* A regular expression object (class RE) is compiled by constructing it
* from a String, StringBuffer or character array, with optional
* compilation flags (below)
* and an optional syntax specification (see RESyntax; if not specified,
* RESyntax.RE_SYNTAX_PERL5 is used).
*
* Various methods attempt to match input text against a compiled * regular expression. These methods are: *
isMatch: returns true if the input text in its entirety
* matches the regular expression pattern.
* getMatch: returns the first match found in the input text,
* or null if no match is found.
* getAllMatches: returns an array of all non-overlapping
* matches found in the input text. If no matches are found, the array is
* zero-length.
* substitute: substitute the first occurence of the pattern
* in the input text with a replacement string (which may include
* metacharacters $0-$9, see REMatch.substituteInto).
* substituteAll: same as above, but repeat for each match
* before returning.
* getMatchEnumeration: returns an REMatchEnumeration object
* that allows iteration over the matches (see REMatchEnumeration for some
* reasons why you may want to do this instead of using getAllMatches.
* * * These methods all have similar argument lists. The input can be a * String, a character array, a StringBuffer, a Reader or an * InputStream of some sort. Note that when using a Reader or * InputStream, the stream read position cannot be guaranteed after * attempting a match (this is not a bug, but a consequence of the way * regular expressions work). Using an REMatchEnumeration can * eliminate most positioning problems. * *
* * The optional index argument specifies the offset from the beginning * of the text at which the search should start (see the descriptions * of some of the execution flags for how this can affect positional * pattern operators). For a Reader or InputStream, this means an * offset from the current read position, so subsequent calls with the * same index argument on a Reader or an InputStream will not * necessarily access the same position on the stream, whereas * repeated searches at a given index in a fixed string will return * consistent results. * *
* You can optionally affect the execution environment by using a * combination of execution flags (constants listed below). * *
* All operations on a regular expression are performed in a * thread-safe manner. * * @author Wes Biggs * @version 1.1.4-dev, to be released */ public static class RE extends REToken { // This String will be returned by getVersion() private static final String VERSION = "1.1.4-dev"; // The localized strings are kept in a separate file private static ResourceBundle messages = new ListResourceBundle() { public Object[][] getContents() { return stuff; } private Object[][] stuff = new Object[][] { { "error.prefix", "At position {0} in regular expression pattern:" }, { "repeat.assertion", "repeated token is zero-width assertion" }, { "repeat.chained", "attempted to repeat a token that is already repeated" }, { "repeat.no.token", "quantifier (?*+{}) without preceding token" }, { "repeat.empty.token", "repeated token may be empty" }, { "unmatched.brace", "unmatched brace" }, { "unmatched.bracket", "unmatched bracket" }, { "unmatched.paren", "unmatched parenthesis" }, { "interval.no.end", "expected end of interval" }, { "class.no.end", "expected end of character class" }, { "subexpr.no.end", "expected end of subexpression" }, { "interval.order", "interval minimum is greater than maximum" }, { "interval.error", "interval is empty or contains illegal chracters" }, { "ends.with.backslash", "backslash at end of pattern" }, { "syntax.final", "Syntax has been declared final and cannot be modified" } }; }; // These are, respectively, the first and last tokens in our linked list // If there is only one token, firstToken == lastToken private REToken firstToken, lastToken; // This is the number of subexpressions in this regular expression, // with a minimum value of zero. Returned by getNumSubs() private int numSubs; /** Minimum length, in characters, of any possible match. */ private int minimumLength; /** * Compilation flag. Do not differentiate case. Subsequent * searches using this RE will be case insensitive. */ public static final int REG_ICASE = 2; /** * Compilation flag. The match-any-character operator (dot) * will match a newline character. When set this overrides the syntax * bit RE_DOT_NEWLINE (see RESyntax for details). This is equivalent to * the "/s" operator in Perl. */ public static final int REG_DOT_NEWLINE = 4; /** * Compilation flag. Use multiline mode. In this mode, the ^ and $ * anchors will match based on newlines within the input. This is * equivalent to the "/m" operator in Perl. */ public static final int REG_MULTILINE = 8; /** * Execution flag. * The match-beginning operator (^) will not match at the beginning * of the input string. Useful for matching on a substring when you * know the context of the input is such that position zero of the * input to the match test is not actually position zero of the text. *
* This example demonstrates the results of various ways of matching on * a substring. *
*
* // Results:
* String s = "food bar fool";
*/
public static final int REG_NOTBOL = 16;
/**
* Execution flag.
* The match-end operator ($) does not match at the end
* of the input string. Useful for matching on substrings.
*/
public static final int REG_NOTEOL = 32;
/**
* Execution flag.
* When a match method is invoked that starts matching at a non-zero
* index into the input, treat the input as if it begins at the index
* given. The effect of this flag is that the engine does not "see"
* any text in the input before the given index. This is useful so
* that the match-beginning operator (^) matches not at position 0
* in the input string, but at the position the search started at
* (based on the index input given to the getMatch function). See
* the example under REG_NOTBOL. It also affects the use of the \<
* and \b operators.
*/
public static final int REG_ANCHORINDEX = 64;
/**
* Execution flag.
* The substitute and substituteAll methods will not attempt to
* interpolate occurrences of $1-$9 in the replacement text with
* the corresponding subexpressions. For example, you may want to
* replace all matches of "one dollar" with "$1".
*/
public static final int REG_NO_INTERPOLATE = 128;
/** Returns a string representing the version of the gnu.regexp package. */
public static final String version() {
return VERSION;
}
// Retrieves a message from the ResourceBundle
static final String getLocalizedMessage(String key) {
return messages.getString(key);
}
/**
* Constructs a regular expression pattern buffer without any compilation
* flags set, and using the default syntax (RESyntax.RE_SYNTAX_PERL5).
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer or char[]. Other input types will be converted to
* strings using the toString() method.
