comparison src/lread.c @ 43080:a6382f0fcb2a

(read1): Redesign strategy for force_multibyte and force_singlebyte. Now is_multibyte records whether read_buffer is multibyte. Encountering any multibyte character makes it so.

43079:a202d9fb56eb

    case '"':
      {
	char *p = read_buffer;
	char *end = read_buffer + read_buffer_size;
	register int c;
	/* Nonzero if we saw an escape sequence specifying
	   a multibyte character.  */
	int force_multibyte = 0;
	/* Nonzero if we saw an escape sequence specifying
	   a single-byte character.  */
	int force_singlebyte = 0;
	int cancel = 0;
	int nchars;

	while ((c = READCHAR) >= 0
	       && c != '\"')
	  {
	    if (end - p < MAX_MULTIBYTE_LENGTH)

		  force_singlebyte = 1;
		else if (byterep == 2)
		  force_multibyte = 1;
	      }

	    if (! SINGLE_BYTE_CHAR_P ((c & ~CHAR_MODIFIER_MASK)))
	      {
		/* Any modifiers for a multibyte character are invalid.  */
		if (c & CHAR_MODIFIER_MASK)
		  error ("Invalid modifier in string");
		p += CHAR_STRING (c, p);
		force_multibyte = 1;
	      }
	    else
	      {
		/* Allow `\C- ' and `\C-?'.  */
		if (c == (CHAR_CTL | ' '))
		  c = 0;
		else if (c == (CHAR_CTL | '?'))
		  c = 127;

		if (c & CHAR_SHIFT)
		  {
		    /* Shift modifier is valid only with [A-Za-z].  */
		    if ((c & 0377) >= 'A' && (c & 0377) <= 'Z')
		      c &= ~CHAR_SHIFT;
		    else if ((c & 0377) >= 'a' && (c & 0377) <= 'z')
		      c = (c & ~CHAR_SHIFT) - ('a' - 'A');
		  }

		if (c & CHAR_META)
		  /* Move the meta bit to the right place for a string.  */
		  c = (c & ~CHAR_META) | 0x80;
		if (c & ~0xff)
		  error ("Invalid modifier in string");
		*p++ = c;
	      }
	  }
	if (c < 0)
	  end_of_file_error ();

	/* If purifying, and string starts with \ newline,
	   return zero instead.  This is for doc strings
	   that we are really going to find in etc/DOC.nn.nn  */
	if (!NILP (Vpurify_flag) && NILP (Vdoc_file_name) && cancel)
	  return make_number (0);

	if (force_multibyte)
	  to_multibyte (&p, &end, &nchars);
	else if (force_singlebyte)
	  nchars = p - read_buffer;
	else if (load_convert_to_unibyte)
	  {
	    Lisp_Object string;
	    to_multibyte (&p, &end, &nchars);
	    if (p - read_buffer != nchars)
	      {
		string = make_multibyte_string (read_buffer, nchars,
						p - read_buffer);
		return Fstring_make_unibyte (string);
	      }
	  }
	else if (EQ (readcharfun, Qget_file_char)
		 || EQ (readcharfun, Qlambda))
	  {
	    /* Nowadays, reading directly from a file is used only for
	       compiled Emacs Lisp files, and those always use the
	       Emacs internal encoding.  Meanwhile, Qlambda is used
	       for reading dynamic byte code (compiled with
	       byte-compile-dynamic = t).  */
	    to_multibyte (&p, &end, &nchars);
	  }
	else
	  /* In all other cases, if we read these bytes as
	     separate characters, treat them as separate characters now.  */
	  nchars = p - read_buffer;

	if (read_pure)
	  return make_pure_string (read_buffer, nchars, p - read_buffer,
				   (force_multibyte
				    || (p - read_buffer != nchars)));
	return make_specified_string (read_buffer, nchars, p - read_buffer,
				      (force_multibyte
				       || (p - read_buffer != nchars)));
      }

    case '.':
      {
	int next_char = READCHAR;

