lread.c 144 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

75 76 77 78
/* 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.  */
79
static Lisp_Object read_objects;
80

81
/* File for get_file_char to read from.  Use by load.  */
Jim Blandy's avatar
Jim Blandy committed
82 83
static FILE *instream;

84
/* For use within read-from-string (this reader is non-reentrant!!)  */
85 86 87
static ptrdiff_t read_from_string_index;
static ptrdiff_t read_from_string_index_byte;
static ptrdiff_t read_from_string_limit;
88

89
/* Number of characters read in the current call to Fread or
90
   Fread_from_string.  */
91
static EMACS_INT readchar_count;
92

93
/* This contains the last string skipped with #@.  */
94 95
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
96
static ptrdiff_t saved_doc_string_size;
97
/* Length of actual data in saved_doc_string.  */
98
static ptrdiff_t saved_doc_string_length;
99
/* This is the file position that string came from.  */
100
static file_offset saved_doc_string_position;
101

102 103 104 105 106
/* 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.  */
107
static ptrdiff_t prev_saved_doc_string_size;
108
/* Length of actual data in prev_saved_doc_string.  */
109
static ptrdiff_t prev_saved_doc_string_length;
110
/* This is the file position that string came from.  */
111
static file_offset prev_saved_doc_string_position;
112

113
/* True means inside a new-style backquote
114
   with no surrounding parentheses.
115
   Fread initializes this to false, so we need not specbind it
116
   or worry about what happens to it when there is an error.  */
117
static bool new_backquote_flag;
118 119 120 121

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

Lute Kamstra's avatar
Lute Kamstra committed
122
static Lisp_Object Vloads_in_progress;
123

124 125
static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
126

127
static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
128 129
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);
Jim Blandy's avatar
Jim Blandy committed
130

Kenichi Handa's avatar
Kenichi Handa committed
131 132 133 134 135 136
/* 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.  */

137 138 139
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
140

Jim Blandy's avatar
Jim Blandy committed
141 142
/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
Karl Heuer's avatar
Karl Heuer committed
143 144
   UNREAD(c) to unread c to be read again.

Kenichi Handa's avatar
Kenichi Handa committed
145
   These macros correctly read/unread multibyte characters.  */
Jim Blandy's avatar
Jim Blandy committed
146

147
#define READCHAR readchar (readcharfun, NULL)
Jim Blandy's avatar
Jim Blandy committed
148 149
#define UNREAD(c) unreadchar (readcharfun, c)

150 151 152
/* 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
153
/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
Kenichi Handa's avatar
Kenichi Handa committed
154 155
   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
156
   means that there's no unread character.  */
Kenichi Handa's avatar
Kenichi Handa committed
157 158
static int unread_char;

Jim Blandy's avatar
Jim Blandy committed
159
static int
160
readchar (Lisp_Object readcharfun, bool *multibyte)
Jim Blandy's avatar
Jim Blandy committed
161 162
{
  Lisp_Object tem;
163
  register int c;
164
  int (*readbyte) (int, Lisp_Object);
Kenichi Handa's avatar
Kenichi Handa committed
165 166
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
167
  bool emacs_mule_encoding = 0;
Jim Blandy's avatar
Jim Blandy committed
168

169 170 171
  if (multibyte)
    *multibyte = 0;

172
  readchar_count++;
173

174
  if (BUFFERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
175
    {
176
      register struct buffer *inbuffer = XBUFFER (readcharfun);
Jim Blandy's avatar
Jim Blandy committed
177

178
      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);
179

180 181 182
      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

183 184
      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;
Jim Blandy's avatar
Jim Blandy committed
185

Tom Tromey's avatar
Tom Tromey committed
186
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
187
	{
188
	  /* Fetch the character code from the buffer.  */
189 190
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
191
	  c = STRING_CHAR (p);
192 193
	  if (multibyte)
	    *multibyte = 1;
194 195 196 197
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
198
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
199
	    c = BYTE8_TO_CHAR (c);
200
	  pt_byte++;
201
	}
202
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);
203

