Java Token.EOF屬性代碼示例

@Override
protected void reportNoViableAlternative(Parser recognizer, NoViableAltException e) {
	// change error message from default implementation
	TokenStream tokens = recognizer.getInputStream();
	String input;
	if (tokens != null) {
		if (e.getStartToken().getType() == Token.EOF) {
			input = "the end";
		} else {
			input = escapeWSAndQuote(tokens.getText(e.getStartToken(), e.getOffendingToken()));
		}
	} else {
		input = escapeWSAndQuote("<unknown input>");
	}
	String msg = "inadmissible input at " + input;
	recognizer.notifyErrorListeners(e.getOffendingToken(), msg, e);
}
 

@Override
protected String getTokenErrorDisplay(Token t) {
	// overwrite standard behavior to use "the end" instead of <EOF>
	if (t == null) {
		return "<no token>";
	}
	String s = getSymbolText(t).replace("<EOF>", "the end");
	if (s == null) {
		if (getSymbolType(t) == Token.EOF) {
			s = "the end";
		} else {
			s = escapeWSAndQuote("<" + getSymbolType(t) + ">");
		}
	}
	return s;
}
 

public StatementSplitter(String sql, Set<String> delimiters)
{
    TokenSource tokens = getLexer(sql, delimiters);
    ImmutableList.Builder<Statement> list = ImmutableList.builder();
    StringBuilder sb = new StringBuilder();
    while (true) {
        Token token = tokens.nextToken();
        if (token.getType() == Token.EOF) {
            break;
        }
        if (token.getType() == SqlBaseParser.DELIMITER) {
            String statement = sb.toString().trim();
            if (!statement.isEmpty()) {
                list.add(new Statement(statement, token.getText()));
            }
            sb = new StringBuilder();
        }
        else {
            sb.append(token.getText());
        }
    }
    this.completeStatements = list.build();
    this.partialStatement = sb.toString().trim();
}
 

public static String squeezeStatement(String sql)
{
    TokenSource tokens = getLexer(sql, ImmutableSet.of());
    StringBuilder sb = new StringBuilder();
    while (true) {
        Token token = tokens.nextToken();
        if (token.getType() == Token.EOF) {
            break;
        }
        if (token.getType() == SqlBaseLexer.WS) {
            sb.append(' ');
        }
        else {
            sb.append(token.getText());
        }
    }
    return sb.toString().trim();
}
 

public void computeFeatureVectorForToken(int i) {
	Token curToken = tokens.get(i);
	if ( curToken.getType()==Token.EOF ) return;

	int[] features = getFeatures(i);

	int injectNL_WS = getInjectWSCategory(tokens, i);

	int aligned = -1; // "don't care"
	if ( (injectNL_WS&0xFF)==CAT_INJECT_NL ) {
		TerminalNode node = tokenToNodeMap.get(curToken);
		aligned = getAlignmentCategory(doc, node, indentSize);
	}

	// track feature -> injectws, align decisions for token i
	corpus.addExemplar(doc, features, injectNL_WS, aligned);
}
 

public static int getMatchingSymbolEndsLine(Corpus corpus,
	                                            InputDocument doc,
	                                            TerminalNode node)
	{
		TerminalNode matchingLeftNode = getMatchingLeftSymbol(corpus, doc, node);
		if ( matchingLeftNode != null ) {
			Token matchingLeftToken = matchingLeftNode.getSymbol();
			int i = matchingLeftToken.getTokenIndex();
			Token tokenAfterMatchingToken = doc.tokens.getNextRealToken(i);
//			System.out.printf("doc=%s node=%s, pair=%s, after=%s\n",
//			                  new File(doc.fileName).getName(), node.getSymbol(), matchingLeftToken, tokenAfterMatchingToken);
			if ( tokenAfterMatchingToken!=null ) {
				if ( tokenAfterMatchingToken.getType()==Token.EOF ) {
					return 1;
				}
				return tokenAfterMatchingToken.getLine()>matchingLeftToken.getLine() ? 1 : 0;
			}
		}
		return NOT_PAIR;
	}
 

/**
 * Gets the name by which a token can be referenced in the generated code.
 * For tokens defined in a {@code tokens{}} block or via a lexer rule, this
 * is the declared name of the token. For token types generated by the use
 * of a string literal within a parser rule of a combined grammar, this is
 * the automatically generated token type which includes the
 * {@link #AUTO_GENERATED_TOKEN_NAME_PREFIX} prefix. For types which are not
 * associated with a defined token, this method returns
 * {@link #INVALID_TOKEN_NAME}.
 *
 * @param ttype The token type.
 * @return The name of the token with the specified type.
 */

public String getTokenName(int ttype) {
	// inside any target's char range and is lexer grammar?
	if ( isLexer() &&
		 ttype >= Lexer.MIN_CHAR_VALUE && ttype <= Lexer.MAX_CHAR_VALUE )
	{
		return CharSupport.getANTLRCharLiteralForChar(ttype);
	}

	if ( ttype==Token.EOF ) {
		return "EOF";
	}

	if (ttype >= 0 && ttype < typeToTokenList.size() && typeToTokenList.get(ttype) != null) {
		return typeToTokenList.get(ttype);
	}

	return INVALID_TOKEN_NAME;
}
 

public Token nextToken() {
  initIfFirstRun();
  Token t = dentsBuffer.isEmpty()
    ? pullToken()
    : dentsBuffer.remove();
  if (reachedEof) {
    return t;
  }
  final Token r;
  if (t.getType() == nlToken) {
    r = handleNewlineToken(t);
  } else if (t.getType() == Token.EOF) {
    r = eofHandler.apply(t);

