lread.c

/* Lisp parsing and input streams.

Copyright (C) 1985-1989, 1993-1995, 1997-2017 Free Software Foundation,
Inc.

This file is part of GNU Emacs.

GNU Emacs is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or (at
your option) any later version.

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
along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */

/* Tell globals.h to define tables needed by init_obarray.  */
#define DEFINE_SYMBOLS

#include <config.h>
#include "sysstdio.h"
#include <stdlib.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/file.h>
#include <errno.h>
#include <math.h>
#include <stat-time.h>
#include "lisp.h"
#include "dispextern.h"
#include "intervals.h"
#include "character.h"
#include "buffer.h"
#include "charset.h"
#include <epaths.h>
#include "commands.h"
#include "keyboard.h"
#include "systime.h"
#include "termhooks.h"
#include "blockinput.h"
#include <c-ctype.h>

#ifdef MSDOS
#include "msdos.h"
#if __DJGPP__ == 2 && __DJGPP_MINOR__ < 5
# define INFINITY  __builtin_inf()
# define NAN       __builtin_nan("")
#endif
#endif

#ifdef HAVE_NS
#include "nsterm.h"
#endif

#include <unistd.h>

#ifdef HAVE_SETLOCALE
#include <locale.h>
#endif /* HAVE_SETLOCALE */

#include <fcntl.h>

#ifdef HAVE_FSEEKO
#define file_offset off_t
#define file_tell ftello
#else
#define file_offset long
#define file_tell ftell
#endif

#ifndef HAVE_GETC_UNLOCKED
#define getc_unlocked getc
#endif

/* The objects or placeholders read with the #n=object form.

   A hash table maps a number to either a placeholder (while the
   object is still being parsed, in case it's referenced within its
   own definition) or to the completed object.  With small integers
   for keys, it's effectively little more than a vector, but it'll
   manage any needed resizing for us.

   The variable must be reset to an empty hash table before all
   top-level calls to read0.  In between calls, it may be an empty
   hash table left unused from the previous call (to reduce
   allocations), or nil.  */
static Lisp_Object read_objects_map;

/* The recursive objects read with the #n=object form.

   Objects that might have circular references are stored here, so
   that recursive substitution knows not to keep processing them
   multiple times.

   Only objects that are completely processed, including substituting
   references to themselves (but not necessarily replacing
   placeholders for other objects still being read), are stored.

   A hash table is used for efficient lookups of keys.  We don't care
   what the value slots hold.  The variable must be set to an empty
   hash table before all top-level calls to read0.  In between calls,
   it may be an empty hash table left unused from the previous call
   (to reduce allocations), or nil.  */
static Lisp_Object read_objects_completed;

/* File for get_file_char to read from.  Use by load.  */
static FILE *instream;

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

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

/* This contains the last string skipped with #@.  */
static char *saved_doc_string;
/* Length of buffer allocated in saved_doc_string.  */
static ptrdiff_t saved_doc_string_size;
/* Length of actual data in saved_doc_string.  */
static ptrdiff_t saved_doc_string_length;
/* This is the file position that string came from.  */
static file_offset saved_doc_string_position;

/* 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.  */
static ptrdiff_t prev_saved_doc_string_size;
/* Length of actual data in prev_saved_doc_string.  */
static ptrdiff_t prev_saved_doc_string_length;
/* This is the file position that string came from.  */
static file_offset prev_saved_doc_string_position;

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

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

static Lisp_Object Vloads_in_progress;

static int read_emacs_mule_char (int, int (*) (int, Lisp_Object),
                                 Lisp_Object);

static void readevalloop (Lisp_Object, FILE *, Lisp_Object, bool,
                          Lisp_Object, Lisp_Object,
                          Lisp_Object, Lisp_Object);

/* 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.  */

static int readbyte_for_lambda (int, Lisp_Object);
static int readbyte_from_file (int, Lisp_Object);
static int readbyte_from_string (int, Lisp_Object);

/* Handle unreading and rereading of characters.
   Write READCHAR to read a character,
   UNREAD(c) to unread c to be read again.