* @exception REException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public RE(Object pattern) throws REException {
this(pattern,0,RESyntax.RE_SYNTAX_PERL5,0,0);
}
/**
* Constructs a regular expression pattern buffer using the specified
* compilation flags and the default syntax (RESyntax.RE_SYNTAX_PERL5).
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer, or char[]. Other input types will be converted to
* strings using the toString() method.
* @param cflags The logical OR of any combination of the compilation flags listed above.
* @exception REException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public RE(Object pattern, int cflags) throws REException {
this(pattern,cflags,RESyntax.RE_SYNTAX_PERL5,0,0);
}
/**
* Constructs a regular expression pattern buffer using the specified
* compilation flags and regular expression syntax.
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer, or char[]. Other input types will be converted to
* strings using the toString() method.
* @param cflags The logical OR of any combination of the compilation flags listed above.
* @param syntax The type of regular expression syntax to use.
* @exception REException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public RE(Object pattern, int cflags, RESyntax syntax) throws REException {
this(pattern,cflags,syntax,0,0);
}
// internal constructor used for alternation
private RE(REToken first, REToken last,int subs, int subIndex, int minLength) {
super(subIndex);
firstToken = first;
lastToken = last;
numSubs = subs;
minimumLength = minLength;
addToken(new RETokenEndSub(subIndex));
}
private RE(Object patternObj, int cflags, RESyntax syntax, int myIndex, int nextSub) throws REException {
super(myIndex); // Subexpression index of this token.
initialize(patternObj, cflags, syntax, myIndex, nextSub);
}
// For use by subclasses
protected RE() { super(0); }
// The meat of construction
protected void initialize(Object patternObj, int cflags, RESyntax syntax, int myIndex, int nextSub) throws REException {
char[] pattern;
if (patternObj instanceof String) {
pattern = ((String) patternObj).toCharArray();
} else if (patternObj instanceof char[]) {
pattern = (char[]) patternObj;
} else if (patternObj instanceof StringBuffer) {
pattern = new char [((StringBuffer) patternObj).length()];
((StringBuffer) patternObj).getChars(0,pattern.length,pattern,0);
} else {
pattern = patternObj.toString().toCharArray();
}
int pLength = pattern.length;
numSubs = 0; // Number of subexpressions in this token.
Vector branches = null;
// linked list of tokens (sort of -- some closed loops can exist)
firstToken = lastToken = null;
// Precalculate these so we don't pay for the math every time we
// need to access them.
boolean insens = ((cflags & REG_ICASE) > 0);
// Parse pattern into tokens. Does anyone know if it's more efficient
// to use char[] than a String.charAt()? I'm assuming so.
// index tracks the position in the char array
int index = 0;
// this will be the current parse character (pattern[index])
CharUnit unit = new CharUnit();
// This is used for {x,y} calculations
IntPair minMax = new IntPair();
// Buffer a token so we can create a TokenRepeated, etc.
REToken currentToken = null;
char ch;
while (index < pLength) {
// read the next character unit (including backslash escapes)
index = getCharUnit(pattern,index,unit);
// ALTERNATION OPERATOR
// \| or | (if RE_NO_BK_VBAR) or newline (if RE_NEWLINE_ALT)
// not available if RE_LIMITED_OPS is set
// TODO: the '\n' literal here should be a test against REToken.newline,
// which unfortunately may be more than a single character.
if ( ( (unit.ch == '|' && (syntax.get(RESyntax.RE_NO_BK_VBAR) ^ unit.bk))
|| (syntax.get(RESyntax.RE_NEWLINE_ALT) && (unit.ch == '\n') && !unit.bk) )
&& !syntax.get(RESyntax.RE_LIMITED_OPS)) {
// make everything up to here be a branch. create vector if nec.
addToken(currentToken);
RE theBranch = new RE(firstToken, lastToken, numSubs, subIndex, minimumLength);
minimumLength = 0;
if (branches == null) {
branches = new Vector();
}
branches.addElement(theBranch);
firstToken = lastToken = currentToken = null;
}
// INTERVAL OPERATOR:
// {x} | {x,} | {x,y} (RE_INTERVALS && RE_NO_BK_BRACES)
// \{x\} | \{x,\} | \{x,y\} (RE_INTERVALS && !RE_NO_BK_BRACES)
//
// OPEN QUESTION:
// what is proper interpretation of '{' at start of string?
else if ((unit.ch == '{') && syntax.get(RESyntax.RE_INTERVALS) && (syntax.get(RESyntax.RE_NO_BK_BRACES) ^ unit.bk)) {
int newIndex = getMinMax(pattern,index,minMax,syntax);
if (newIndex > index) {
if (minMax.first > minMax.second)
throw new REException(getLocalizedMessage("interval.order"),REException.REG_BADRPT,newIndex);
if (currentToken == null)
throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,newIndex);
if (currentToken instanceof RETokenRepeated)
throw new REException(getLocalizedMessage("repeat.chained"),REException.REG_BADRPT,newIndex);
if (currentToken instanceof RETokenWordBoundary || currentToken instanceof RETokenWordBoundary)
throw new REException(getLocalizedMessage("repeat.assertion"),REException.REG_BADRPT,newIndex);
if ((currentToken.getMinimumLength() == 0) && (minMax.second == Integer.MAX_VALUE))
throw new REException(getLocalizedMessage("repeat.empty.token"),REException.REG_BADRPT,newIndex);
index = newIndex;
currentToken = setRepeated(currentToken,minMax.first,minMax.second,index);
}
else {
addToken(currentToken);
currentToken = new RETokenChar(subIndex,unit.ch,insens);
}
}
// LIST OPERATOR:
// [...] | [^...]
