safescanner.c

bonddb 是一个源于PostgreSQL封装包的对象。它是一个由C/C++编写的快速数据提取层应用软件

  config = scanner->config;
  (*value_p).v_int = 0;
  
  if ((scanner->text >= scanner->text_end && scanner->input_fd < 0) ||
      scanner->token == SAFE_TOKEN_EOF)
    {
      *token_p = SAFE_TOKEN_EOF;
      return;
    }
  
  in_comment_multi = FALSE;
  in_comment_single = FALSE;
  in_string_sq = FALSE;
  in_string_dq = FALSE;
  gstring = NULL;
  i_saw_escape_char = FALSE;
  
  do /* while (ch != 0) */
    {
      gboolean dotted_float = FALSE;
      
      ch = safe_scanner_get_char (scanner, line_p, position_p);
      
      value.v_int = 0;
      token = SAFE_TOKEN_NONE;
      
      /* this is *evil*, but needed ;(
       * we first check for identifier first character, because	 it
       * might interfere with other key chars like slashes or numbers
       */
      if (config->scan_identifier &&
	  ch && strchr (config->cset_identifier_first, ch))
	goto identifier_precedence;
      
      switch (ch)
	{
	case 0:
	  token = SAFE_TOKEN_EOF;
	  (*position_p)++;
	  /* ch = 0; */
	  break;
	  
	case '/':
	  if (!config->scan_comment_multi ||
	      safe_scanner_peek_next_char (scanner) != '*')
	    goto default_case;
	  safe_scanner_get_char (scanner, line_p, position_p);
	  token = SAFE_TOKEN_COMMENT_MULTI;
	  in_comment_multi = TRUE;
	  gstring = g_string_new ("");
	  while ((ch = safe_scanner_get_char (scanner, line_p, position_p)) != 0)
	    {
	      if (ch == '*' && safe_scanner_peek_next_char (scanner) == '/')
		{
		  safe_scanner_get_char (scanner, line_p, position_p);
		  in_comment_multi = FALSE;
		  break;
		}
	      else
		gstring = g_string_append_c (gstring, ch);
	    }
	  ch = 0;
	  break;
	  
	case '\'':
/*printf("I saw sq.\n");*/
	  if (!config->scan_string_sq)
	    goto default_case;

	  token = SAFE_TOKEN_STRING;
	  in_string_sq = TRUE;
	  gstring = g_string_new ("");
	  while ((ch = safe_scanner_get_char (scanner, line_p, position_p)) != 0)
	    {
/*printf("%c\n", ch);*/
	      if (ch == '\'')
		{
		  in_string_sq = FALSE;
/*printf("Ending...\n");*/
		  i_saw_escape_char = FALSE;
		  break;
		}
	      else if (ch == '\\')
		    {
		      ch = safe_scanner_get_char (scanner, line_p, position_p);
		      switch (ch)
			{
			  guint	i;
			  guint	fchar;
			  
			case 0:
			  break;

			case '\\':
			  gstring = g_string_append_c (gstring, '\\');
/*printf("I see '\\'\n");*/
			  i_saw_escape_char = TRUE;
			  break;
			  
			case 'n':
			  gstring = g_string_append_c (gstring, '\n');
			  break;
			  
			case 't':
			  gstring = g_string_append_c (gstring, '\t');
			  break;
			  
			case 'r':
			  gstring = g_string_append_c (gstring, '\r');
			  break;
			  
			case 'b':
			  gstring = g_string_append_c (gstring, '\b');
			  break;
			  
			case 'f':
			  gstring = g_string_append_c (gstring, '\f');
			  break;
			  
			case '0':
			case '1':
			case '2':
			case '3':
			case '4':
			case '5':
			case '6':
			case '7':
			  i = ch - '0';
			  fchar = safe_scanner_peek_next_char (scanner);
			  if (fchar >= '0' && fchar <= '7')
			    {
			      ch = safe_scanner_get_char (scanner, line_p, position_p);
			      i = i * 8 + ch - '0';
			      fchar = safe_scanner_peek_next_char (scanner);
			      if (fchar >= '0' && fchar <= '7')
				{
				  ch = safe_scanner_get_char (scanner, line_p, position_p);
				  i = i * 8 + ch - '0';
				}
			    }
			  gstring = g_string_append_c (gstring, i);
			  break;
			  
			default:
			  gstring = g_string_append_c (gstring, ch);
			  break;
			}
		    }
		  else
			  gstring = g_string_append_c (gstring, ch);
	    }
	  ch = 0;
	  break;
	  