   These macros correctly read/unread multibyte characters.  */

#define READCHAR readchar (readcharfun, NULL)
#define UNREAD(c) unreadchar (readcharfun, c)

/* Same as READCHAR but set *MULTIBYTE to the multibyteness of the source.  */
#define READCHAR_REPORT_MULTIBYTE(multibyte) readchar (readcharfun, multibyte)

/* When READCHARFUN is Qget_file_char, Qget_emacs_mule_file_char,
   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
   means that there's no unread character.  */
static int unread_char;

static int
readchar (Lisp_Object readcharfun, bool *multibyte)
{
  Lisp_Object tem;
  register int c;
  int (*readbyte) (int, Lisp_Object);
  unsigned char buf[MAX_MULTIBYTE_LENGTH];
  int i, len;
  bool emacs_mule_encoding = 0;

  if (multibyte)
    *multibyte = 0;

  readchar_count++;

  if (BUFFERP (readcharfun))
    {
      register struct buffer *inbuffer = XBUFFER (readcharfun);

      ptrdiff_t pt_byte = BUF_PT_BYTE (inbuffer);

      if (! BUFFER_LIVE_P (inbuffer))
	return -1;

      if (pt_byte >= BUF_ZV_BYTE (inbuffer))
	return -1;

      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
	{
	  /* Fetch the character code from the buffer.  */
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, pt_byte);
	  BUF_INC_POS (inbuffer, pt_byte);
	  c = STRING_CHAR (p);
	  if (multibyte)
	    *multibyte = 1;
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, pt_byte);
	  if (! ASCII_CHAR_P (c))
	    c = BYTE8_TO_CHAR (c);
	  pt_byte++;
	}
      SET_BUF_PT_BOTH (inbuffer, BUF_PT (inbuffer) + 1, pt_byte);

      return c;
    }
  if (MARKERP (readcharfun))
    {
      register struct buffer *inbuffer = XMARKER (readcharfun)->buffer;

      ptrdiff_t bytepos = marker_byte_position (readcharfun);

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

      if (! NILP (BVAR (inbuffer, enable_multibyte_characters)))
	{
	  /* Fetch the character code from the buffer.  */
	  unsigned char *p = BUF_BYTE_ADDRESS (inbuffer, bytepos);
	  BUF_INC_POS (inbuffer, bytepos);
	  c = STRING_CHAR (p);
	  if (multibyte)
	    *multibyte = 1;
	}
      else
	{
	  c = BUF_FETCH_BYTE (inbuffer, bytepos);
	  if (! ASCII_CHAR_P (c))
	    c = BYTE8_TO_CHAR (c);
	  bytepos++;
	}

      XMARKER (readcharfun)->bytepos = bytepos;
      XMARKER (readcharfun)->charpos++;

      return c;
    }

  if (EQ (readcharfun, Qlambda))
    {
      readbyte = readbyte_for_lambda;
      goto read_multibyte;
    }

  if (EQ (readcharfun, Qget_file_char))
    {
      readbyte = readbyte_from_file;
      goto read_multibyte;
    }

  if (STRINGP (readcharfun))
    {
      if (read_from_string_index >= read_from_string_limit)
	c = -1;
      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);
	}
      else
	{
	  c = SREF (readcharfun, read_from_string_index_byte);
	  read_from_string_index++;
	  read_from_string_index_byte++;
	}
      return c;
    }

  if (CONSP (readcharfun) && STRINGP (XCAR (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;
    }

  if (EQ (readcharfun, Qget_emacs_mule_file_char))
    {
      readbyte = readbyte_from_file;
      emacs_mule_encoding = 1;
      goto read_multibyte;
    }

  tem = call0 (readcharfun);

  if (NILP (tem))
    return -1;
  return XINT (tem);

 read_multibyte:
  if (unread_char >= 0)
    {
      c = unread_char;
      unread_char = -1;
      return c;
    }
  c = (*readbyte) (-1, readcharfun);
  if (c < 0)
    return c;
  if (multibyte)
    *multibyte = 1;
  if (ASCII_CHAR_P (c))
    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;
    }
  return STRING_CHAR (buf);
}

#define FROM_FILE_P(readcharfun)			\
  (EQ (readcharfun, Qget_file_char)			\
   || EQ (readcharfun, Qget_emacs_mule_file_char))

static void
skip_dyn_bytes (Lisp_Object readcharfun, ptrdiff_t n)
{
  if (FROM_FILE_P (readcharfun))
    {
      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');
    }
}

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);
}

/* 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
unreadchar (Lisp_Object readcharfun, int c)
{
  readchar_count--;
  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.  */
    ;
  else if (BUFFERP (readcharfun))
    {
      struct buffer *b = XBUFFER (readcharfun);
      ptrdiff_t charpos = BUF_PT (b);
      ptrdiff_t bytepos = BUF_PT_BYTE (b);

      if (! NILP (BVAR (b, enable_multibyte_characters)))
	BUF_DEC_POS (b, bytepos);
      else
	bytepos--;