else if ((unit.ch == '[') && !unit.bk) {
Vector options = new Vector();
boolean negative = false;
char lastChar = 0;
if (index == pLength) throw new REException(getLocalizedMessage("unmatched.bracket"),REException.REG_EBRACK,index);
// Check for initial caret, negation
if ((ch = pattern[index]) == '^') {
negative = true;
if (++index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
ch = pattern[index];
}
// Check for leading right bracket literal
if (ch == ']') {
lastChar = ch;
if (++index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
}
while ((ch = pattern[index++]) != ']') {
if ((ch == '-') && (lastChar != 0)) {
if (index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
if ((ch = pattern[index]) == ']') {
options.addElement(new RETokenChar(subIndex,lastChar,insens));
lastChar = '-';
} else {
options.addElement(new RETokenRange(subIndex,lastChar,ch,insens));
lastChar = 0;
index++;
}
} else if ((ch == '\\') && syntax.get(RESyntax.RE_BACKSLASH_ESCAPE_IN_LISTS)) {
if (index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
int posixID = -1;
boolean negate = false;
char asciiEsc = 0;
if (("dswDSW".indexOf(pattern[index]) != -1) && syntax.get(RESyntax.RE_CHAR_CLASS_ESC_IN_LISTS)) {
switch (pattern[index]) {
case 'D':
negate = true;
case 'd':
posixID = RETokenPOSIX.DIGIT;
break;
case 'S':
negate = true;
case 's':
posixID = RETokenPOSIX.SPACE;
break;
case 'W':
negate = true;
case 'w':
posixID = RETokenPOSIX.ALNUM;
break;
}
}
else if ("nrt".indexOf(pattern[index]) != -1) {
switch (pattern[index]) {
case 'n':
asciiEsc = '\n';
break;
case 't':
asciiEsc = '\t';
break;
case 'r':
asciiEsc = '\r';
break;
}
}
if (lastChar != 0) options.addElement(new RETokenChar(subIndex,lastChar,insens));
if (posixID != -1) {
options.addElement(new RETokenPOSIX(subIndex,posixID,insens,negate));
} else if (asciiEsc != 0) {
lastChar = asciiEsc;
} else {
lastChar = pattern[index];
}
++index;
} else if ((ch == '[') && (syntax.get(RESyntax.RE_CHAR_CLASSES)) && (index < pLength) && (pattern[index] == ':')) {
StringBuffer posixSet = new StringBuffer();
index = getPosixSet(pattern,index+1,posixSet);
int posixId = RETokenPOSIX.intValue(posixSet.toString());
if (posixId != -1)
options.addElement(new RETokenPOSIX(subIndex,posixId,insens,false));
} else {
if (lastChar != 0) options.addElement(new RETokenChar(subIndex,lastChar,insens));
lastChar = ch;
}
if (index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
} // while in list
// Out of list, index is one past ']'
if (lastChar != 0) options.addElement(new RETokenChar(subIndex,lastChar,insens));
// Create a new RETokenOneOf
addToken(currentToken);
options.trimToSize();
currentToken = new RETokenOneOf(subIndex,options,negative);
}
// SUBEXPRESSIONS
// (...) | \(...\) depending on RE_NO_BK_PARENS
else if ((unit.ch == '(') && (syntax.get(RESyntax.RE_NO_BK_PARENS) ^ unit.bk)) {
boolean pure = false;
boolean comment = false;
boolean lookAhead = false;
boolean negativelh = false;
if ((index+1 < pLength) && (pattern[index] == '?')) {
switch (pattern[index+1]) {
case '!':
if (syntax.get(RESyntax.RE_LOOKAHEAD)) {
pure = true;
negativelh = true;
lookAhead = true;
index += 2;
}
break;
case '=':
if (syntax.get(RESyntax.RE_LOOKAHEAD)) {
pure = true;
lookAhead = true;
index += 2;
}
break;
case ':':
if (syntax.get(RESyntax.RE_PURE_GROUPING)) {
pure = true;
index += 2;
}
break;
case '#':
if (syntax.get(RESyntax.RE_COMMENTS)) {
comment = true;
}
break;
default:
throw new REException(getLocalizedMessage("repeat.no.token"), REException.REG_BADRPT, index);
}
}
if (index >= pLength) {
throw new REException(getLocalizedMessage("unmatched.paren"), REException.REG_ESUBREG,index);
}
// find end of subexpression
int endIndex = index;
int nextIndex = index;
int nested = 0;
while ( ((nextIndex = getCharUnit(pattern,endIndex,unit)) > 0)
&& !(nested == 0 && (unit.ch == ')') && (syntax.get(RESyntax.RE_NO_BK_PARENS) ^ unit.bk)) )
if ((endIndex = nextIndex) >= pLength)
throw new REException(getLocalizedMessage("subexpr.no.end"),REException.REG_ESUBREG,nextIndex);
else if (unit.ch == '(' && (syntax.get(RESyntax.RE_NO_BK_PARENS) ^ unit.bk))
nested++;
else if (unit.ch == ')' && (syntax.get(RESyntax.RE_NO_BK_PARENS) ^ unit.bk))
nested--;
// endIndex is now position at a ')','\)'
// nextIndex is end of string or position after ')' or '\)'
if (comment) index = nextIndex;
else { // not a comment
// create RE subexpression as token.
addToken(currentToken);
if (!pure) {
numSubs++;
}
int useIndex = (pure || lookAhead) ? 0 : nextSub + numSubs;
currentToken = new RE(String.valueOf(pattern,index,endIndex-index).toCharArray(),cflags,syntax,useIndex,nextSub + numSubs);
numSubs += ((RE) currentToken).getNumSubs();
if (lookAhead) {
currentToken = new RETokenLookAhead(currentToken,negativelh);
}
index = nextIndex;
} // not a comment
} // subexpression
// UNMATCHED RIGHT PAREN
// ) or \) throw exception if
// !syntax.get(RESyntax.RE_UNMATCHED_RIGHT_PAREN_ORD)
else if (!syntax.get(RESyntax.RE_UNMATCHED_RIGHT_PAREN_ORD) && ((unit.ch == ')') && (syntax.get(RESyntax.RE_NO_BK_PARENS) ^ unit.bk))) {
throw new REException(getLocalizedMessage("unmatched.paren"),REException.REG_EPAREN,index);
}
// START OF LINE OPERATOR
// ^
else if ((unit.ch == '^') && !unit.bk) {
addToken(currentToken);
currentToken = null;
addToken(new RETokenStart(subIndex,((cflags & REG_MULTILINE) > 0) ? syntax.getLineSeparator() : null));
}
// END OF LINE OPERATOR
// $
else if ((unit.ch == '$') && !unit.bk) {
addToken(currentToken);
currentToken = null;
addToken(new RETokenEnd(subIndex,((cflags & REG_MULTILINE) > 0) ? syntax.getLineSeparator() : null));
}
// MATCH-ANY-CHARACTER OPERATOR (except possibly newline and null)
// .