43080:a6382f0fcb2a

    case '"':
      {
	char *p = read_buffer;
	char *end = read_buffer + read_buffer_size;
	register int c;
	/* 1 if we saw an escape sequence specifying
	   a multibyte character, or a multibyte character.  */
	int force_multibyte = 0;
	/* 1 if we saw an escape sequence specifying
	   a single-byte character.  */
	int force_singlebyte = 0;
	/* 1 if read_buffer contains multibyte text now.  */
	int is_multibyte = 0;
	int cancel = 0;
	int nchars = 0;

	while ((c = READCHAR) >= 0
	       && c != '\"')
	  {
	    if (end - p < MAX_MULTIBYTE_LENGTH)

		  force_singlebyte = 1;
		else if (byterep == 2)
		  force_multibyte = 1;
	      }

	    /* A character that must be multibyte forces multibyte.  */
	    if (! SINGLE_BYTE_CHAR_P (c & ~CHAR_MODIFIER_MASK))
	      force_multibyte = 1;

	    /* If we just discovered the need to be multibyte,
	       convert the text accumulated thus far.  */
	    if (force_multibyte && ! is_multibyte)
	      {
		is_multibyte = 1;
		to_multibyte (&p, &end, &nchars);
	      }

	    /* Allow `\C- ' and `\C-?'.  */
	    if (c == (CHAR_CTL | ' '))
	      c = 0;
	    else if (c == (CHAR_CTL | '?'))
	      c = 127;

	    if (c & CHAR_SHIFT)
	      {
		/* Shift modifier is valid only with [A-Za-z].  */
		if ((c & 0377) >= 'A' && (c & 0377) <= 'Z')
		  c &= ~CHAR_SHIFT;
		else if ((c & 0377) >= 'a' && (c & 0377) <= 'z')
		  c = (c & ~CHAR_SHIFT) - ('a' - 'A');
	      }

	    if (c & CHAR_META)
	      /* Move the meta bit to the right place for a string.  */
	      c = (c & ~CHAR_META) | 0x80;
	    if (c & CHAR_MODIFIER_MASK)
	      error ("Invalid modifier in string");

	    if (is_multibyte)
	      p += CHAR_STRING (c, p);
	    else
	      *p++ = c;

	    nchars++;
	  }

	if (c < 0)
	  end_of_file_error ();

	/* If purifying, and string starts with \ newline,
	   return zero instead.  This is for doc strings
	   that we are really going to find in etc/DOC.nn.nn  */
	if (!NILP (Vpurify_flag) && NILP (Vdoc_file_name) && cancel)
	  return make_number (0);

	if (is_multibyte || force_singlebyte)
	  ;
	else if (load_convert_to_unibyte)
	  {
	    Lisp_Object string;
	    to_multibyte (&p, &end, &nchars);
	    if (p - read_buffer != nchars)
	      {
		string = make_multibyte_string (read_buffer, nchars,
						p - read_buffer);
		return Fstring_make_unibyte (string);
	      }
	    /* We can make a unibyte string directly.  */
	    is_multibyte = 0;
	  }
	else if (EQ (readcharfun, Qget_file_char)
		 || EQ (readcharfun, Qlambda))
	  {
	    /* Nowadays, reading directly from a file is used only for
	       compiled Emacs Lisp files, and those always use the
	       Emacs internal encoding.  Meanwhile, Qlambda is used
	       for reading dynamic byte code (compiled with
	       byte-compile-dynamic = t).  */
	    to_multibyte (&p, &end, &nchars);
	    is_multibyte = 1;
	  }
	else
	  /* In all other cases, if we read these bytes as
	     separate characters, treat them as separate characters now.  */
	  ;

	if (read_pure)
	  return make_pure_string (read_buffer, nchars, p - read_buffer,
				   is_multibyte);
	return make_specified_string (read_buffer, nchars, p - read_buffer,
				      is_multibyte);
      }

    case '.':
      {
	int next_char = READCHAR;