      SET_BUF_PT_BOTH (b, charpos - 1, bytepos);
    }
  else if (MARKERP (readcharfun))
    {
      struct buffer *b = XMARKER (readcharfun)->buffer;
      ptrdiff_t bytepos = XMARKER (readcharfun)->bytepos;

      XMARKER (readcharfun)->charpos--;
      if (! NILP (BVAR (b, enable_multibyte_characters)))
	BUF_DEC_POS (b, bytepos);
      else
	bytepos--;

      XMARKER (readcharfun)->bytepos = bytepos;
    }
  else if (STRINGP (readcharfun))
    {
      read_from_string_index--;
      read_from_string_index_byte
	= string_char_to_byte (readcharfun, read_from_string_index);
    }
  else if (CONSP (readcharfun) && STRINGP (XCAR (readcharfun)))
    {
      unread_char = c;
    }
  else if (EQ (readcharfun, Qlambda))
    {
      unread_char = c;
    }
  else if (FROM_FILE_P (readcharfun))
    {
      unread_char = c;
    }
  else
    call1 (readcharfun, make_number (c));
}

static int
readbyte_for_lambda (int c, Lisp_Object readcharfun)
{
  return read_bytecode_char (c >= 0);
}


static int
readbyte_from_file (int c, Lisp_Object readcharfun)
{
  if (c >= 0)
    {
      block_input ();
      ungetc (c, instream);
      unblock_input ();
      return 0;
    }

  block_input ();
  c = getc_unlocked (instream);

  /* Interrupted reads have been observed while reading over the network.  */
  while (c == EOF && ferror (instream) && errno == EINTR)
    {
      unblock_input ();
      maybe_quit ();
      block_input ();
      clearerr (instream);
      c = getc_unlocked (instream);
    }

  unblock_input ();

  return (c == EOF ? -1 : c);
}

static int
readbyte_from_string (int c, Lisp_Object readcharfun)
{
  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);
    }

  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
read_emacs_mule_char (int c, int (*readbyte) (int, Lisp_Object), Lisp_Object readcharfun)
{
  /* 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)
    {
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
      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)
	{
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
	  code = buf[2] & 0x7F;
	}
      else
	{
	  charset = CHARSET_FROM_ID (emacs_mule_charset[buf[0]]);
	  code = ((buf[1] << 8) | buf[2]) & 0x7F7F;
	}
    }
  else
    {
      charset = CHARSET_FROM_ID (emacs_mule_charset[buf[1]]);
      code = ((buf[2] << 8) | buf[3]) & 0x7F7F;
    }
  c = DECODE_CHAR (charset, code);
  if (c < 0)
    Fsignal (Qinvalid_read_syntax,
	     list1 (build_string ("invalid multibyte form")));
  return c;
}


static Lisp_Object read_internal_start (Lisp_Object, Lisp_Object,
                                        Lisp_Object);
static Lisp_Object read0 (Lisp_Object);
static Lisp_Object read1 (Lisp_Object, int *, bool);

static Lisp_Object read_list (bool, Lisp_Object);
static Lisp_Object read_vector (Lisp_Object, bool);

static Lisp_Object substitute_object_recurse (Lisp_Object, Lisp_Object,
                                              Lisp_Object);
static void substitute_in_interval (INTERVAL, Lisp_Object);


/* Get a character from the tty.  */

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

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

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

   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
   character.

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

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

static Lisp_Object
read_filtered_event (bool no_switch_frame, bool ascii_required,
		     bool error_nonascii, bool input_method, Lisp_Object seconds)
{
  Lisp_Object val, delayed_switch_frame;
  struct timespec end_time;

#ifdef HAVE_WINDOW_SYSTEM
  if (display_hourglass_p)
    cancel_hourglass ();
#endif

  delayed_switch_frame = Qnil;

  /* Compute timeout.  */
  if (NUMBERP (seconds))
    {
      double duration = XFLOATINT (seconds);
      struct timespec wait_time = dtotimespec (duration);
      end_time = timespec_add (current_timespec (), wait_time);
    }

  /* Read until we get an acceptable event.  */
 retry:
  do
    val = read_char (0, Qnil, (input_method ? Qnil : Qt), 0,
		     NUMBERP (seconds) ? &end_time : NULL);
  while (INTEGERP (val) && XINT (val) == -2); /* wrong_kboard_jmpbuf */

  if (BUFFERP (val))
    goto retry;

