private int parseRegexpFlags(final LexerSource src) throws java.io.IOException {
char kcode = 0;
int options = 0;
int c;
StringBuilder unknownFlags = new StringBuilder(10);
for (c = src.read(); c != RubyYaccLexer.EOF && Character.isLetter(c); c = src.read()) {
switch (c) {
case 'i':
options |= ReOptions.RE_OPTION_IGNORECASE;
break;
case 'x':
options |= ReOptions.RE_OPTION_EXTENDED;
break;
case 'm':
options |= ReOptions.RE_OPTION_MULTILINE;
break;
case 'o':
options |= ReOptions.RE_OPTION_ONCE;
break;
case 'n':
kcode = 16;
break;
case 'e':
kcode = 32;
break;
case 's':
kcode = 48;
break;
case 'u':
kcode = 64;
break;
case 'j':
options |= 256; // Regexp engine 'java'
break;
default:
unknownFlags.append((char) c);
break;
}
}
src.unread(c);
if (unknownFlags.length() != 0) {
throw new SyntaxException(
PID.REGEXP_UNKNOWN_OPTION,
src.getPosition(),
"unknown regexp option"
+ (unknownFlags.length() > 1 ? "s" : "")
+ " - "
+ unknownFlags.toString(),
unknownFlags.toString());
}
return options | kcode;
}
public void testWasBeginOfLineTellsIfPreviousCharactersWereNewline() throws IOException {
final LexerSource src = newSource(".\n=end");
assertTrue(src.lastWasBeginOfLine());
assertFalse(src.wasBeginOfLine());
src.read();
src.read();
assertTrue(src.lastWasBeginOfLine());
assertFalse(src.wasBeginOfLine());
src.read();
assertFalse(src.lastWasBeginOfLine());
assertTrue(src.wasBeginOfLine());
src.read();
assertFalse(src.lastWasBeginOfLine());
assertFalse(src.wasBeginOfLine());
}
public void testCurrentLineGivesAnErrorLocation() throws IOException {
LexerSource src = newSource("111111\n222222\n333333");
for (int i = 0; i < 10; i++) {
src.read();
}
assertEquals("222222\n ^", src.getCurrentLine());
}
@SuppressWarnings("empty-statement")
public void testLinesAndOffsetsAreReported() throws IOException {
LexerSource src = newSource("a\nb\nc\nd\n");
while (src.read() != RubyLexer.EOF) ;
assertEquals(4, src.getLine());
assertEquals(8, src.getOffset());
}
public void testCaptureLinesWithCarriageReturn() throws IOException {
List<String> lines = new ArrayList<String>();
LexerSource src = newSource("1\r\n2\r\n3", lines);
while (src.read() != RubyLexer.EOF) ;
assertEquals(3, lines.size());
assertEquals("1\r\n", lines.get(0));
assertEquals("2\r\n", lines.get(1));
assertEquals("3", lines.get(2));
}
public int parseSimpleStringIntoBuffer(LexerSource src, ByteList buffer)
throws java.io.IOException {
int c;
while ((c = src.read()) != RubyYaccLexer.EOF) {
if (c == end) {
src.unread(c);
break;
} else if (c == '\\') {
c = src.read();
if ((c == '\n' || c != end) && c != '\\') buffer.append('\\');
}
buffer.append(c);
}
return c;
}
@SuppressWarnings("empty-statement")
public void testCaptureLines() throws IOException {
List<String> lines = new ArrayList<String>();
LexerSource src = newSource("111111\n222222\n333333", lines);
while (src.read() != RubyLexer.EOF) ;
assertEquals(3, lines.size());
assertEquals("111111\n", lines.get(0));
assertEquals("222222\n", lines.get(1));
assertEquals("333333", lines.get(2));
}
// This one currently fails with InputStreamLexerSource
public void testCarriageReturnsShouldBeUnreadAutomatically() throws IOException {
LexerSource src = newSource("abc\r\ndef");
assertReadShouldProduce(src, "abc\ndef");
assertEquals(8, src.getOffset());
src.unreadMany("\ndef");
assertEquals(3, src.getOffset());
src.read();
assertEquals(5, src.getOffset());
assertReadShouldProduce(src, "def");