  } else {
    r = t;
  }
  return r;
}
 

@Test
public void compareWithAntrl() {
	String s = "select " + "*" + "from dbo.test";
	AntrlResult result = Antlr4Utils.getFull(s);
	SourceLinesProvider p = new SourceLinesProvider();
	SourceLine[] lines = p.getLines(new StringBufferInputStream(s), Charset.defaultCharset());
	FillerRequest file = new FillerRequest(null, null, result.getTree(), lines);
	for (Token t : result.getStream().getTokens()) {
		if (t.getType() == Token.EOF) {
			continue;
		}
		int[] start = file.getLineAndColumn(t.getStartIndex());
		int[] end = file.getLineAndColumn(t.getStopIndex());
		Assert.assertNotNull(start);
		Assert.assertNotNull(end);
		Assert.assertEquals(t.getLine(), start[0]);
		System.out.println(t.getText() + Arrays.toString(start) + " " + t.getCharPositionInLine() + " "
				+ t.getLine() + " " + Arrays.toString(end));
		Assert.assertEquals(t.getCharPositionInLine(), start[1]);
	}
}
 

@Test
public void circularInclude() {
    final URL url = getClass().getResource("/includes/circular-include.raml");
    final RAMLCustomLexer lexer = lexer(url);

    for (Token token = lexer.nextToken(); token.getType() != Token.EOF; token = lexer.nextToken()) {

    }
}
 

public static boolean isEmptyStatement(String sql)
{
    TokenSource tokens = getLexer(sql, ImmutableSet.of());
    while (true) {
        Token token = tokens.nextToken();
        if (token.getType() == Token.EOF) {
            return true;
        }
        if (token.getChannel() != Token.HIDDEN_CHANNEL) {
            return false;
        }
    }
}
 

/**
 * @requires the tokens of this are not initialized yet && source != null
 * @modifies this
 * @effects Initializes the tokens of this with the given token source.
 */
private void initTokens(TokenSource source) {
    Assert.isTrue(tokens.isEmpty());

    Token token;

    do {
        token = source.nextToken();
        tokens.add(token);
    } while (token.getType() != Token.EOF);
}
 

public static String noViableAlt(Parser recognizer, NoViableAltException e) {
    TokenStream tokens = recognizer.getInputStream();
    String input = null;
    if (tokens != null) {
        Token startToken = e.getStartToken();
        if (startToken.getType() == Token.EOF) {
            input = "<EOF>";
        } else {
            input = tokens.getText(
                    startToken, e.getOffendingToken()
            );
        }
    }
    return "syntax error at input:" + input;
}
 

/** Compute a document difference metric 0-1.0 between two documents that
	 *  are identical other than (likely) the whitespace and comments.
	 *
	 *  1.0 means the docs are maximally different and 0 means docs are identical.
	 *
	 *  The Levenshtein distance between the docs counts only
	 *  whitespace diffs as the non-WS content is identical.
	 *  Levenshtein distance is bounded by 0..max(len(doc1),len(doc2)) so
	 *  we normalize the distance by dividing by max WS count.
	 *
	 *  TODO: can we simplify this to a simple walk with two
	 *  cursors through the original vs formatted counting
	 *  mismatched whitespace? real text are like anchors.
	 */
	public static double docDiff(String original,
	                             String formatted,
	                             Class<? extends Lexer> lexerClass)
		throws Exception
	{
		// Grammar must strip all but real tokens and whitespace (and put that on hidden channel)
		CodeBuffTokenStream original_tokens = Tool.tokenize(original, lexerClass);
//		String s = original_tokens.getText();
		CodeBuffTokenStream formatted_tokens = Tool.tokenize(formatted, lexerClass);
//		String t = formatted_tokens.getText();

		// walk token streams and examine whitespace in between tokens
		int i = -1;
		int ws_distance = 0;
		int original_ws = 0;
		int formatted_ws = 0;
		while ( true ) {
			Token ot = original_tokens.LT(i); // TODO: FIX THIS! can't use LT()
			if ( ot==null || ot.getType()==Token.EOF ) break;
			List<Token> ows = original_tokens.getHiddenTokensToLeft(ot.getTokenIndex());
			original_ws += tokenText(ows).length();

			Token ft = formatted_tokens.LT(i); // TODO: FIX THIS! can't use LT()
			if ( ft==null || ft.getType()==Token.EOF ) break;
			List<Token> fws = formatted_tokens.getHiddenTokensToLeft(ft.getTokenIndex());
			formatted_ws += tokenText(fws).length();

			ws_distance += whitespaceEditDistance(tokenText(ows), tokenText(fws));
			i++;
		}
		// it's probably ok to ignore ws diffs after last real token

		int max_ws = Math.max(original_ws, formatted_ws);
		double normalized_ws_distance = ((float) ws_distance)/max_ws;
		return normalized_ws_distance;
	}
 

public static List<Token> getRealTokens(CommonTokenStream tokens) {
	List<Token> real = new ArrayList<>();
	for (int i=0; i<tokens.size(); i++) {
		Token t = tokens.get(i);
		if ( t.getType()!=Token.EOF &&
			t.getChannel()==Lexer.DEFAULT_TOKEN_CHANNEL )
		{
			real.add(t);
		}
	}
	return real;
}