lread.c 137 KB
Newer Older
Jim Blandy's avatar
Jim Blandy committed
1
/* Lisp parsing and input streams.
2

3 4
Copyright (C) 1985-1989, 1993-1995, 1997-2014 Free Software Foundation,
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


Richard M. Stallman's avatar
Richard M. Stallman committed
22
#include <config.h>
23
#include "sysstdio.h"
Jim Blandy's avatar
Jim Blandy committed
24 25 26
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/file.h>
27
#include <errno.h>
28
#include <limits.h>	/* For CHAR_BIT.  */
29
#include <stat-time.h>
Jim Blandy's avatar
Jim Blandy committed
30
#include "lisp.h"
31
#include "intervals.h"
Kenichi Handa's avatar
Kenichi Handa committed
32
#include "character.h"
33
#include "buffer.h"
Karl Heuer's avatar
Karl Heuer committed
34
#include "charset.h"
Kenichi Handa's avatar
Kenichi Handa committed
35
#include "coding.h"
36
#include <epaths.h>
Jim Blandy's avatar
Jim Blandy committed
37
#include "commands.h"
Jim Blandy's avatar
Jim Blandy committed
38
#include "keyboard.h"
39
#include "frame.h"
40
#include "termhooks.h"
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
41
#include "blockinput.h"
Jim Blandy's avatar
Jim Blandy committed
42

43 44 45 46
#ifdef MSDOS
#include "msdos.h"
#endif

47 48 49 50
#ifdef HAVE_NS
#include "nsterm.h"
#endif

Andreas Schwab's avatar
Andreas Schwab committed
51
#include <unistd.h>
Jim Blandy's avatar
Jim Blandy committed
52

Richard M. Stallman's avatar
Richard M. Stallman committed
53 54 55 56
#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

Dave Love's avatar
Dave Love committed
57
#include <fcntl.h>
Richard M. Stallman's avatar
Richard M. Stallman committed
58

59
#ifdef HAVE_FSEEKO
60 61 62 63 64 65 66
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

67
/* Hash table read constants.  */
68 69 70 71 72 73 74 75
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;
76
Lisp_Object Qvariable_documentation;
77
static Lisp_Object Qascii_character, Qload, Qload_file_name;
78
Lisp_Object Qbackquote, Qcomma, Qcomma_at, Qcomma_dot, Qfunction;
79 80
static Lisp_Object Qinhibit_file_name_operation;
static Lisp_Object Qeval_buffer_list;
81
Lisp_Object Qlexical_binding;
82
static Lisp_Object Qfile_truename, Qdo_after_load_evaluation; /* ACM 2006/5/16 */
83

Kenichi Handa's avatar
Kenichi Handa committed
84 85 86 87 88 89
/* 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;

90
static Lisp_Object Qload_in_progress;
Ken Raeburn's avatar
Ken Raeburn committed
91

92 93 94 95
/* 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.  */
96
static Lisp_Object read_objects;
97

98
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
99 100
static FILE *instream;

101
/* For use within read-from-string (this reader is non-reentrant!!)  */
102 103 104
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
105

106
/* Number of characters read in the current call to Fread or
107
   Fread_from_string.  */
108
static EMACS_INT readchar_count;
109

110
/* This contains the last string skipped with #@.  */
111 112
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
113
static ptrdiff_t saved_doc_string_size;
114
/* Length of actual data in saved_doc_string.  */
115
static ptrdiff_t saved_doc_string_length;
116
/* This is the file position that string came from.  */
117
static file_offset saved_doc_string_position;
118

119 120 121 122 123
/* 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.  */
124
static ptrdiff_t prev_saved_doc_string_size;
125
/* Length of actual data in prev_saved_doc_string.  */
126
static ptrdiff_t prev_saved_doc_string_length;
127
/* This is the file position that string came from.  */
128
static file_offset prev_saved_doc_string_position;
129

130
/* True means inside a new-style backquote
131
   with no surrounding parentheses.
132
   Fread initializes this to false, so we need not specbind it
133
   or worry about what happens to it when there is an error.  */
134
static bool new_backquote_flag;
135
static Lisp_Object Qold_style_backquotes;
136 137 138 139