204
      return c;
Jim Blandy's avatar
Jim Blandy committed
205
    }
206
  if (MARKERP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
207
    {
208
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;
Jim Blandy's avatar
Jim Blandy committed
209

210
      ptrdiff_t bytepos = marker_byte_position (readcharfun);
211

212 213
      if (bytepos >= BUF_ZV_BYTE (inbuffer))
	return -1;
214

Tom Tromey's avatar
Tom Tromey committed
215
      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
216
	{
217
	  /* Fetch the character code from the buffer.  */
218 219
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
220
	  c = STRING_CHAR (p);
221 222
	  if (multibyte)
	    *multibyte = 1;
223 224 225 226
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
227
	  if (! ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
228
	    c = BYTE8_TO_CHAR (c);
229
	  bytepos++;
230 231
	}

232 233 234 235
      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
Jim Blandy's avatar
Jim Blandy committed
236
    }
237 238

  if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
239 240 241 242
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }
243

Jim Blandy's avatar
Jim Blandy committed
244
  if (EQ (readcharfun, Qget_file_char))
245
    {
Kenichi Handa's avatar
Kenichi Handa committed
246 247
      readbyte = readbyte_from_file;
      goto read_multibyte;
248
    }
Jim Blandy's avatar
Jim Blandy committed
249

250
  if (STRINGP (readcharfun))
Jim Blandy's avatar
Jim Blandy committed
251
    {
252
      if (read_from_string_index >= read_from_string_limit)
Jim Blandy's avatar
Jim Blandy committed
253
	c = -1;
254 255 256 257 258 259 260 261
      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);
	}
262
      else
263 264 265 266 267
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
Jim Blandy's avatar
Jim Blandy committed
268 269 270
      return c;
    }

271
  if (CONSP (readcharfun) && STRINGP (XCAR (readcharfun)))
Kenichi Handa's avatar
Kenichi Handa committed
272 273 274 275 276 277 278 279 280 281 282
    {
      /* 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
283

Kenichi Handa's avatar
Kenichi Handa committed
284 285 286 287 288 289
  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
290 291 292

  tem = call0 (readcharfun);

Jim Blandy's avatar
Jim Blandy committed
293
  if (NILP (tem))
Jim Blandy's avatar
Jim Blandy committed
294 295
    return -1;
  return XINT (tem);
Kenichi Handa's avatar
Kenichi Handa committed
296 297 298 299 300 301 302 303 304

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
305
  if (c < 0)
306 307 308
    return c;
  if (multibyte)
    *multibyte = 1;
309
  if (ASCII_CHAR_P (c))
Kenichi Handa's avatar
Kenichi Handa committed
310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
    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;
    }
327
  return STRING_CHAR (buf);
Jim Blandy's avatar
Jim Blandy committed
328 329
}

Stefan Monnier's avatar
Stefan Monnier committed
330 331 332 333
#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

334 335 336
static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
Stefan Monnier's avatar
Stefan Monnier committed
337
  if (FROM_FILE_P (readcharfun))
338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356
    {
      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');
    }
}

357 358 359 360 361 362 363 364 365 366 367 368 369
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
370 371 372 373
/* 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
374
unreadchar (Lisp_Object readcharfun, int c)
Jim Blandy's avatar
Jim Blandy committed
375
{
376
  readchar_count--;
377 378 379 380
  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.  */
    ;
381
  else if (BUFFERP (readcharfun))
382
    {
383
      struct buffer *b = XBUFFER (readcharfun);
384 385
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);
386

Tom Tromey's avatar
Tom Tromey committed
387
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
388
	BUF_DEC_POS (b, bytepos);
389
      else
Kenichi Handa's avatar
Kenichi Handa committed
390
	bytepos--;
391