}
// Was a goto in original ruby lexer
private void escaped(LexerSource src, ByteList buffer) throws java.io.IOException {
int c;
switch (c = src.read()) {
case '\\':
parseEscapeIntoBuffer(src, buffer);
break;
case RubyYaccLexer.EOF:
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
default:
buffer.append(c);
}
}
public void testGetRemainingOutputAsStream() throws IOException {
LexerSource src = newSource("111111\n222222\n333333\n");
for (int i = 0; i < 10; i++) {
src.read();
}
final InputStream in = src.getRemainingAsStream();
assertActionShouldProduce(
"222\n333333\n",
new Callable<byte[]>() {
public byte[] call() throws Exception {
ByteList buf = new ByteList();
int c;
while ((c = in.read()) != -1) {
buf.append(c);
}
return buf.bytes();
}
});
}
private void assertReadShouldProduce(LexerSource src, String expected) {
final ByteList actual = new ByteList(expected.length());
for (int i = 0; i < expected.length(); i++) {
try {
int c = src.read();
if (c == RubyLexer.EOF) {
break;
}
actual.append(c);
} catch (IOException ex) {
fail(ex.getMessage());
}
}
assertActionShouldProduce(
expected,
new Callable<byte[]>() {
public byte[] call() throws Exception {
return actual.bytes();
}
});
}
public void testWasBeginOfLineBehaviorAtBeginningOfInput() throws IOException {
final LexerSource src = newSource("=begin");
src.read();
assertFalse(src.lastWasBeginOfLine());
assertTrue(src.wasBeginOfLine());
}
public int parseString(RubyYaccLexer lexer, LexerSource src) throws java.io.IOException {
boolean spaceSeen = false;
int c;
// FIXME: How much more obtuse can this be?
// Heredoc already parsed this and saved string...Do not parse..just return
if (flags == -1) {
lexer.setValue(new Token("\"", lexer.getPosition()));
return Tokens.tSTRING_END;
}
c = src.read();
if ((flags & RubyYaccLexer.STR_FUNC_QWORDS) != 0 && Character.isWhitespace(c)) {
do {
c = src.read();
} while (Character.isWhitespace(c));
spaceSeen = true;
}
if (c == end && nest == 0) {
if ((flags & RubyYaccLexer.STR_FUNC_QWORDS) != 0) {
flags = -1;
lexer.getPosition();
return ' ';
}
if ((flags & RubyYaccLexer.STR_FUNC_REGEXP) != 0) {
lexer.setValue(
new RegexpNode(src.getPosition(), ByteList.create(""), parseRegexpFlags(src)));
return Tokens.tREGEXP_END;
}
lexer.setValue(new Token("\"", lexer.getPosition()));
return Tokens.tSTRING_END;
}
if (spaceSeen) {
src.unread(c);
lexer.getPosition();
return ' ';
}
// Single-quote fast path
if (begin == '\0' && flags == 0) {
ByteList buffer = new ByteList();
src.unread(c);
if (parseSimpleStringIntoBuffer(src, buffer) == RubyYaccLexer.EOF) {
throw new SyntaxException(
PID.STRING_HITS_EOF,
src.getPosition(),
src.getCurrentLine(),
"unterminated string meets end of file");
}
/*
ByteList buffer;
src.unread(c);
if ((buffer = src.readUntil(end)) == null) {
throw new SyntaxException(src.getPosition(), "unterminated string meets end of file");
}
*/
lexer.setValue(new StrNode(lexer.getPosition(), buffer));
return Tokens.tSTRING_CONTENT;
}
ByteList buffer = new ByteList();
if ((flags & RubyYaccLexer.STR_FUNC_EXPAND) != 0 && c == '#') {
c = src.read();
switch (c) {
case '$':
case '@':
src.unread(c);
lexer.setValue(new Token("#" + c, lexer.getPosition()));
return Tokens.tSTRING_DVAR;
case '{':
lexer.setValue(new Token("#" + c, lexer.getPosition()));
return Tokens.tSTRING_DBEG;
}
buffer.append((byte) '#');
}
src.unread(c);
if (parseStringIntoBuffer(lexer, src, buffer) == RubyYaccLexer.EOF) {