else if ((unit.ch == '.') && !unit.bk) {
addToken(currentToken);
currentToken = new RETokenAny(subIndex,syntax.get(RESyntax.RE_DOT_NEWLINE) || ((cflags & REG_DOT_NEWLINE) > 0),syntax.get(RESyntax.RE_DOT_NOT_NULL));
}
// ZERO-OR-MORE REPEAT OPERATOR
// *
else if ((unit.ch == '*') && !unit.bk) {
if (currentToken == null)
throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,index);
if (currentToken instanceof RETokenRepeated)
throw new REException(getLocalizedMessage("repeat.chained"),REException.REG_BADRPT,index);
if (currentToken instanceof RETokenWordBoundary || currentToken instanceof RETokenWordBoundary)
throw new REException(getLocalizedMessage("repeat.assertion"),REException.REG_BADRPT,index);
if (currentToken.getMinimumLength() == 0)
throw new REException(getLocalizedMessage("repeat.empty.token"),REException.REG_BADRPT,index);
currentToken = setRepeated(currentToken,0,Integer.MAX_VALUE,index);
}
// ONE-OR-MORE REPEAT OPERATOR
// + | \+ depending on RE_BK_PLUS_QM
// not available if RE_LIMITED_OPS is set
else if ((unit.ch == '+') && !syntax.get(RESyntax.RE_LIMITED_OPS) && (!syntax.get(RESyntax.RE_BK_PLUS_QM) ^ unit.bk)) {
if (currentToken == null)
throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,index);
if (currentToken instanceof RETokenRepeated)
throw new REException(getLocalizedMessage("repeat.chained"),REException.REG_BADRPT,index);
if (currentToken instanceof RETokenWordBoundary || currentToken instanceof RETokenWordBoundary)
throw new REException(getLocalizedMessage("repeat.assertion"),REException.REG_BADRPT,index);
if (currentToken.getMinimumLength() == 0)
throw new REException(getLocalizedMessage("repeat.empty.token"),REException.REG_BADRPT,index);
currentToken = setRepeated(currentToken,1,Integer.MAX_VALUE,index);
}
// ZERO-OR-ONE REPEAT OPERATOR / STINGY MATCHING OPERATOR
// ? | \? depending on RE_BK_PLUS_QM
// not available if RE_LIMITED_OPS is set
// stingy matching if RE_STINGY_OPS is set and it follows a quantifier
else if ((unit.ch == '?') && !syntax.get(RESyntax.RE_LIMITED_OPS) && (!syntax.get(RESyntax.RE_BK_PLUS_QM) ^ unit.bk)) {
if (currentToken == null) throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,index);
// Check for stingy matching on RETokenRepeated
if (currentToken instanceof RETokenRepeated) {
if (syntax.get(RESyntax.RE_STINGY_OPS) && !((RETokenRepeated)currentToken).isStingy())
((RETokenRepeated)currentToken).makeStingy();
else
throw new REException(getLocalizedMessage("repeat.chained"),REException.REG_BADRPT,index);
}
else if (currentToken instanceof RETokenWordBoundary || currentToken instanceof RETokenWordBoundary)
throw new REException(getLocalizedMessage("repeat.assertion"),REException.REG_BADRPT,index);
else
currentToken = setRepeated(currentToken,0,1,index);
}
// BACKREFERENCE OPERATOR
// \1 \2 ... \9
// not available if RE_NO_BK_REFS is set
else if (unit.bk && Character.isDigit(unit.ch) && !syntax.get(RESyntax.RE_NO_BK_REFS)) {
addToken(currentToken);
currentToken = new RETokenBackRef(subIndex,Character.digit(unit.ch,10),insens);
}
// START OF STRING OPERATOR
// \A if RE_STRING_ANCHORS is set
else if (unit.bk && (unit.ch == 'A') && syntax.get(RESyntax.RE_STRING_ANCHORS)) {
addToken(currentToken);
currentToken = new RETokenStart(subIndex,null);
}
// WORD BREAK OPERATOR
// \b if ????
else if (unit.bk && (unit.ch == 'b') && syntax.get(RESyntax.RE_STRING_ANCHORS)) {
addToken(currentToken);
currentToken = new RETokenWordBoundary(subIndex, RETokenWordBoundary.BEGIN | RETokenWordBoundary.END, false);
}
// WORD BEGIN OPERATOR
// \< if ????
else if (unit.bk && (unit.ch == '<')) {
addToken(currentToken);
currentToken = new RETokenWordBoundary(subIndex, RETokenWordBoundary.BEGIN, false);
}
// WORD END OPERATOR
// \> if ????
else if (unit.bk && (unit.ch == '>')) {
addToken(currentToken);
currentToken = new RETokenWordBoundary(subIndex, RETokenWordBoundary.END, false);
}
// NON-WORD BREAK OPERATOR
// \B if ????