  /* `switch-frame' events are put off until after the next ASCII
     character.  This is better than signaling an error just because
     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)
      && EQ (EVENT_HEAD_KIND (EVENT_HEAD (val)), Qswitch_frame))
    {
      delayed_switch_frame = val;
      goto retry;
    }

  if (ascii_required && !(NUMBERP (seconds) && NILP (val)))
    {
      /* Convert certain symbols to their ASCII equivalents.  */
      if (SYMBOLP (val))
	{
	  Lisp_Object tem, tem1;
	  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))
		XSETFASTINT (val, XINT (tem1) | XINT (Fcar (Fcdr (tem))));
	    }
	}

      /* If we don't have a character now, deal with it appropriately.  */
      if (!INTEGERP (val))
	{
	  if (error_nonascii)
	    {
	      Vunread_command_events = list1 (val);
	      error ("Non-character input-event");
	    }
	  else
	    goto retry;
	}
    }

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

#if 0

#ifdef HAVE_WINDOW_SYSTEM
  if (display_hourglass_p)
    start_hourglass ();
#endif

#endif

  return val;
}

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

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
exception, switch-frame events are put off until non-character events
can be read.
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
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.  */)
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
{
  Lisp_Object val;

  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
  val = read_filtered_event (1, 1, 1, ! NILP (inherit_input_method), seconds);

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
}

DEFUN ("read-event", Fread_event, Sread_event, 0, 3, 0,
       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
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.  */)
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
{
  if (! NILP (prompt))
    message_with_string ("%s", prompt, 0);
  return read_filtered_event (0, 0, 0, ! NILP (inherit_input_method), seconds);
}

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

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
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.  */)
  (Lisp_Object prompt, Lisp_Object inherit_input_method, Lisp_Object seconds)
{
  Lisp_Object val;

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

  val = read_filtered_event (1, 1, 0, ! NILP (inherit_input_method), seconds);

  return (NILP (val) ? Qnil
	  : make_number (char_resolve_modifier_mask (XINT (val))));
}

DEFUN ("get-file-char", Fget_file_char, Sget_file_char, 0, 0, 0,
       doc: /* Don't use this yourself.  */)
  (void)
{
  register Lisp_Object val;
  block_input ();
  XSETINT (val, getc_unlocked (instream));
  unblock_input ();
  return val;
}




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

static bool
lisp_file_lexically_bound_p (Lisp_Object readcharfun)
{
  int ch = READCHAR;

  if (ch == '#')
    {
      ch = READCHAR;
      if (ch != '!')
        {
          UNREAD (ch);
          UNREAD ('#');
          return 0;
        }
      while (ch != '\n' && ch != EOF)
        ch = READCHAR;
      if (ch == '\n') ch = READCHAR;
      /* It is OK to leave the position after a #! line, since
         that is what read1 does.  */
    }

  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.  */
    {
      bool rv = 0;
      enum {
	NOMINAL, AFTER_FIRST_DASH, AFTER_ASTERIX
      } beg_end_state = NOMINAL;
      bool in_file_vars = 0;

#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;					\
    }

      /* 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)
	{
	  char var[100], val[100];
	  unsigned i;

	  ch = READCHAR;

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

	  i = 0;
	  while (ch != ':' && ch != '\n' && ch != EOF && in_file_vars)
	    {
	      if (i < sizeof var - 1)
		var[i++] = ch;
	      UPDATE_BEG_END_STATE (ch);
	      ch = READCHAR;
	    }

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

	  while (i > 0 && (var[i - 1] == ' ' || var[i - 1] == '\t'))
	    i--;
	  var[i] = '\0';

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

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

	      i = 0;
	      while (ch != ';' && ch != '\n' && ch != EOF && in_file_vars)
		{
		  if (i < sizeof val - 1)
		    val[i++] = ch;
		  UPDATE_BEG_END_STATE (ch);
		  ch = READCHAR;
		}
	      if (! in_file_vars)
		/* The value was terminated by an end-marker, which remove.  */
		i -= 3;
	      while (i > 0 && (val[i - 1] == ' ' || val[i - 1] == '\t'))
		i--;
	      val[i] = '\0';

	      if (strcmp (var, "lexical-binding") == 0)
		/* This is it...  */
		{
		  rv =