scan-gram.l

GNU的词法/语法分析器bison源码

  {splice}	 STRING_GROW;
  <<EOF>>	 BEGIN context_state;
}


  /*------------------------------------------------.
  | Scanning a Bison string, including its escapes. |
  | The initial quote is already eaten.             |
  `------------------------------------------------*/

<SC_ESCAPED_STRING>
{
  "\"" {
    STRING_FINISH;
    loc->start = token_start;
    val->chars = last_string;
    rule_length++;
    BEGIN INITIAL;
    return STRING;
  }
  \n		unexpected_newline (token_start, "\"");	BEGIN INITIAL;
  <<EOF>>	unexpected_eof (token_start, "\"");	BEGIN INITIAL;
}

  /*----------------------------------------------------------.
  | Scanning a Bison character literal, decoding its escapes. |
  | The initial quote is already eaten.			      |
  `----------------------------------------------------------*/

<SC_ESCAPED_CHARACTER>
{
  "'" {
    unsigned char last_string_1;
    STRING_GROW;
    STRING_FINISH;
    loc->start = token_start;
    val->symbol = symbol_get (quotearg_style (escape_quoting_style,
					      last_string),
			      *loc);
    symbol_class_set (val->symbol, token_sym, *loc);
    last_string_1 = last_string[1];
    symbol_user_token_number_set (val->symbol, last_string_1, *loc);
    STRING_FREE;
    rule_length++;
    BEGIN INITIAL;
    return ID;
  }
  \n		unexpected_newline (token_start, "'");	BEGIN INITIAL;
  <<EOF>>	unexpected_eof (token_start, "'");	BEGIN INITIAL;
}

<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING>
{
  \0	    complain_at (*loc, _("invalid null character"));
}


  /*----------------------------.
  | Decode escaped characters.  |
  `----------------------------*/

<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
{
  \\[0-7]{1,3} {
    unsigned long int c = strtoul (yytext + 1, 0, 8);
    if (UCHAR_MAX < c)
      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
    else if (! c)
      complain_at (*loc, _("invalid null character: %s"), quote (yytext));
    else
      obstack_1grow (&obstack_for_string, c);
  }

  \\x[0-9abcdefABCDEF]+ {
    unsigned long int c;
    set_errno (0);
    c = strtoul (yytext + 2, 0, 16);
    if (UCHAR_MAX < c || get_errno ())
      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
    else if (! c)
      complain_at (*loc, _("invalid null character: %s"), quote (yytext));
    else
      obstack_1grow (&obstack_for_string, c);
  }

  \\a	obstack_1grow (&obstack_for_string, '\a');
  \\b	obstack_1grow (&obstack_for_string, '\b');
  \\f	obstack_1grow (&obstack_for_string, '\f');
  \\n	obstack_1grow (&obstack_for_string, '\n');
  \\r	obstack_1grow (&obstack_for_string, '\r');
  \\t	obstack_1grow (&obstack_for_string, '\t');
  \\v	obstack_1grow (&obstack_for_string, '\v');

  /* \\[\"\'?\\] would be shorter, but it confuses xgettext.  */
  \\("\""|"'"|"?"|"\\")  obstack_1grow (&obstack_for_string, yytext[1]);

  \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} {
    int c = convert_ucn_to_byte (yytext);
    if (c < 0)
      complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext));
    else if (! c)
      complain_at (*loc, _("invalid null character: %s"), quote (yytext));
    else
      obstack_1grow (&obstack_for_string, c);
  }
  \\(.|\n)	{
    complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext));
    STRING_GROW;
  }
}

  /*--------------------------------------------.
  | Scanning user-code characters and strings.  |
  `--------------------------------------------*/

<SC_CHARACTER,SC_STRING>
{
  {splice}|\\{splice}[^\n$@\[\]]	STRING_GROW;
}

<SC_CHARACTER>
{
  "'"		STRING_GROW; BEGIN context_state;
  \n		unexpected_newline (token_start, "'"); BEGIN context_state;
  <<EOF>>	unexpected_eof (token_start, "'"); BEGIN context_state;
}

<SC_STRING>
{
  "\""		STRING_GROW; BEGIN context_state;
  \n		unexpected_newline (token_start, "\""); BEGIN context_state;
  <<EOF>>	unexpected_eof (token_start, "\""); BEGIN context_state;
}


  /*---------------------------------------------------.
  | Strings, comments etc. can be found in user code.  |
  `---------------------------------------------------*/

<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
{
  "'" {
    STRING_GROW;
    context_state = YY_START;
    token_start = loc->start;
    BEGIN SC_CHARACTER;
  }
  "\"" {
    STRING_GROW;
    context_state = YY_START;
    token_start = loc->start;
    BEGIN SC_STRING;
  }
  "/"{splice}"*" {
    STRING_GROW;
    context_state = YY_START;
    token_start = loc->start;
    BEGIN SC_COMMENT;
  }
  "/"{splice}"/" {
    STRING_GROW;
    context_state = YY_START;
    BEGIN SC_LINE_COMMENT;
  }
}


  /*---------------------------------------------------------------.
  | Scanning after %union etc., possibly followed by white space.  |
  | For %union only, allow arbitrary C code to appear before the   |
  | following brace, as an extension to POSIX.			   |
  `---------------------------------------------------------------*/