else if (unit.bk && (unit.ch == 'B') && syntax.get(RESyntax.RE_STRING_ANCHORS)) {
addToken(currentToken);
currentToken = new RETokenWordBoundary(subIndex, RETokenWordBoundary.BEGIN | RETokenWordBoundary.END, true);
}
// DIGIT OPERATOR
// \d if RE_CHAR_CLASS_ESCAPES is set
else if (unit.bk && (unit.ch == 'd') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.DIGIT,insens,false);
}
// NON-DIGIT OPERATOR
// \D
else if (unit.bk && (unit.ch == 'D') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.DIGIT,insens,true);
}
// NEWLINE ESCAPE
// \n
else if (unit.bk && (unit.ch == 'n')) {
addToken(currentToken);
currentToken = new RETokenChar(subIndex,'\n',false);
}
// RETURN ESCAPE
// \r
else if (unit.bk && (unit.ch == 'r')) {
addToken(currentToken);
currentToken = new RETokenChar(subIndex,'\r',false);
}
// WHITESPACE OPERATOR
// \s if RE_CHAR_CLASS_ESCAPES is set
else if (unit.bk && (unit.ch == 's') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.SPACE,insens,false);
}
// NON-WHITESPACE OPERATOR
// \S
else if (unit.bk && (unit.ch == 'S') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.SPACE,insens,true);
}
// TAB ESCAPE
// \t
else if (unit.bk && (unit.ch == 't')) {
addToken(currentToken);
currentToken = new RETokenChar(subIndex,'\t',false);
}
// ALPHANUMERIC OPERATOR
// \w
else if (unit.bk && (unit.ch == 'w') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.ALNUM,insens,false);
}
// NON-ALPHANUMERIC OPERATOR
// \W
else if (unit.bk && (unit.ch == 'W') && syntax.get(RESyntax.RE_CHAR_CLASS_ESCAPES)) {
addToken(currentToken);
currentToken = new RETokenPOSIX(subIndex,RETokenPOSIX.ALNUM,insens,true);
}
// END OF STRING OPERATOR
// \Z
else if (unit.bk && (unit.ch == 'Z') && syntax.get(RESyntax.RE_STRING_ANCHORS)) {
addToken(currentToken);
currentToken = new RETokenEnd(subIndex,null);
}
// NON-SPECIAL CHARACTER (or escape to make literal)
// c | \* for example
else { // not a special character
addToken(currentToken);
currentToken = new RETokenChar(subIndex,unit.ch,insens);
}
} // end while
// Add final buffered token and an EndSub marker
addToken(currentToken);
if (branches != null) {
branches.addElement(new RE(firstToken,lastToken,numSubs,subIndex,minimumLength));
branches.trimToSize(); // compact the Vector
minimumLength = 0;
firstToken = lastToken = null;
addToken(new RETokenOneOf(subIndex,branches,false));
}
else addToken(new RETokenEndSub(subIndex));
}
private static int getCharUnit(char[] input, int index, CharUnit unit) throws REException {
unit.ch = input[index++];
if (unit.bk = (unit.ch == '\\'))
if (index < input.length)
unit.ch = input[index++];
else throw new REException(getLocalizedMessage("ends.with.backslash"),REException.REG_ESCAPE,index);
return index;
}
/**
* Checks if the regular expression matches the input in its entirety.
*
* @param input The input text.
*/
public boolean isMatch(Object input) {
return isMatch(input,0,0);
}
/**
* Checks if the input string, starting from index, is an exact match of
* this regular expression.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
*/
public boolean isMatch(Object input,int index) {
return isMatch(input,index,0);
}
/**
* Checks if the input, starting from index and using the specified
* execution flags, is an exact match of this regular expression.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
*/
public boolean isMatch(Object input,int index,int eflags) {
return isMatchImpl(makeCharIndexed(input,index),index,eflags);
}
private boolean isMatchImpl(CharIndexed input, int index, int eflags) {
if (firstToken == null) // Trivial case
return (input.charAt(0) == CharIndexed.OUT_OF_BOUNDS);
REMatch m = new REMatch(numSubs, index, eflags);
if (firstToken.match(input, m)) {
while (m != null) {
if (input.charAt(m.index) == CharIndexed.OUT_OF_BOUNDS) {
return true;
}
m = m.next;
}
}
return false;
}
/**
* Returns the maximum number of subexpressions in this regular expression.
* If the expression contains branches, the value returned will be the
* maximum subexpressions in any of the branches.
*/
public int getNumSubs() {
return numSubs;
}
// Overrides REToken.setUncle
void setUncle(REToken uncle) {
if (lastToken != null) {
lastToken.setUncle(uncle);
} else super.setUncle(uncle); // to deal with empty subexpressions
}
// Overrides REToken.chain
boolean chain(REToken next) {
super.chain(next);
setUncle(next);
return true;
}
/**
* Returns the minimum number of characters that could possibly
* constitute a match of this regular expression.
*/
public int getMinimumLength() {
return minimumLength;
}
/**
* Returns an array of all matches found in the input.
*
* If the regular expression allows the empty string to match, it will
* substitute matches at all positions except the end of the input.
*
* @param input The input text.
* @return a non-null (but possibly zero-length) array of matches
*/
public REMatch[] getAllMatches(Object input) {
return getAllMatches(input,0,0);
}
/**
* Returns an array of all matches found in the input,
* beginning at the specified index position.
*
* If the regular expression allows the empty string to match, it will
* substitute matches at all positions except the end of the input.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @return a non-null (but possibly zero-length) array of matches
*/
public REMatch[] getAllMatches(Object input, int index) {
return getAllMatches(input,index,0);
}
/**
* Returns an array of all matches found in the input string,
* beginning at the specified index position and using the specified
* execution flags.
*
* If the regular expression allows the empty string to match, it will
* substitute matches at all positions except the end of the input.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @return a non-null (but possibly zero-length) array of matches
*/
public REMatch[] getAllMatches(Object input, int index, int eflags) {
return getAllMatchesImpl(makeCharIndexed(input,index),index,eflags);
}
// this has been changed since 1.03 to be non-overlapping matches
private REMatch[] getAllMatchesImpl(CharIndexed input, int index, int eflags) {
Vector all = new Vector();
REMatch m = null;
while ((m = getMatchImpl(input,index,eflags,null)) != null) {
all.addElement(m);
index = m.getEndIndex();
if (m.end[0] == 0) { // handle pathological case of zero-length match
index++;
input.move(1);
} else {
input.move(m.end[0]);
}
if (!input.isValid()) break;
}
REMatch[] mset = new REMatch[all.size()];
all.copyInto(mset);
return mset;
}
/* Implements abstract method REToken.match() */
boolean match(CharIndexed input, REMatch mymatch) {
if (firstToken == null) return next(input, mymatch);
// Note the start of this subexpression
mymatch.start[subIndex] = mymatch.index;
return firstToken.match(input, mymatch);
}
/**
* Returns the first match found in the input. If no match is found,
* null is returned.
