
exception, switch-frame events are put off until non-character events
can be read.
691 692 693 694 695 696
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
697 698 699 700 701
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
702
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
703
{
704 705
  Lisp_Object val;

706 707
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
708
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
709 710 711

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
712 713
}

Paul Eggert's avatar
Paul Eggert committed
714
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
715 716 717 718
       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
719 720 721 722 723
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
724
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
725
{
726 727
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
728
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
729 730
}

731
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
732 733
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
734 735
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
736 737 738 739

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
740 741 742 743 744
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
745
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
746
{
747 748
  Lisp_Object val;

749 750
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
751

752
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
753 754 755

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
756 757 758
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
759
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
760
  (void)
Jim Blandy's avatar
Jim Blandy committed
761 762
{
  register Lisp_Object val;
763
  block_input ();
Ken Raeburn's avatar
Ken Raeburn committed
764
  XSETINT (val, getc_unlocked (instream));
765
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
766 767
  return val;
}
768 769


770 771 772 773


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
774 775
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
776

777
static bool
Stefan Monnier's avatar
Stefan Monnier committed
778
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
779 780
{
  int ch = READCHAR;
781 782 783 784 785 786 787 788 789 790 791 792 793

  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
794 795
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
796 797
    }

798 799 800 801 802 803 804 805 806
  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.  */
    {
807
      bool rv = 0;
808
      enum {
Paul Eggert's avatar
Paul Eggert committed
809
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
810
      } beg_end_state = NOMINAL;
811
      bool in_file_vars = 0;
812

813 814 815 816 817 818 819 820 821 822
#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;					\
823 824 825 826 827 828 829 830 831 832 833 834
    }

      /* 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)
	{
835
	  char var[100], val[100];
836
	  unsigned i;
837 838 839 840 841 842 843

	  ch = READCHAR;

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

844
	  i = 0;
845
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
846
	    {
847 848
	      if (i < sizeof var - 1)
		var[i++] = ch;
849 850 851
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
852

853
	  /* Stop scanning if no colon was found before end marker.  */
854
	  if (!in_file_vars || ch == '\n' || ch == EOF)
855 856
	    break;

857 858 859
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
860 861 862 863 864 865 866 867 868

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

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

869
	      i = 0;
870 871
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
872 873
		  if (i < sizeof val - 1)
		    val[i++] = ch;
874 875 876 877
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
878
		/* The value was terminated by an end-marker, which remove.  */
879 880 881 882
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';
883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898

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

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

      return rv;
    }
}
899

Kenichi Handa's avatar
Kenichi Handa committed
900
/* Value is a version number of byte compiled code if the file
Miles Bader's avatar
Miles Bader committed
901
   associated with file descriptor FD is a compiled Lisp file that's
Kenichi Handa's avatar
Kenichi Handa committed
902 903 904
   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.  */
905 906

static int
907
safe_to_load_version (int fd)
908 909 910
{
  char buf[512];
  int nbytes, i;
Kenichi Handa's avatar
Kenichi Handa committed
911
  int version = 1;
912 913 914

  /* Read the first few bytes from the file, and look for a line
     specifying the byte compiler version used.  */
Paul Eggert's avatar
Paul Eggert committed
915
  nbytes = emacs_read_quit (fd, buf, sizeof buf);
916 917 918
  if (nbytes > 0)
    {
      /* Skip to the next newline, skipping over the initial `ELC'
Dave Love's avatar
Dave Love committed
919
	 with NUL bytes following it, but note the version.  */
920
      for (i = 0; i < nbytes && buf[i] != '\n'; ++i)
Dave Love's avatar
Dave Love committed
921
	if (i == 4)
Kenichi Handa's avatar
Kenichi Handa committed
922
	  version = buf[i];
923

924
      if (i >= nbytes
925
	  || fast_c_string_match_ignore_case (Vbytecomp_version_regexp,
926
					      buf + i, nbytes - i) < 0)