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

Lute Kamstra's avatar
Lute Kamstra committed
140
static Lisp_Object Vloads_in_progress;
141

142 143
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
144

145
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
146 147
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
148

Kenichi Handa's avatar
Kenichi Handa committed
149 150 151 152 153 154
/* 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.  */

155 156 157
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
158

Jim Blandy's avatar
Jim Blandy committed
159 160
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
161 162
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
163
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
164

165
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
166 167
#define UNREAD(c) unreadchar (readcharfun, c)

168 169 170
/* 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
171
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
172 173
   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
174
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
175 176
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
177
static int
178
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
179 180
{
  Lisp_Object tem;
181
  register int c;
182
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
183 184
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
185
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
186

187 188 189
  if (multibyte)
    *multibyte = 0;

190
  readchar_count++;
191

192
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
193
    {
194
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
195

196
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
197

198 199 200
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

201 202
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
203

Tom Tromey's avatar
Tom Tromey committed
204
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
205
	{
206
	  /* Fetch the character code from the buffer.  */
207 208
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
209
	  c = STRING_CHAR (p);
210 211
	  if (multibyte)
	    *multibyte = 1;
212 213 214 215
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
Kenichi Handa's avatar
Kenichi Handa committed
216 217
	  if (! ASCII_BYTE_P (c))
	    c = BYTE8_TO_CHAR (c);
218
	  pt_byte++;
219
	}
220
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
221

222
      return c;
Jim Blandy's avatar
Jim Blandy committed
223
    }
224
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
225
    {
226
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
227

228
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
229

230 231
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
232

Tom Tromey's avatar
Tom Tromey committed
233
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
234
	{
235
	  /* Fetch the character code from the buffer.  */
236 237
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
238
	  c = STRING_CHAR (p);
239 240
	  if (multibyte)
	    *multibyte = 1;
241 242 243 244
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
Kenichi Handa's avatar
Kenichi Handa committed
245 246
	  if (! ASCII_BYTE_P (c))
	    c = BYTE8_TO_CHAR (c);
247
	  bytepos++;
248 249
	}

250 251 252 253
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
254
    }
255 256

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
257 258 259 260
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
261

Jim Blandy's avatar
Jim Blandy committed
262
  if (EQ (readcharfun, Qget_file_char))
263
    {
Kenichi Handa's avatar
Kenichi Handa committed
264 265
      readbyte = readbyte_from_file;
      goto read_multibyte;
266
    }
Jim Blandy's avatar
Jim Blandy committed
267

268
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
269
    {
270
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
271
	c = -1;
272 273 274 275 276 277 278 279
      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);
	}
280
      else
281 282 283 284 285
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
286 287 288
      return c;
    }

Kenichi Handa's avatar
Kenichi Handa committed
289 290 291 292 293 294 295 296 297 298 299 300
  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
301

Kenichi Handa's avatar
Kenichi Handa committed
302 303 304 305 306 307
  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
308 309 310

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
311
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
312 313
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
314 315 316 317 318 319 320 321 322

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
323
  if (c < 0)
324 325 326 327
    return c;
  if (multibyte)
    *multibyte = 1;
  if (ASCII_BYTE_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344
    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;
    }
345
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
346 347
}

Stefan Monnier's avatar
Stefan Monnier committed
348 349 350 351
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

352 353 354
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
355
  if (FROM_FILE_P (readcharfun))
356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
    {
      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');
    }
}

375 376 377 378 379 380 381 382 383 384 385 386 387
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
388 389 390 391
/* 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
392
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
393
{
394
  readchar_count--;
395 396 397 398
  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.  */
    ;
399
  else if (BUFFERP (readcharfun))
400
    {
401
      struct buffer *b = XBUFFER (readcharfun);
402 403
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
404

Tom Tromey's avatar
Tom Tromey committed
405
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
406
	BUF_DEC_POS (b, bytepos);
407
      else
Kenichi Handa's avatar
Kenichi Handa committed
408
	bytepos--;
409

410
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
411
    }
412
  else if (MARKERP (readcharfun))
413
    {
414
      struct buffer *b = XMARKER (readcharfun)->buffer;
415
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
416