*
* @param input The input text.
* @return An REMatch instance referencing the match, or null if none.
*/
public REMatch getMatch(Object input) {
return getMatch(input,0,0);
}
/**
* Returns the first match found in the input, beginning
* the search at the specified index. If no match is found,
* returns null.
*
* @param input The input text.
* @param index The offset within the text to begin looking for a match.
* @return An REMatch instance referencing the match, or null if none.
*/
public REMatch getMatch(Object input, int index) {
return getMatch(input,index,0);
}
/**
* Returns the first match found in the input, beginning
* the search at the specified index, and using the specified
* execution flags. If no match is found, returns null.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @return An REMatch instance referencing the match, or null if none.
*/
public REMatch getMatch(Object input, int index, int eflags) {
return getMatch(input,index,eflags,null);
}
/**
* Returns the first match found in the input, beginning the search
* at the specified index, and using the specified execution flags.
* If no match is found, returns null. If a StringBuffer is
* provided and is non-null, the contents of the input text from the
* index to the beginning of the match (or to the end of the input,
* if there is no match) are appended to the StringBuffer.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @param buffer The StringBuffer to save pre-match text in.
* @return An REMatch instance referencing the match, or null if none. */
public REMatch getMatch(Object input, int index, int eflags, StringBuffer buffer) {
return getMatchImpl(makeCharIndexed(input,index),index,eflags,buffer);
}
REMatch getMatchImpl(CharIndexed input, int anchor, int eflags, StringBuffer buffer) {
// Create a new REMatch to hold results
REMatch mymatch = new REMatch(numSubs, anchor, eflags);
do {
// Optimization: check if anchor + minimumLength > length
if (minimumLength == 0 || input.charAt(minimumLength-1) != CharIndexed.OUT_OF_BOUNDS) {
if (match(input, mymatch)) {
// Find longest match of them all to observe leftmost longest
REMatch longest = mymatch;
while ((mymatch = mymatch.next) != null) {
if (mymatch.index > longest.index) {
longest = mymatch;
}
}
longest.end[0] = longest.index;
longest.finish(input);
return longest;
}
}
mymatch.clear(++anchor);
// Append character to buffer if needed
if (buffer != null && input.charAt(0) != CharIndexed.OUT_OF_BOUNDS) {
buffer.append(input.charAt(0));
}
} while (input.move(1));
return null;
}
/**
* Returns an REMatchEnumeration that can be used to iterate over the
* matches found in the input text.
*
* @param input The input text.
* @return A non-null REMatchEnumeration instance.
*/
public REMatchEnumeration getMatchEnumeration(Object input) {
return getMatchEnumeration(input,0,0);
}
/**
* Returns an REMatchEnumeration that can be used to iterate over the
* matches found in the input text.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @return A non-null REMatchEnumeration instance, with its input cursor
* set to the index position specified.
*/
public REMatchEnumeration getMatchEnumeration(Object input, int index) {
return getMatchEnumeration(input,index,0);
}
/**
* Returns an REMatchEnumeration that can be used to iterate over the
* matches found in the input text.
*
* @param input The input text.
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @return A non-null REMatchEnumeration instance, with its input cursor
* set to the index position specified.
*/
public REMatchEnumeration getMatchEnumeration(Object input, int index, int eflags) {
return new REMatchEnumeration(this,makeCharIndexed(input,index),index,eflags);
}
/**
* Substitutes the replacement text for the first match found in the input.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @return A String interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substitute(Object input,String replace) {
return substitute(input,replace,0,0);
}
/**
* Substitutes the replacement text for the first match found in the input
* beginning at the specified index position. Specifying an index
* effectively causes the regular expression engine to throw away the
* specified number of characters.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @param index The offset index at which the search should be begin.
* @return A String containing the substring of the input, starting
* at the index position, and interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substitute(Object input,String replace,int index) {
return substitute(input,replace,index,0);
}
/**
* Substitutes the replacement text for the first match found in the input
* string, beginning at the specified index position and using the
* specified execution flags.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @return A String containing the substring of the input, starting
* at the index position, and interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substitute(Object input,String replace,int index,int eflags) {
return substituteImpl(makeCharIndexed(input,index),replace,index,eflags);
}
private String substituteImpl(CharIndexed input,String replace,int index,int eflags) {
StringBuffer buffer = new StringBuffer();
REMatch m = getMatchImpl(input,index,eflags,buffer);
if (m==null) return buffer.toString();
buffer.append( ((eflags & REG_NO_INTERPOLATE) > 0) ?
replace : m.substituteInto(replace) );
if (input.move(m.end[0])) {
do {
buffer.append(input.charAt(0));
} while (input.move(1));
}
return buffer.toString();
}
/**
* Substitutes the replacement text for each non-overlapping match found
* in the input text.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @return A String interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substituteAll(Object input,String replace) {
return substituteAll(input,replace,0,0);
}
/**
* Substitutes the replacement text for each non-overlapping match found
* in the input text, starting at the specified index.
*
* If the regular expression allows the empty string to match, it will
* substitute matches at all positions except the end of the input.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @param index The offset index at which the search should be begin.
* @return A String containing the substring of the input, starting
* at the index position, and interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substituteAll(Object input,String replace,int index) {
return substituteAll(input,replace,index,0);
}
/**
* Substitutes the replacement text for each non-overlapping match found
* in the input text, starting at the specified index and using the
* specified execution flags.
*
* @param input The input text.
* @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
* @param index The offset index at which the search should be begin.
* @param eflags The logical OR of any execution flags above.
