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

3
Copyright (C) 1985-1989, 1993-1995, 1997-2013 Free Software Foundation, Inc.
Jim Blandy's avatar
Jim Blandy committed
4 5 6

This file is part of GNU Emacs.

7
GNU Emacs is free software: you can redistribute it and/or modify
Jim Blandy's avatar
Jim Blandy committed
8
it under the terms of the GNU General Public License as published by
9 10
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Jim Blandy's avatar
Jim Blandy committed
11 12 13 14 15 16 17

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
18
along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
Jim Blandy's avatar
Jim Blandy committed
19 20


Richard M. Stallman's avatar
Richard M. Stallman committed
21
#include <config.h>
22
#include "sysstdio.h"
Jim Blandy's avatar
Jim Blandy committed
23 24 25
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/file.h>
26
#include <errno.h>
27
#include <limits.h>	/* For CHAR_BIT.  */
28
#include <stat-time.h>
Jim Blandy's avatar
Jim Blandy committed
29
#include "lisp.h"
30
#include "intervals.h"
Kenichi Handa's avatar
Kenichi Handa committed
31
#include "character.h"
32
#include "buffer.h"
Karl Heuer's avatar
Karl Heuer committed
33
#include "charset.h"
Kenichi Handa's avatar
Kenichi Handa committed
34
#include "coding.h"
35
#include <epaths.h>
Jim Blandy's avatar
Jim Blandy committed
36
#include "commands.h"
Jim Blandy's avatar
Jim Blandy committed
37
#include "keyboard.h"
38
#include "frame.h"
39
#include "termhooks.h"
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
40
#include "blockinput.h"
Jim Blandy's avatar
Jim Blandy committed
41

42 43 44 45
#ifdef MSDOS
#include "msdos.h"
#endif

46 47 48 49
#ifdef HAVE_NS
#include "nsterm.h"
#endif

Andreas Schwab's avatar
Andreas Schwab committed
50
#include <unistd.h>
Jim Blandy's avatar
Jim Blandy committed
51

Richard M. Stallman's avatar
Richard M. Stallman committed
52 53 54 55
#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

Dave Love's avatar
Dave Love committed
56
#include <fcntl.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
57

58
#ifdef HAVE_FSEEKO
59 60 61 62 63 64 65
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

66
/* Hash table read constants.  */
67 68 69 70 71 72 73 74
static Lisp_Object Qhash_table, Qdata;
static Lisp_Object Qtest, Qsize;
static Lisp_Object Qweakness;
static Lisp_Object Qrehash_size;
static Lisp_Object Qrehash_threshold;

static Lisp_Object Qread_char, Qget_file_char, Qcurrent_load_list;
Lisp_Object Qstandard_input;
75
Lisp_Object Qvariable_documentation;
76
static Lisp_Object Qascii_character, Qload, Qload_file_name;
77
Lisp_Object Qbackquote, Qcomma, Qcomma_at, Qcomma_dot, Qfunction;
78 79
static Lisp_Object Qinhibit_file_name_operation;
static Lisp_Object Qeval_buffer_list;
80
Lisp_Object Qlexical_binding;
81
static Lisp_Object Qfile_truename, Qdo_after_load_evaluation; /* ACM 2006/5/16 */
82

Kenichi Handa's avatar
Kenichi Handa committed
83 84 85 86 87 88
/* Used instead of Qget_file_char while loading *.elc files compiled
   by Emacs 21 or older.  */
static Lisp_Object Qget_emacs_mule_file_char;

static Lisp_Object Qload_force_doc_strings;

89
static Lisp_Object Qload_in_progress;
Ken Raeburn's avatar
Ken Raeburn committed
90

91 92 93 94
/* 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.  */
95
static Lisp_Object read_objects;
96

97
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
98 99
static FILE *instream;

100
/* For use within read-from-string (this reader is non-reentrant!!)  */
101 102 103
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
104

105
/* Number of characters read in the current call to Fread or
106
   Fread_from_string.  */
107
static EMACS_INT readchar_count;
108

109
/* This contains the last string skipped with #@.  */
110 111
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
112
static ptrdiff_t saved_doc_string_size;
113
/* Length of actual data in saved_doc_string.  */
114
static ptrdiff_t saved_doc_string_length;
115
/* This is the file position that string came from.  */
116
static file_offset saved_doc_string_position;
117