Kenichi Handa's avatar
Kenichi Handa committed
417
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
418
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
419
	BUF_DEC_POS (b, bytepos);
420
      else
Kenichi Handa's avatar
Kenichi Handa committed
421
	bytepos--;
422

Kenichi Handa's avatar
Kenichi Handa committed
423
      XMARKER (readcharfun)->bytepos = bytepos;
424
    }
425
  else if (STRINGP (readcharfun))
426 427 428 429 430
    {
      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
431 432 433 434
  else if (CONSP (readcharfun))
    {
      unread_char = c;
    }
435
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
436 437 438
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
439
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
440
    {
441
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
442
    }
Jim Blandy's avatar
Jim Blandy committed
443 444 445 446
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
447
static int
448
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
449 450 451 452 453 454
{
  return read_bytecode_char (c >= 0);
}


static int
455
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
456 457 458
{
  if (c >= 0)
    {
459
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
460
      ungetc (c, instream);
461
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
462 463 464
      return 0;
    }

465
  block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
466
  c = getc (instream);
Miles Bader's avatar
Miles Bader committed
467

468
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
469 470
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
471
      unblock_input ();
Miles Bader's avatar
Miles Bader committed
472
      QUIT;
473
      block_input ();
Miles Bader's avatar
Miles Bader committed
474
      clearerr (instream);
Miles Bader's avatar
Miles Bader committed
475
      c = getc (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
476
    }
Miles Bader's avatar
Miles Bader committed
477

478
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
479

Kenichi Handa's avatar
Kenichi Handa committed
480 481 482 483
  return (c == EOF ? -1 : c);
}

static int
484
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
485 486 487 488 489 490 491 492 493
{
  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
494

Kenichi Handa's avatar
Kenichi Handa committed
495 496 497 498 499 500 501 502 503 504 505 506 507 508 509
  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
510
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
{
  /* 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)
    {
539
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
540 541 542 543 544 545 546
      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)
	{
547
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
548 549 550 551
	  code = buf[2] & 0x7F;
	}
      else
	{
552
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
553 554 555 556 557
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
558
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
559
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
560 561 562 563
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
564
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
565 566 567 568
  return c;
}


569 570 571
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
572
static Lisp_Object read1 (Lisp_Object, int *, bool);
573

574 575
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
576

577 578 579 580 581
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);
582

Jim Blandy's avatar
Jim Blandy committed
583

584
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
585

586 587
/* Read input events until we get one that's acceptable for our purposes.

588
   If NO_SWITCH_FRAME, switch-frame events are stashed
589 590 591
   until we get a character we like, and then stuffed into
   unread_switch_frame.

592
   If ASCII_REQUIRED, check function key events to see
593 594 595
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

596 597 598
   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
599 600
   character.

601
   If INPUT_METHOD, invoke the current input method
602 603
   if the character warrants that.

604
   If SECONDS is a number, wait that many seconds for input, and
605
   return Qnil if no input arrives within that time.  */
606

Andreas Schwab's avatar
Andreas Schwab committed
607
static Lisp_Object
608 609
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
610
{
611
  Lisp_Object val, delayed_switch_frame;
612
  struct timespec end_time;
613

614
#ifdef HAVE_WINDOW_SYSTEM
615 616
  if (display_hourglass_p)
    cancel_hourglass ();
617
#endif
618

619
  delayed_switch_frame = Qnil;
620

621 622 623
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
624
      double duration = extract_float (seconds);
625 626
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
627 628
    }

629
  /* Read until we get an acceptable event.  */
630
 retry:
631
  do
632
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
633
		     NUMBERP (seconds) ? &end_time : NULL);
634
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
635

636
  if (BUFFERP (val))
637 638
    goto retry;

639
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
640
     character.  This is better than signaling an error just because
641 642 643 644 645
     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)
646
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
647 648 649 650 651
    {
      delayed_switch_frame = val;
      goto retry;
    }

652
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
653 654
    {
      /* Convert certain symbols to their ASCII equivalents.  */
655
      if (SYMBOLP (val))
656
	{
657
	  Lisp_Object tem, tem1;
658 659 660 661 662 663 664
	  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))
665
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
666 667
	    }
	}
668

669
      /* If we don't have a character now, deal with it appropriately.  */
670
      if (!INTEGERP (val))
671 672 673
	{
	  if (error_nonascii)
	    {
674
	      Vunread_command_events = list1 (val);
675 676 677 678 679 680 681 682 683 684
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