	case '"':
	  if (!config->scan_string_dq)
	    goto default_case;
	  token = SAFE_TOKEN_STRING;
	  in_string_dq = TRUE;
	  gstring = g_string_new ("");
	  while ((ch = safe_scanner_get_char (scanner, line_p, position_p)) != 0)
	    {
	      if (ch == '"')
		{
		  in_string_dq = FALSE;
		  break;
		}
	      else
		{
		  if (ch == '\\')
		    {
		      ch = safe_scanner_get_char (scanner, line_p, position_p);
		      switch (ch)
			{
			  guint	i;
			  guint	fchar;
			  
			case 0:
			  break;
			  
			case '\\':
			  gstring = g_string_append_c (gstring, '\\');
			  break;
			  
			case 'n':
			  gstring = g_string_append_c (gstring, '\n');
			  break;
			  
			case 't':
			  gstring = g_string_append_c (gstring, '\t');
			  break;
			  
			case 'r':
			  gstring = g_string_append_c (gstring, '\r');
			  break;
			  
			case 'b':
			  gstring = g_string_append_c (gstring, '\b');
			  break;
			  
			case 'f':
			  gstring = g_string_append_c (gstring, '\f');
			  break;
			  
			case '0':
			case '1':
			case '2':
			case '3':
			case '4':
			case '5':
			case '6':
			case '7':
			  i = ch - '0';
			  fchar = safe_scanner_peek_next_char (scanner);
			  if (fchar >= '0' && fchar <= '7')
			    {
			      ch = safe_scanner_get_char (scanner, line_p, position_p);
			      i = i * 8 + ch - '0';
			      fchar = safe_scanner_peek_next_char (scanner);
			      if (fchar >= '0' && fchar <= '7')
				{
				  ch = safe_scanner_get_char (scanner, line_p, position_p);
				  i = i * 8 + ch - '0';
				}
			    }
			  gstring = g_string_append_c (gstring, i);
			  break;
			  
			default:
			  gstring = g_string_append_c (gstring, ch);
			  break;
			}
		    }
		  else
		    gstring = g_string_append_c (gstring, ch);
		}
	    }
	  ch = 0;
	  break;
	  
	case '.':
	  if (!config->scan_float)
	    goto default_case;
	  token = SAFE_TOKEN_FLOAT;
	  dotted_float = TRUE;
	  ch = safe_scanner_get_char (scanner, line_p, position_p);
	  goto number_parsing;
	  
	case '$':
	  if (!config->scan_hex_dollar)
	    goto default_case;
	  token = SAFE_TOKEN_HEX;
	  ch = safe_scanner_get_char (scanner, line_p, position_p);
	  goto number_parsing;
	  
	case '0':
	  if (config->scan_octal)
	    token = SAFE_TOKEN_OCTAL;
	  else
	    token = SAFE_TOKEN_INT;
	  ch = safe_scanner_peek_next_char (scanner);
	  if (config->scan_hex && (ch == 'x' || ch == 'X'))
	    {
	      token = SAFE_TOKEN_HEX;
	      safe_scanner_get_char (scanner, line_p, position_p);
	      ch = safe_scanner_get_char (scanner, line_p, position_p);
	      if (ch == 0)
		{
		  token = SAFE_TOKEN_ERROR;
		  value.v_error = G_ERR_UNEXP_EOF;
		  (*position_p)++;
		  break;
		}
	      if (safe_scanner_char_2_num (ch, 16) < 0)
		{
		  token = SAFE_TOKEN_ERROR;
		  value.v_error = G_ERR_DIGIT_RADIX;
		  ch = 0;
		  break;
		}
	    }
	  else if (config->scan_binary && (ch == 'b' || ch == 'B'))
	    {
	      token = SAFE_TOKEN_BINARY;
	      safe_scanner_get_char (scanner, line_p, position_p);
	      ch = safe_scanner_get_char (scanner, line_p, position_p);
	      if (ch == 0)
		{
		  token = SAFE_TOKEN_ERROR;
		  value.v_error = G_ERR_UNEXP_EOF;
		  (*position_p)++;
		  break;
		}
	      if (safe_scanner_char_2_num (ch, 10) < 0)
		{
		  token = SAFE_TOKEN_ERROR;
		  value.v_error = G_ERR_NON_DIGIT_IN_CONST;
		  ch = 0;
		  break;
		}
	    }
	  else
	    ch = '0';
	  /* fall through */
	case '1':
	case '2':
	case '3':
	case '4':
	case '5':
	case '6':
	case '7':
	case '8':
	case '9':
	number_parsing:
	{
          gboolean in_number = TRUE;
	  gchar *endptr;
	  