<SC_PRE_CODE>
{
  . {
    bool valid = yytext[0] == '{' || token_type == PERCENT_UNION;
    scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
    yyless (0);

    if (valid)
      {
	braces_level = -1;
	code_start = loc->start;
	BEGIN SC_BRACED_CODE;
      }
    else
      {
	complain_at (*loc, _("missing `{' in %s"),
		     token_name (token_type));
	obstack_sgrow (&obstack_for_string, "{}");
	STRING_FINISH;
	val->chars = last_string;
	BEGIN INITIAL;
	return token_type;
      }
  }

  <<EOF>>  unexpected_eof (scanner_cursor, "{}"); BEGIN INITIAL;
}


  /*---------------------------------------------------------------.
  | Scanning some code in braces (%union and actions). The initial |
  | "{" is already eaten.                                          |
  `---------------------------------------------------------------*/

<SC_BRACED_CODE>
{
  "{"|"<"{splice}"%"  STRING_GROW; braces_level++;
  "%"{splice}">"      STRING_GROW; braces_level--;
  "}" {
    bool outer_brace = --braces_level < 0;

    /* As an undocumented Bison extension, append `;' before the last
       brace in braced code, so that the user code can omit trailing
       `;'.  But do not append `;' if emulating Yacc, since Yacc does
       not append one.

       FIXME: Bison should warn if a semicolon seems to be necessary
       here, and should omit the semicolon if it seems unnecessary
       (e.g., after ';', '{', or '}', each followed by comments or
       white space).  Such a warning shouldn't depend on --yacc; it
       should depend on a new --pedantic option, which would cause
       Bison to warn if it detects an extension to POSIX.  --pedantic
       should also diagnose other Bison extensions like %yacc.
       Perhaps there should also be a GCC-style --pedantic-errors
       option, so that such warnings are diagnosed as errors.  */
    if (outer_brace && token_type == BRACED_CODE && ! yacc_flag)
      obstack_1grow (&obstack_for_string, ';');

    obstack_1grow (&obstack_for_string, '}');

    if (outer_brace)
      {
	STRING_FINISH;
	rule_length++;
	loc->start = code_start;
	val->chars = last_string;
	BEGIN INITIAL;
	return token_type;
      }
  }

  /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
     (as `<' `<%').  */
  "<"{splice}"<"  STRING_GROW;

  "$"("<"{tag}">")?(-?[0-9]+|"$")  handle_dollar (token_type, yytext, *loc);
  "@"(-?[0-9]+|"$")		   handle_at (token_type, yytext, *loc);

  <<EOF>>  unexpected_eof (code_start, "}"); BEGIN INITIAL;
}


  /*--------------------------------------------------------------.
  | Scanning some prologue: from "%{" (already scanned) to "%}".  |
  `--------------------------------------------------------------*/

<SC_PROLOGUE>
{
  "%}" {
    STRING_FINISH;
    loc->start = code_start;
    val->chars = last_string;
    BEGIN INITIAL;
    return PROLOGUE;
  }

  <<EOF>>  unexpected_eof (code_start, "%}"); BEGIN INITIAL;
}


  /*---------------------------------------------------------------.
  | Scanning the epilogue (everything after the second "%%", which |
  | has already been eaten).                                       |
  `---------------------------------------------------------------*/

<SC_EPILOGUE>
{
  <<EOF>> {
    STRING_FINISH;
    loc->start = code_start;
    val->chars = last_string;
    BEGIN INITIAL;
    return EPILOGUE;
  }
}


  /*-----------------------------------------.
  | Escape M4 quoting characters in C code.  |
  `-----------------------------------------*/

<SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
{
  \$	obstack_sgrow (&obstack_for_string, "$][");
  \@	obstack_sgrow (&obstack_for_string, "@@");
  \[	obstack_sgrow (&obstack_for_string, "@{");
  \]	obstack_sgrow (&obstack_for_string, "@}");
}


  /*-----------------------------------------------------.
  | By default, grow the string obstack with the input.  |
  `-----------------------------------------------------*/

<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>.	|
<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n	STRING_GROW;

%%

/* Keeps track of the maximum number of semantic values to the left of
   a handle (those referenced by $0, $-1, etc.) are required by the
   semantic actions of this grammar. */
int max_left_semantic_context = 0;

/* Set *LOC and adjust scanner cursor to account for token TOKEN of
   size SIZE.  */

static void
adjust_location (location *loc, char const *token, size_t size)
{
  int line = scanner_cursor.line;
  int column = scanner_cursor.column;
  char const *p0 = token;
  char const *p = token;
  char const *lim = token + size;

  loc->start = scanner_cursor;

  for (p = token; p < lim; p++)
    switch (*p)
      {
      case '\n':
	line++;
	column = 1;
	p0 = p + 1;
	break;

      case '\t':
	column += mbsnwidth (p0, p - p0, 0);
	column += 8 - ((column - 1) & 7);
	p0 = p + 1;
	break;
      }

  scanner_cursor.line = line;
  scanner_cursor.column = column + mbsnwidth (p0, p - p0, 0);

  loc->end = scanner_cursor;
}


/* Read bytes from FP into buffer BUF of size SIZE.  Return the
   number of bytes read.  Remove '\r' from input, treating \r\n
   and isolated \r as \n.  */

static size_t
no_cr_read (FILE *fp, char *buf, size_t size)
{
  size_t bytes_read = fread (buf, 1, size, fp);
  if (bytes_read)