685 686
#if 0

687
#ifdef HAVE_WINDOW_SYSTEM
688 689
  if (display_hourglass_p)
    start_hourglass ();
690
#endif
691 692 693

#endif

694 695 696
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
697
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
698 699
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
700 701 702
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

703 704
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
705 706
exception, switch-frame events are put off until non-character events
can be read.
707 708 709 710 711 712
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
713 714 715 716 717
is used for reading a character.
If the optional argument SECONDS is non-nil, it should be a number
specifying the maximum number of seconds to wait for input.  If no
input arrives in that time, return nil.  SECONDS may be a
floating-point value.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
718
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
719
{
720 721
  Lisp_Object val;

722 723
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
724
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
725 726 727

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
728 729
}

Paul Eggert's avatar
Paul Eggert committed
730
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
731 732 733 734
       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
735 736 737 738 739
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
740
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
741
{
742 743
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
744
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
745 746
}

747
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
748 749
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
750 751
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
752 753 754 755

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
756 757 758 759 760
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
761
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
762
{
763 764
  Lisp_Object val;

765 766
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
767

768
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
769 770 771

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
772 773 774
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
775
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
776
  (void)
Jim Blandy's avatar
Jim Blandy committed
777 778
{
  register Lisp_Object val;
779
  block_input ();
780
  XSETINT (val, getc (instream));
781
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
782 783
  return val;
}
784 785


786 787 788 789


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
790 791
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
792

793
static bool
Stefan Monnier's avatar
Stefan Monnier committed
794
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
795 796
{
  int ch = READCHAR;
797 798 799 800 801 802 803 804 805 806 807 808 809

  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
810 811
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
812 813
    }

814 815 816 817 818 819 820 821 822
  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.  */
    {
823
      bool rv = 0;
824
      enum {
Paul Eggert's avatar
Paul Eggert committed
825
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
826
      } beg_end_state = NOMINAL;
827
      bool in_file_vars = 0;
828

829 830 831 832 833 834 835 836 837 838
#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;					\
839 840 841 842 843 844 845 846 847 848 849 850
    }

      /* 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)
	{
851
	  char var[100], val[100];
852
	  unsigned i;
853 854 855 856 857 858 859

	  ch = READCHAR;

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

860
	  i = 0;
861
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
862
	    {
863 864
	      if (i < sizeof var - 1)
		var[i++] = ch;
865 866 867
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
868

869
	  /* Stop scanning if no colon was found before end marker.  */
870
	  if (!in_file_vars || ch == '\n' || ch == EOF)
871 872
	    break;

873 874 875
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
876 877 878 879 880 881 882 883 884

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

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

885
	      i = 0;
886 887
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
888 889
		  if (i < sizeof val - 1)
		    val[i++] = ch;
890 891 892 893
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
894
		/* The value was terminated by an end-marker, which remove.  */
895 896 897 898
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';
899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914

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

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

      return rv;
    }
}
915

Kenichi Handa's avatar
Kenichi Handa committed
916
/* Value is a version number of byte compiled code if the file
Miles Bader's avatar
Miles Bader committed
917
   associated with file descriptor FD is a compiled Lisp file that's
Kenichi Handa's avatar
Kenichi Handa committed
918 919 920
   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.  */
921 922

static int
923
safe_to_load_version (int fd)
924 925 926
{
  char buf[512];
  int nbytes, i;
Kenichi Handa's avatar
Kenichi Handa committed
927
  int version = 1;
928 929 930

  /* Read the first few bytes from the file, and look for a line
     specifying the byte compiler version used.  */
931
  nbytes = emacs_read (fd, buf, sizeof buf);
932 933 934
  if (nbytes > 0)
    {
      /* Skip to the next newline, skipping over the initial `ELC'
Dave Love's avatar
Dave Love committed
935
	 with NUL bytes following it, but note the version.  */
936
      for (i = 0; i < nbytes && buf[i] != '\n'; ++i)
Dave Love's avatar
Dave Love committed
937
	if (i == 4)
Kenichi Handa's avatar
Kenichi Handa committed
938
	  version = buf[i];