118 119 120 121 122
/* 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.  */
123
static ptrdiff_t prev_saved_doc_string_size;
124
/* Length of actual data in prev_saved_doc_string.  */
125
static ptrdiff_t prev_saved_doc_string_length;
126
/* This is the file position that string came from.  */
127
static file_offset prev_saved_doc_string_position;
128

129
/* True means inside a new-style backquote
130
   with no surrounding parentheses.
131
   Fread initializes this to false, so we need not specbind it
132
   or worry about what happens to it when there is an error.  */
133
static bool new_backquote_flag;
134
static Lisp_Object Qold_style_backquotes;
135 136 137 138

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

Lute Kamstra's avatar
Lute Kamstra committed
139
static Lisp_Object Vloads_in_progress;
140

141 142
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
143

144
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
145 146
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
147

Kenichi Handa's avatar
Kenichi Handa committed
148 149 150 151 152 153
/* 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.  */

154 155 156
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
157

Jim Blandy's avatar
Jim Blandy committed
158 159
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
160 161
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
162
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
163

164
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
165 166
#define UNREAD(c) unreadchar (readcharfun, c)

167 168 169
/* 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
170
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
171 172
   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
173
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
174 175
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
176
static int
177
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
178 179
{
  Lisp_Object tem;
180
  register int c;
181
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
182 183
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
184
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
185

186 187 188
  if (multibyte)
    *multibyte = 0;

189
  readchar_count++;
190

191
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
192
    {
193
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
194

195
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
196

197 198 199
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

200 201
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
202

Tom Tromey's avatar
Tom Tromey committed
203
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
204
	{
205
	  /* Fetch the character code from the buffer.  */
206 207
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
208
	  c = STRING_CHAR (p);
209 210
	  if (multibyte)
	    *multibyte = 1;
211 212 213 214
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
Kenichi Handa's avatar
Kenichi Handa committed
215 216
	  if (! ASCII_BYTE_P (c))
	    c = BYTE8_TO_CHAR (c);
217
	  pt_byte++;
218
	}
219
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
220

221
      return c;
Jim Blandy's avatar
Jim Blandy committed
222
    }
223
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
224
    {
225
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
226

227
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
228

229 230
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
231

Tom Tromey's avatar
Tom Tromey committed
232
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
233
	{
234
	  /* Fetch the character code from the buffer.  */
235 236
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
237
	  c = STRING_CHAR (p);
238 239
	  if (multibyte)
	    *multibyte = 1;
240 241 242 243
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
Kenichi Handa's avatar
Kenichi Handa committed
244 245
	  if (! ASCII_BYTE_P (c))
	    c = BYTE8_TO_CHAR (c);
246
	  bytepos++;
247 248
	}

249 250 251 252
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
253
    }
254 255

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
256 257 258 259
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
260

Jim Blandy's avatar
Jim Blandy committed
261
  if (EQ (readcharfun, Qget_file_char))
262
    {
Kenichi Handa's avatar
Kenichi Handa committed
263 264
      readbyte = readbyte_from_file;
      goto read_multibyte;
265
    }
Jim Blandy's avatar
Jim Blandy committed
266

267
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
268
    {
269
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
270
	c = -1;
271 272 273 274 275 276 277 278
      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);
	}
279
      else
280 281 282 283 284
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
285 286 287
      return c;
    }

Kenichi Handa's avatar
Kenichi Handa committed
288 289 290 291 292 293 294 295 296 297 298 299
  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
300

Kenichi Handa's avatar
Kenichi Handa committed
301 302 303 304 305 306
  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
307 308 309

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
310
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
311 312
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
313 314 315 316 317 318 319 320 321

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
322
  if (c < 0)
323 324 325 326
    return c;
  if (multibyte)
    *multibyte = 1;
  if (ASCII_BYTE_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343
    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;
    }
344
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
345 346
}

Stefan Monnier's avatar
Stefan Monnier committed
347 348 349 350
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

351 352 353
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
354
  if (FROM_FILE_P (readcharfun))
355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373
    {
      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');
    }
}