	  if (token == SAFE_TOKEN_NONE)
	    token = SAFE_TOKEN_INT;
	  
	  gstring = g_string_new (dotted_float ? "0." : "");
	  gstring = g_string_append_c (gstring, ch);
	  
	  do /* while (in_number) */
	    {
	      gboolean is_E;
	      
	      is_E = token == SAFE_TOKEN_FLOAT && (ch == 'e' || ch == 'E');
	      
	      ch = safe_scanner_peek_next_char (scanner);
	      
	      if (safe_scanner_char_2_num (ch, 36) >= 0 ||
		  (config->scan_float && ch == '.') ||
		  (is_E && (ch == '+' || ch == '-')))
		{
		  ch = safe_scanner_get_char (scanner, line_p, position_p);
		  
		  switch (ch)
		    {
		    case '.':
		      if (token != SAFE_TOKEN_INT && token != SAFE_TOKEN_OCTAL)
			{
			  value.v_error = token == SAFE_TOKEN_FLOAT ? G_ERR_FLOAT_MALFORMED : G_ERR_FLOAT_RADIX;
			  token = SAFE_TOKEN_ERROR;
			  in_number = FALSE;
			}
		      else
			{
			  token = SAFE_TOKEN_FLOAT;
			  gstring = g_string_append_c (gstring, ch);
			}
		      break;
		      
		    case '0':
		    case '1':
		    case '2':
		    case '3':
		    case '4':
		    case '5':
		    case '6':
		    case '7':
		    case '8':
		    case '9':
		      gstring = g_string_append_c (gstring, ch);
		      break;
		      
		    case '-':
		    case '+':
		      if (token != SAFE_TOKEN_FLOAT)
			{
			  token = SAFE_TOKEN_ERROR;
			  value.v_error = G_ERR_NON_DIGIT_IN_CONST;
			  in_number = FALSE;
			}
		      else
			gstring = g_string_append_c (gstring, ch);
		      break;
		      
		    case 'e':
		    case 'E':
		      if ((token != SAFE_TOKEN_HEX && !config->scan_float) ||
			  (token != SAFE_TOKEN_HEX &&
			   token != SAFE_TOKEN_OCTAL &&
			   token != SAFE_TOKEN_FLOAT &&
			   token != SAFE_TOKEN_INT))
			{
			  token = SAFE_TOKEN_ERROR;
			  value.v_error = G_ERR_NON_DIGIT_IN_CONST;
			  in_number = FALSE;
			}
		      else
			{
			  if (token != SAFE_TOKEN_HEX)
			    token = SAFE_TOKEN_FLOAT;
			  gstring = g_string_append_c (gstring, ch);
			}
		      break;
		      
		    default:
		      if (token != SAFE_TOKEN_HEX)
			{
			  token = SAFE_TOKEN_ERROR;
			  value.v_error = G_ERR_NON_DIGIT_IN_CONST;
			  in_number = FALSE;
			}
		      else
			gstring = g_string_append_c (gstring, ch);
		      break;
		    }
		}
	      else
		in_number = FALSE;
	    }
	  while (in_number);
	  
	  endptr = NULL;
	  switch (token)
	    {
	    case SAFE_TOKEN_BINARY:
	      value.v_binary = strtol (gstring->str, &endptr, 2);
	      break;
	      
	    case SAFE_TOKEN_OCTAL:
	      value.v_octal = strtol (gstring->str, &endptr, 8);
	      break;
	      
	    case SAFE_TOKEN_INT:
	      value.v_int = strtol (gstring->str, &endptr, 10);
	      break;
	      
	    case SAFE_TOKEN_FLOAT:
	      value.v_float = g_strtod (gstring->str, &endptr);
	      break;
	      
	    case SAFE_TOKEN_HEX:
	      value.v_hex = strtol (gstring->str, &endptr, 16);
	      break;
	      
	    default:
	      break;
	    }
	  if (endptr && *endptr)
	    {
	      token = SAFE_TOKEN_ERROR;
	      if (*endptr == 'e' || *endptr == 'E')
		value.v_error = G_ERR_NON_DIGIT_IN_CONST;
	      else
		value.v_error = G_ERR_DIGIT_RADIX;
	    }
	  g_string_free (gstring, TRUE);
	  gstring = NULL;
	  ch = 0;
	} /* number_parsing:... */
	break;
	