392
      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
393
    }
394
  else if (MARKERP (readcharfun))
395
    {
396
      struct buffer *b = XMARKER (readcharfun)->buffer;
397
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;
398

Kenichi Handa's avatar
Kenichi Handa committed
399
      XMARKER (readcharfun)->charpos--;
Tom Tromey's avatar
Tom Tromey committed
400
      if (! NILP (BVAR (b, enable_multibyte_characters)))
Kenichi Handa's avatar
Kenichi Handa committed
401
	BUF_DEC_POS (b, bytepos);
402
      else
Kenichi Handa's avatar
Kenichi Handa committed
403
	bytepos--;
404

Kenichi Handa's avatar
Kenichi Handa committed
405
      XMARKER (readcharfun)->bytepos = bytepos;
406
    }
407
  else if (STRINGP (readcharfun))
408 409 410 411 412
    {
      read_from_string_index--;
      read_from_string_index_byte
	= string_char_to_byte (readcharfun, read_from_string_index);
    }
413
  else if (CONSP (readcharfun) && STRINGP (XCAR (readcharfun)))
Kenichi Handa's avatar
Kenichi Handa committed
414 415 416
    {
      unread_char = c;
    }
417
  else if (EQ (readcharfun, Qlambda))
Kenichi Handa's avatar
Kenichi Handa committed
418 419 420
    {
      unread_char = c;
    }
Stefan Monnier's avatar
Stefan Monnier committed
421
  else if (FROM_FILE_P (readcharfun))
Kenichi Handa's avatar
Kenichi Handa committed
422
    {
423
      unread_char = c;
Kenichi Handa's avatar
Kenichi Handa committed
424
    }
Jim Blandy's avatar
Jim Blandy committed
425 426 427 428
  else
    call1 (readcharfun, make_number (c));
}

Kenichi Handa's avatar
Kenichi Handa committed
429
static int
430
readbyte_for_lambda (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
431 432 433 434 435 436
{
  return read_bytecode_char (c >= 0);
}


static int
437
readbyte_from_file (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
438 439 440
{
  if (c >= 0)
    {
441
      block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
442
      ungetc (c, instream);
443
      unblock_input ();
Kenichi Handa's avatar
Kenichi Handa committed
444 445 446
      return 0;
    }

447
  block_input ();
Kenichi Handa's avatar
Kenichi Handa committed
448
  c = getc (instream);
Miles Bader's avatar
Miles Bader committed
449

450
  /* Interrupted reads have been observed while reading over the network.  */
Miles Bader's avatar
Miles Bader committed
451 452
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
453
      unblock_input ();
Paul Eggert's avatar
Paul Eggert committed
454
      maybe_quit ();
455
      block_input ();
Miles Bader's avatar
Miles Bader committed
456
      clearerr (instream);
Miles Bader's avatar
Miles Bader committed
457
      c = getc (instream);
YAMAMOTO Mitsuharu's avatar
YAMAMOTO Mitsuharu committed
458
    }
Miles Bader's avatar
Miles Bader committed
459

460
  unblock_input ();
Miles Bader's avatar
Miles Bader committed
461

Kenichi Handa's avatar
Kenichi Handa committed
462 463 464 465
  return (c == EOF ? -1 : c);
}

static int
466
readbyte_from_string (int c, Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
467 468 469 470 471 472 473 474 475
{
  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
476

Kenichi Handa's avatar
Kenichi Handa committed
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
  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
492
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
Kenichi Handa's avatar
Kenichi Handa committed
493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520
{
  /* 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)
    {
521
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
522 523 524 525 526 527 528
      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)
	{
529
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
530 531 532 533
	  code = buf[2] & 0x7F;
	}
      else
	{
534
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
Kenichi Handa's avatar
Kenichi Handa committed
535 536 537 538 539
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
540
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
Kenichi Handa's avatar
Kenichi Handa committed
541
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
Kenichi Handa's avatar
Kenichi Handa committed
542 543 544 545
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
546
	     list1 (build_string ("invalid multibyte form")));
Kenichi Handa's avatar
Kenichi Handa committed
547 548 549 550
  return c;
}