374 375 376 377 378 379 380 381 382 383 384 385 386
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
387 388 389 390
/* 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
391
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
392
{
393
  readchar_count--;
394 395 396 397
  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.  */
    ;
398
  else if (BUFFERP (readcharfun))
399
    {
400
      struct buffer *b = XBUFFER (readcharfun);
401 402
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
403

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

409
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
410
    }
411
  else if (MARKERP (readcharfun))
412
    {
413
      struct buffer *b = XMARKER (readcharfun)->buffer;
414
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
415

Kenichi Handa's avatar
Kenichi Handa committed
416
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
417
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
418
	BUF_DEC_POS (b, bytepos);
419
      else
Kenichi Handa's avatar
Kenichi Handa committed
420
	bytepos--;
421

Kenichi Handa's avatar
Kenichi Handa committed
422
      XMARKER (readcharfun)->bytepos = bytepos;
423
    }
424
  else if (STRINGP (readcharfun))
425 426 427 428 429
    {
      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
430 431 432 433
  else if (CONSP (readcharfun))
    {
      unread_char = c;
    }
434
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
435 436 437
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
438
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
439
    {
440
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
441
    }
Jim Blandy's avatar
Jim Blandy committed
442 443 444 445
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
446
static int
447
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
448 449 450 451 452 453
{
  return read_bytecode_char (c >= 0);
}


static int
454
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
455 456 457
{
  if (c >= 0)
    {
458
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
459
      ungetc (c, instream);
460
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
461 462 463
      return 0;
    }

464
  block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
465
  c = getc (instream);
Miles Bader's avatar
Miles Bader committed
466

467
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
468 469
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
470
      unblock_input ();
Miles Bader's avatar
Miles Bader committed
471
      QUIT;
472
      block_input ();
Miles Bader's avatar
Miles Bader committed
473
      clearerr (instream);
Miles Bader's avatar
Miles Bader committed
474
      c = getc (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
475
    }
Miles Bader's avatar
Miles Bader committed
476

477
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
478

Kenichi Handa's avatar
Kenichi Handa committed
479 480 481 482
  return (c == EOF ? -1 : c);
}

static int
483
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
484 485 486 487 488 489 490 491 492
{
  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
493

Kenichi Handa's avatar
Kenichi Handa committed
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
  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
509
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537
{
  /* 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)
    {
538
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
539 540 541 542 543 544 545
      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)
	{
546
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
547 548 549 550
	  code = buf[2] & 0x7F;
	}
      else
	{
551
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
552 553 554 555 556
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
557
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
558
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
559 560 561 562
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
563
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
564 565 566 567
  return c;
}


568 569 570
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
571
static Lisp_Object read1 (Lisp_Object, int *, bool);
572

573 574
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
575

576 577 578 579 580
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);
581

Jim Blandy's avatar
Jim Blandy committed
582

583
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
584

585 586
/* Read input events until we get one that's acceptable for our purposes.

587
   If NO_SWITCH_FRAME, switch-frame events are stashed
588 589 590
   until we get a character we like, and then stuffed into
   unread_switch_frame.

591
   If ASCII_REQUIRED, check function key events to see
592 593 594
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

595 596 597
   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
598 599
   character.

600
   If INPUT_METHOD, invoke the current input method
601 602
   if the character warrants that.

603
   If SECONDS is a number, wait that many seconds for input, and
604
   return Qnil if no input arrives within that time.  */
605

Andreas Schwab's avatar
Andreas Schwab committed
606
static Lisp_Object
607 608
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
609
{
610
  Lisp_Object val, delayed_switch_frame;
611
  struct timespec end_time;
612

613
#ifdef HAVE_WINDOW_SYSTEM
614 615
  if (display_hourglass_p)
    cancel_hourglass ();
616
#endif
617

618
  delayed_switch_frame = Qnil;
619

620 621 622
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
623
      double duration = extract_float (seconds);
624 625
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
626 627
    }

628
  /* Read until we get an acceptable event.  */
629
 retry:
630
  do
631
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
632
		     NUMBERP (seconds) ? &end_time : NULL);
633
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
634

635
  if (BUFFERP (val))
636 637
    goto retry;

638
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
639
     character.  This is better than signaling an error just because
640 641 642 643 644
     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)
645
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
646 647 648 649 650
    {
      delayed_switch_frame = val;
      goto retry;
    }