	default:
	default_case:
	{
	  if (config->cpair_comment_single &&
	      ch == config->cpair_comment_single[0])
	    {
	      token = SAFE_TOKEN_COMMENT_SINGLE;
	      in_comment_single = TRUE;
	      gstring = g_string_new ("");
	      ch = safe_scanner_get_char (scanner, line_p, position_p);
	      while (ch != 0)
		{
		  if (ch == config->cpair_comment_single[1])
		    {
		      in_comment_single = FALSE;
		      ch = 0;
		      break;
		    }
		  
		  gstring = g_string_append_c (gstring, ch);
		  ch = safe_scanner_get_char (scanner, line_p, position_p);
		}
	    }
	  else if (config->scan_identifier && ch &&
		   strchr (config->cset_identifier_first, ch))
	    {
	    identifier_precedence:
	      
	      if (config->cset_identifier_nth && ch &&
		  strchr (config->cset_identifier_nth,
			  safe_scanner_peek_next_char (scanner)))
		{
		  token = SAFE_TOKEN_IDENTIFIER;
		  gstring = g_string_new ("");
		  gstring = g_string_append_c (gstring, ch);
		  do
		    {
		      ch = safe_scanner_get_char (scanner, line_p, position_p);
		      gstring = g_string_append_c (gstring, ch);
		      ch = safe_scanner_peek_next_char (scanner);
		    }
		  while (ch && strchr (config->cset_identifier_nth, ch));
		  ch = 0;
		}
	      else if (config->scan_identifier_1char)
		{
		  token = SAFE_TOKEN_IDENTIFIER;
		  value.v_identifier = g_new0 (gchar, 2);
		  value.v_identifier[0] = ch;
		  ch = 0;
		}
	    }
	  if (ch)
	    {
	      if (config->char_2_token)
		token = ch;
	      else
		{
		  token = SAFE_TOKEN_CHAR;
		  value.v_char = ch;
		}
	      ch = 0;
	    }
	} /* default_case:... */
	break;
	}
      g_assert (ch == 0 && token != SAFE_TOKEN_NONE); /* paranoid */
    }
  while (ch != 0);
  
  if (in_comment_multi || in_comment_single ||
      in_string_sq || in_string_dq)
    {
printf("Still error!\n");
      token = SAFE_TOKEN_ERROR;
      if (gstring)
	{
	  g_string_free (gstring, TRUE);
	  gstring = NULL;
	}
      (*position_p)++;
      if (in_comment_multi || in_comment_single)
	value.v_error = G_ERR_UNEXP_EOF_IN_COMMENT;
      else /* (in_string_sq || in_string_dq) */
	value.v_error = G_ERR_UNEXP_EOF_IN_STRING;
    }
  
  if (gstring)
    {
/*printf("hey: %s\n", gstring->str);*/
      value.v_string = gstring->str;
      g_string_free (gstring, FALSE);
      gstring = NULL;
    }
  
  if (token == SAFE_TOKEN_IDENTIFIER)
    {
      if (config->scan_symbols)
	{
	  SafeScannerKey *key;
	  guint scope_id;
	  
	  scope_id = scanner->scope_id;
	  key = safe_scanner_lookup_internal (scanner, scope_id, value.v_identifier);
	  if (!key && scope_id && scanner->config->scope_0_fallback)
	    key = safe_scanner_lookup_internal (scanner, 0, value.v_identifier);
	  
	  if (key)
	    {
	      g_free (value.v_identifier);
	      token = SAFE_TOKEN_SYMBOL;
	      value.v_symbol = key->value;
	    }
	}
      
      if (token == SAFE_TOKEN_IDENTIFIER &&
	  config->scan_identifier_NULL &&
	  strlen (value.v_identifier) == 4)
	{
	  gchar *null_upper = "NULL";
	  gchar *null_lower = "null";
	  
	  if (scanner->config->case_sensitive)
	    {
	      if (value.v_identifier[0] == null_upper[0] &&
		  value.v_identifier[1] == null_upper[1] &&
		  value.v_identifier[2] == null_upper[2] &&
		  value.v_identifier[3] == null_upper[3])
		token = SAFE_TOKEN_IDENTIFIER_NULL;
	    }
	  else
	    {
	      if ((value.v_identifier[0] == null_upper[0] ||
		   value.v_identifier[0] == null_lower[0]) &&
		  (value.v_identifier[1] == null_upper[1] ||
		   value.v_identifier[1] == null_lower[1]) &&
		  (value.v_identifier[2] == null_upper[2] ||
		   value.v_identifier[2] == null_lower[2]) &&
		  (value.v_identifier[3] == null_upper[3] ||
		   value.v_identifier[3] == null_lower[3]))
		token = SAFE_TOKEN_IDENTIFIER_NULL;
	    }
	}
    }
  
  *token_p = token;
  *value_p = value;
}