551 552 553
static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
554
static Lisp_Object read1 (Lisp_Object, int *, bool);
555

556 557
static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);
558

559 560 561
static Lisp_Object substitute_object_recurse (Lisp_Object, Lisp_Object,
                                              Lisp_Object);
static void substitute_in_interval (INTERVAL, Lisp_Object);
562

Jim Blandy's avatar
Jim Blandy committed
563

564
/* Get a character from the tty.  */
Jim Blandy's avatar
Jim Blandy committed
565

566 567
/* Read input events until we get one that's acceptable for our purposes.

568
   If NO_SWITCH_FRAME, switch-frame events are stashed
569 570 571
   until we get a character we like, and then stuffed into
   unread_switch_frame.

572
   If ASCII_REQUIRED, check function key events to see
573 574 575
   if the unmodified version of the symbol has a Qascii_character
   property, and use that character, if present.

576 577 578
   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
579 580
   character.

581
   If INPUT_METHOD, invoke the current input method
582 583
   if the character warrants that.

584
   If SECONDS is a number, wait that many seconds for input, and
585
   return Qnil if no input arrives within that time.  */
586

Andreas Schwab's avatar
Andreas Schwab committed
587
static Lisp_Object
588 589
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
590
{
591
  Lisp_Object val, delayed_switch_frame;
592
  struct timespec end_time;
593

594
#ifdef HAVE_WINDOW_SYSTEM
595 596
  if (display_hourglass_p)
    cancel_hourglass ();
597
#endif
598

599
  delayed_switch_frame = Qnil;
600

601 602 603
  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
604
      double duration = XFLOATINT (seconds);
605 606
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
607 608
    }

609
  /* Read until we get an acceptable event.  */
610
 retry:
611
  do
612
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
Karoly Lorentey's avatar
Karoly Lorentey committed
613
		     NUMBERP (seconds) ? &end_time : NULL);
614
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */
615

616
  if (BUFFERP (val))
617 618
    goto retry;

619
  /* `switch-frame' events are put off until after the next ASCII
Karl Heuer's avatar
Karl Heuer committed
620
     character.  This is better than signaling an error just because
621 622 623 624 625
     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)
626
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
627 628 629 630 631
    {
      delayed_switch_frame = val;
      goto retry;
    }

632
  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
633 634
    {
      /* Convert certain symbols to their ASCII equivalents.  */
635
      if (SYMBOLP (val))
636
	{
637
	  Lisp_Object tem, tem1;
638 639 640 641 642 643 644
	  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))
645
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
646 647
	    }
	}
648

649
      /* If we don't have a character now, deal with it appropriately.  */
650
      if (!INTEGERP (val))
651 652 653
	{
	  if (error_nonascii)
	    {
654
	      Vunread_command_events = list1 (val);
655 656 657 658 659 660 661 662 663 664
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

665 666
#if 0

667
#ifdef HAVE_WINDOW_SYSTEM
668 669
  if (display_hourglass_p)
    start_hourglass ();
670
#endif
671 672 673

#endif

674 675 676
  return val;
}

Paul Eggert's avatar
Paul Eggert committed
677
DEFUN ("read-char", Fread_char, Sread_char, 0, 3, 0,
678 679
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.
680 681 682
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).