throw new SyntaxException(
PID.STRING_HITS_EOF,
src.getPosition(),
src.getCurrentLine(),
"unterminated string meets end of file");
}
lexer.setValue(new StrNode(lexer.getPosition(), buffer));
return Tokens.tSTRING_CONTENT;
}
private void parseEscapeIntoBuffer(LexerSource src, ByteList buffer) throws java.io.IOException {
int c;
switch (c = src.read()) {
case '\n':
break; /* just ignore */
case '0':
case '1':
case '2':
case '3': /* octal constant */
case '4':
case '5':
case '6':
case '7':
buffer.append('\\');
buffer.append(c);
for (int i = 0; i < 2; i++) {
c = src.read();
if (c == RubyYaccLexer.EOF) {
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
}
if (!RubyYaccLexer.isOctChar(c)) {
src.unread(c);
break;
}
buffer.append(c);
}
break;
case 'x': /* hex constant */
buffer.append('\\');
buffer.append(c);
c = src.read();
if (!RubyYaccLexer.isHexChar(c)) {
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
}
buffer.append(c);
c = src.read();
if (RubyYaccLexer.isHexChar(c)) {
buffer.append(c);
} else {
src.unread(c);
}
break;
case 'M':
if ((c = src.read()) != '-') {
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
}
buffer.append(new byte[] {'\\', 'M', '-'});
escaped(src, buffer);
break;
case 'C':
if ((c = src.read()) != '-') {
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
}
buffer.append(new byte[] {'\\', 'C', '-'});
escaped(src, buffer);
break;
case 'c':
buffer.append(new byte[] {'\\', 'c'});
escaped(src, buffer);
break;
case RubyYaccLexer.EOF:
throw new SyntaxException(
PID.INVALID_ESCAPE_SYNTAX,
src.getPosition(),
src.getCurrentLine(),
"Invalid escape character syntax");
default:
if (c != '\\' || c != end) {
buffer.append('\\');
}
buffer.append(c);
}
}
public int parseStringIntoBuffer(RubyYaccLexer lexer, LexerSource src, ByteList buffer)
throws java.io.IOException {
boolean qwords = (flags & RubyYaccLexer.STR_FUNC_QWORDS) != 0;
boolean expand = (flags & RubyYaccLexer.STR_FUNC_EXPAND) != 0;
boolean escape = (flags & RubyYaccLexer.STR_FUNC_ESCAPE) != 0;
boolean regexp = (flags & RubyYaccLexer.STR_FUNC_REGEXP) != 0;
int c;
while ((c = src.read()) != RubyYaccLexer.EOF) {
if (begin != '\0' && c == begin) {
nest++;
} else if (c == end) {
if (nest == 0) {
src.unread(c);
break;
}
nest--;
} else if (c == '#' && expand && !src.peek('\n')) {
int c2 = src.read();
if (c2 == '$' || c2 == '@' || c2 == '{') {
src.unread(c2);
src.unread(c);
break;
}
src.unread(c2);
} else if (c == '\\') {
c = src.read();
switch (c) {
case '\n':
if (qwords) break;
if (expand) continue;
buffer.append('\\');
break;
case '\\':
if (escape) buffer.append(c);
break;
default:
if (regexp) {
src.unread(c);
parseEscapeIntoBuffer(src, buffer);
continue;
} else if (expand) {
src.unread(c);
if (escape) buffer.append('\\');
c = lexer.readEscape();
} else if (qwords && Character.isWhitespace(c)) {
/* ignore backslashed spaces in %w */
} else if (c != end && !(begin != '\0' && c == begin)) {
buffer.append('\\');
}
}
} else if (qwords && Character.isWhitespace(c)) {
src.unread(c);
break;
}
buffer.append(c);
}
return c;
}
public void testUnreadBothNewAndExistingCharacters() throws IOException {
LexerSource src = newSource("abc");
src.read();
src.unreadMany("123a");
assertReadShouldProduce(src, "123abc");
}
public void testPeekGivesTheNextCharaterWithoutAdvancing() throws IOException {
LexerSource src = newSource("abc");
assertTrue(src.peek('a'));
assertEquals('a', src.read());
}