* @return A String containing the substring of the input, starting
* at the index position, and interpolating the substituted text.
* @see REMatch#substituteInto
*/
public String substituteAll(Object input,String replace,int index,int eflags) {
return substituteAllImpl(makeCharIndexed(input,index),replace,index,eflags);
}
private String substituteAllImpl(CharIndexed input,String replace,int index,int eflags) {
StringBuffer buffer = new StringBuffer();
REMatch m;
while ((m = getMatchImpl(input,index,eflags,buffer)) != null) {
buffer.append( ((eflags & REG_NO_INTERPOLATE) > 0) ?
replace : m.substituteInto(replace) );
index = m.getEndIndex();
if (m.end[0] == 0) {
char ch = input.charAt(0);
if (ch != CharIndexed.OUT_OF_BOUNDS)
buffer.append(ch);
input.move(1);
} else {
input.move(m.end[0]);
}
if (!input.isValid()) break;
}
return buffer.toString();
}
/* Helper function for constructor */
private void addToken(REToken next) {
if (next == null) return;
minimumLength += next.getMinimumLength();
if (firstToken == null) {
lastToken = firstToken = next;
} else {
// if chain returns false, it "rejected" the token due to
// an optimization, and next was combined with lastToken
if (lastToken.chain(next)) {
lastToken = next;
}
}
}
private static REToken setRepeated(REToken current, int min, int max, int index) throws REException {
if (current == null) throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,index);
return new RETokenRepeated(current.subIndex,current,min,max);
}
private static int getPosixSet(char[] pattern,int index,StringBuffer buf) {
// Precondition: pattern[index-1] == ':'
// we will return pos of closing ']'.
int i;
for (i=index; i<(pattern.length-1); i++) {
if ((pattern[i] == ':') && (pattern[i+1] == ']'))
return i+2;
buf.append(pattern[i]);
}
return index; // didn't match up
}
private int getMinMax(char[] input,int index,IntPair minMax,RESyntax syntax) throws REException {
// Precondition: input[index-1] == '{', minMax != null
boolean mustMatch = !syntax.get(RESyntax.RE_NO_BK_BRACES);
int startIndex = index;
if (index == input.length) {
if (mustMatch)
throw new REException(getLocalizedMessage("unmatched.brace"),REException.REG_EBRACE,index);
else
return startIndex;
}
int min,max=0;
CharUnit unit = new CharUnit();
StringBuffer buf = new StringBuffer();
// Read string of digits
do {
index = getCharUnit(input,index,unit);
if (Character.isDigit(unit.ch))
buf.append(unit.ch);
} while ((index != input.length) && Character.isDigit(unit.ch));
// Check for {} tomfoolery
if (buf.length() == 0) {
if (mustMatch)
throw new REException(getLocalizedMessage("interval.error"),REException.REG_EBRACE,index);
else
return startIndex;
}
min = Integer.parseInt(buf.toString());
if ((unit.ch == '}') && (syntax.get(RESyntax.RE_NO_BK_BRACES) ^ unit.bk))
max = min;
else if (index == input.length)
if (mustMatch)
throw new REException(getLocalizedMessage("interval.no.end"),REException.REG_EBRACE,index);
else
return startIndex;
else if ((unit.ch == ',') && !unit.bk) {
buf = new StringBuffer();
// Read string of digits
while (((index = getCharUnit(input,index,unit)) != input.length) && Character.isDigit(unit.ch))
buf.append(unit.ch);
if (!((unit.ch == '}') && (syntax.get(RESyntax.RE_NO_BK_BRACES) ^ unit.bk)))
if (mustMatch)
throw new REException(getLocalizedMessage("interval.error"),REException.REG_EBRACE,index);
else
return startIndex;
// This is the case of {x,}
if (buf.length() == 0) max = Integer.MAX_VALUE;
else max = Integer.parseInt(buf.toString());
} else
if (mustMatch)
throw new REException(getLocalizedMessage("interval.error"),REException.REG_EBRACE,index);
else
return startIndex;
// We know min and max now, and they are valid.
minMax.first = min;
minMax.second = max;
// return the index following the '}'
return index;
}
/**
* Return a human readable form of the compiled regular expression,
* useful for debugging.
*/
public String toString() {
StringBuffer sb = new StringBuffer();
dump(sb);
return sb.toString();
}
void dump(StringBuffer os) {
os.append('(');
if (subIndex == 0)
os.append("?:");
if (firstToken != null)
firstToken.dumpAll(os);
os.append(')');
}
// Cast input appropriately or throw exception
private static CharIndexed makeCharIndexed(Object input, int index) {
// We could let a String fall through to final input, but since
// it's the most likely input type, we check it first.
if (input instanceof String)
return new CharIndexedString((String) input,index);
else if (input instanceof char[])
return new CharIndexedCharArray((char[]) input,index);
else if (input instanceof StringBuffer)
return new CharIndexedStringBuffer((StringBuffer) input,index);
else if (input instanceof InputStream)
return new CharIndexedInputStream((InputStream) input,index);
else if (input instanceof Reader)
return new CharIndexedReader((Reader) input, index);
else if (input instanceof CharIndexed)
return (CharIndexed) input; // do we lose index info?
else
return new CharIndexedString(input.toString(), index);
}
}
/*
* gnu/regexp/REException.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/**
* This is the regular expression exception class. An exception of this type
* defines the three attributes:
*
* RE exp = new RE("^foo.");
* REMatch m0 = exp.getMatch(s);
* REMatch m1 = exp.getMatch(s.substring(8));
* REMatch m2 = exp.getMatch(s.substring(8),0,RE.REG_NOTBOL);
* REMatch m3 = exp.getMatch(s,8);
* REMatch m4 = exp.getMatch(s,8,RE.REG_ANCHORINDEX);
*
* // m0 = "food"
* // m1 = "fool"
* // m2 = null
* // m3 = null
* // m4 = "fool"
*
*
* REMatch myMatch = myExpression.getMatch(myString);
*/
public String toString() {
return matchedText;
}
/**
* Returns the index within the input text where the match in its entirety
* began.