683 684
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
685 686
exception, switch-frame events are put off until non-character events
can be read.
687 688 689 690 691 692
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
693 694 695 696 697
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
698
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
699
{
700 701
  Lisp_Object val;

702 703
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
704
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);
705 706 707

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
708 709
}

Paul Eggert's avatar
Paul Eggert committed
710
DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
711 712 713 714
       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
715 716 717 718 719
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
720
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
721
{
722 723
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
724
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
Jim Blandy's avatar
Jim Blandy committed
725 726
}

727
DEFUN ("read-char-exclusive", Fread_char_exclusive, Sread_char_exclusive, 0, 3, 0,
728 729
       doc: /* Read a character from the command input (keyboard or macro).
It is returned as a number.  Non-character events are ignored.
730 731
If the character has modifiers, they are resolved and reflected to the
character code if possible (e.g. C-SPC -> 0).
732 733 734 735

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
736 737 738 739 740
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
741
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
Jim Blandy's avatar
Jim Blandy committed
742
{
743 744
  Lisp_Object val;

745 746
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
747

748
  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);
749 750 751

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
Jim Blandy's avatar
Jim Blandy committed
752 753 754
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
755
       doc: /* Don't use this yourself.  */)
Dan Nicolaescu's avatar
Dan Nicolaescu committed
756
  (void)
Jim Blandy's avatar
Jim Blandy committed
757 758
{
  register Lisp_Object val;
759
  block_input ();
760
  XSETINT (val, getc (instream));
761
  unblock_input ();
Jim Blandy's avatar
Jim Blandy committed
762 763
  return val;
}
764 765


766 767 768 769


/* Return true if the lisp code read using READCHARFUN defines a non-nil
   `lexical-binding' file variable.  After returning, the stream is
770 771
   positioned following the first line, if it is a comment or #! line,
   otherwise nothing is read.  */
772

773
static bool
Stefan Monnier's avatar
Stefan Monnier committed
774
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
775 776
{
  int ch = READCHAR;
777 778 779 780 781 782 783 784 785 786 787 788 789

  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
790 791
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
792 793
    }

794 795 796 797 798 799 800 801 802
  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.  */
    {
803
      bool rv = 0;
804
      enum {
Paul Eggert's avatar
Paul Eggert committed
805
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
806
      } beg_end_state = NOMINAL;
807
      bool in_file_vars = 0;
808

809 810 811 812 813 814 815 816 817 818
#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;					\
819 820 821 822 823 824 825 826 827 828 829 830
    }

      /* 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)
	{
831
	  char var[100], val[100];
832
	  unsigned i;
833 834 835 836 837 838 839

	  ch = READCHAR;

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

840
	  i = 0;
841
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
842
	    {
843 844
	      if (i < sizeof var - 1)
		var[i++] = ch;
845 846 847
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }
848

849
	  /* Stop scanning if no colon was found before end marker.  */
850
	  if (!in_file_vars || ch == '\n' || ch == EOF)
851 852
	    break;

853 854 855
	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';
856 857 858 859 860 861 862 863 864

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

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

865
	      i = 0;
866 867
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
868 869
		  if (i < sizeof val - 1)
		    val[i++] = ch;
870 871 872 873
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
874
		/* The value was terminated by an end-marker, which remove.  */
875 876 877 878
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';
879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894

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

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

      return rv;
    }
}
895

Kenichi Handa's avatar
Kenichi Handa committed
896
/* Value is a version number of byte compiled code if the file
Miles Bader's avatar
Miles Bader committed
897
   associated with file descriptor FD is a compiled Lisp file that's
Kenichi Handa's avatar
Kenichi Handa committed
898 899 900
   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.  */
901 902

static int
903
safe_to_load_version (int fd)
904 905 906
{
  char buf[512];
  int nbytes, i;
Kenichi Handa's avatar
Kenichi Handa committed
907
  int version = 1;
908 909 910

  /* 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
911
  nbytes = emacs_read_quit (fd, buf, sizeof buf);
912 913 914
  if (nbytes > 0)
    {
      /* Skip to the next newline, skipping over the initial `ELC'
Dave Love's avatar
Dave Love committed
915
	 with NUL bytes following it, but note the version.  */
916
      for (i = 0; i < nbytes && buf[i] != '\n'; ++i)
Dave Love's avatar
Dave Love committed
917
	if (i == 4)
Kenichi Handa's avatar
Kenichi Handa committed
918
	  version = buf[i];
919

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