651
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
652 653
    {
      /* Convert certain symbols to their ASCII equivalents.  */
654
      if (SYMBOLP (val))
655
	{
656
	  Lisp_Object tem, tem1;
657 658 659 660 661 662 663
	  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))
664
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
665 666
	    }
	}
667

668
      /* If we don't have a character now, deal with it appropriately.  */
669
      if (!INTEGERP (val))
670 671 672
	{
	  if (error_nonascii)
	    {
673
	      Vunread_command_events = list1 (val);
674 675 676 677 678 679 680 681 682 683
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

684 685
#if 0

686
#ifdef HAVE_WINDOW_SYSTEM
687 688
  if (display_hourglass_p)
    start_hourglass ();
689
#endif
690 691 692

#endif

693 694 695
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
696
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
697 698
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
699 700 701
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

702 703
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
704 705
exception, switch-frame events are put off until non-character events
can be read.
706 707 708 709 710 711
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
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
  Lisp_Object val;

721 722
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
723
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
724 725 726

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
727 728
}

Paul Eggert's avatar
Paul Eggert committed
729
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
730 731 732 733
       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
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
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
743
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
744 745
}

746
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
747 748
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
749 750
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
751 752 753 754

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
755 756 757 758 759
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
760
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
761
{
762 763
  Lisp_Object val;

764 765
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
766

767
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
768 769 770

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
771 772 773
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
774
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
775
  (void)
Jim Blandy's avatar
Jim Blandy committed
776 777
{
  register Lisp_Object val;
778
  block_input ();
779
  XSETINT (val, getc (instream));
780
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
781 782
  return val;
}
783 784


785 786 787 788


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
789 790
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
791

792
static bool
Stefan Monnier's avatar
Stefan Monnier committed
793
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
794 795
{
  int ch = READCHAR;
796 797 798 799 800 801 802 803 804 805 806 807 808

  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
809 810
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
811 812
    }

813 814 815 816 817 818 819 820 821
  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.  */
    {
822
      bool rv = 0;
823
      enum {
Paul Eggert's avatar
Paul Eggert committed
824
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
825
      } beg_end_state = NOMINAL;
826
      bool in_file_vars = 0;
827

828 829 830 831 832 833 834 835 836 837
#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;					\
838 839 840 841 842 843 844 845 846 847 848 849
    }

      /* 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)
	{
850
	  char var[100], val[100];
851
	  unsigned i;
852 853 854 855 856 857 858

	  ch = READCHAR;

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

859
	  i = 0;
860
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
861
	    {
862 863
	      if (i < sizeof var - 1)
		var[i++] = ch;
864 865 866
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
867

868
	  /* Stop scanning if no colon was found before end marker.  */
869
	  if (!in_file_vars || ch == '\n' || ch == EOF)
870 871
	    break;

872 873 874
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
875 876 877 878 879 880 881 882 883

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

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

884
	      i = 0;
885 886
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
887 888
		  if (i < sizeof val - 1)
		    val[i++] = ch;
889 890 891 892
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
893
		/* The value was terminated by an end-marker, which remove.  */
894 895 896 897
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';
898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913

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

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

      return rv;
    }
}
914

Kenichi Handa's avatar
Kenichi Handa committed
915
/* Value is a version number of byte compiled code if the file
Miles Bader's avatar
Miles Bader committed
916
   associated with file descriptor FD is a compiled Lisp file that's
Kenichi Handa's avatar
Kenichi Handa committed
917 918 919
   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.  */
920 921

static int
922
safe_to_load_version (int fd)
923 924 925
{
  char buf[512];
  int nbytes, i;
Kenichi Handa's avatar
Kenichi Handa committed
926
  int version = 1;
927 928 929

  /* Read the first few bytes from the file, and look for a line
     specifying the byte compiler version used.  */
930
  nbytes = emacs_read (fd, buf, sizeof buf);
931 932 933
  if (nbytes > 0)
    {
      /* Skip to the next newline, skipping over the initial `ELC'
Dave Love's avatar
Dave Love committed
934
	 with NUL bytes following it, but note the version.  */
935
      for (i = 0; i < nbytes && buf[i] != '\n'; ++i)
Dave Love's avatar
Dave Love committed
936
	if (i == 4)
Kenichi Handa's avatar
Kenichi Handa committed
937
	  version = buf[i];
Gerd Moellmann's avatar