*/
public int getStartIndex() {
return offset + start[0];
}
/**
* Returns the index within the input string where the match in
* its entirety ends. The return value is the next position after
* the end of the string; therefore, a match created by the
* following call:
*
*
* if (myMatch != null) System.out.println("Regexp found: "+myMatch);
*
* REMatch myMatch = myExpression.getMatch(myString);
*
* can be viewed (given that myMatch is not null) by creating *
* String theMatch = myString.substring(myMatch.getStartIndex(),
* myMatch.getEndIndex());
*
* But you can save yourself that work, since the toString()
* method (above) does exactly that for you.
*/
public int getEndIndex() {
return offset + end[0];
}
/**
* Returns the string matching the given subexpression. The subexpressions
* are indexed starting with one, not zero. That is, the subexpression
* identified by the first set of parentheses in a regular expression
* could be retrieved from an REMatch by calling match.toString(1).
*
* @param sub Index of the subexpression.
*/
public String toString(int sub) {
if ((sub >= start.length) || (start[sub] == -1)) return "";
return (matchedText.substring(start[sub],end[sub]));
}
/**
* Returns the index within the input string used to generate this match
* where subexpression number sub begins, or -1 if
* the subexpression does not exist. The initial position is zero.
*
* @param sub Subexpression index
* @deprecated Use getStartIndex(int) instead.
*/
public int getSubStartIndex(int sub) {
if (sub >= start.length) return -1;
int x = start[sub];
return (x == -1) ? x : offset + x;
}
/**
* Returns the index within the input string used to generate this match
* where subexpression number sub begins, or -1 if
* the subexpression does not exist. The initial position is zero.
*
* @param sub Subexpression index
* @since gnu.regexp 1.1.0
*/
public int getStartIndex(int sub) {
if (sub >= start.length) return -1;
int x = start[sub];
return (x == -1) ? x : offset + x;
}
/**
* Returns the index within the input string used to generate this match
* where subexpression number sub ends, or -1 if
* the subexpression does not exist. The initial position is zero.
*
* @param sub Subexpression index
* @deprecated Use getEndIndex(int) instead
*/
public int getSubEndIndex(int sub) {
if (sub >= start.length) return -1;
int x = end[sub];
return (x == -1) ? x : offset + x;
}
/**
* Returns the index within the input string used to generate this match
* where subexpression number sub ends, or -1 if
* the subexpression does not exist. The initial position is zero.
*
* @param sub Subexpression index
*/
public int getEndIndex(int sub) {
if (sub >= start.length) return -1;
int x = end[sub];
return (x == -1) ? x : offset + x;
}
/**
* Substitute the results of this match to create a new string.
* This is patterned after PERL, so the tokens to watch out for are
* $0 through $9. $0 matches
* the full substring matched; $n matches
* subexpression number n.
*
* @param input A string consisting of literals and $n tokens.
*/
public String substituteInto(String input) {
// a la Perl, $0 is whole thing, $1 - $9 are subexpressions
StringBuffer output = new StringBuffer();
int pos;
for (pos = 0; pos < input.length()-1; pos++) {
if ((input.charAt(pos) == '$') && (Character.isDigit(input.charAt(pos+1)))) {
int val = Character.digit(input.charAt(++pos),10);
if (val < start.length) {
output.append(toString(val));
}
} else output.append(input.charAt(pos));
}
if (pos < input.length()) output.append(input.charAt(pos));
return output.toString();
}
}
/*
* gnu/regexp/REMatchEnumeration.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/**
* An REMatchEnumeration enumerates regular expression matches over a
* given input text. You obtain a reference to an enumeration using
* the getMatchEnumeration() methods on an instance of
* RE.
*
*
*
* REMatchEnumeration does lazy computation; that is, it will not
* search for a match until it needs to. If you'd rather just get all
* the matches at once in a big array, use the
* getAllMatches() methods on RE. However, using an
* enumeration can help speed performance when the entire text does
* not need to be searched immediately.
*
*
*
* The enumerated type is especially useful when searching on a Reader
* or InputStream, because the InputStream read position cannot be
* guaranteed after calling
* Note that this makes UncheckedRE dangerous if constructed with
* dynamic data. Do not use UncheckedRE unless you are completely sure
* that all input being passed to it contains valid, well-formed
* regular expressions for the syntax specified.
*
* @author Wes Biggs
* @see gnu.regexp.RE
* @since gnu.regexp 1.1.4
*/
public final static class UncheckedRE extends RE {
/**
* Constructs a regular expression pattern buffer without any compilation
* flags set, and using the default syntax (RESyntax.RE_SYNTAX_PERL5).
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer or char[]. Other input types will be converted to
* strings using the toString() method.
* @exception RuntimeException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public UncheckedRE(Object pattern) {
this(pattern,0,RESyntax.RE_SYNTAX_PERL5);
}
/**
* Constructs a regular expression pattern buffer using the specified
* compilation flags and the default syntax (RESyntax.RE_SYNTAX_PERL5).
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer, or char[]. Other input types will be converted to
* strings using the toString() method.
* @param cflags The logical OR of any combination of the compilation flags in the RE class.
* @exception RuntimeException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public UncheckedRE(Object pattern, int cflags) {
this(pattern,cflags,RESyntax.RE_SYNTAX_PERL5);
}
/**
* Constructs a regular expression pattern buffer using the specified
* compilation flags and regular expression syntax.
*
* @param pattern A regular expression pattern, in the form of a String,
* StringBuffer, or char[]. Other input types will be converted to
* strings using the toString() method.
* @param cflags The logical OR of any combination of the compilation flags in the RE class.
* @param syntax The type of regular expression syntax to use.
* @exception RuntimeException The input pattern could not be parsed.
* @exception NullPointerException The pattern was null.
*/
public UncheckedRE(Object pattern, int cflags, RESyntax syntax) {
try {
initialize(pattern,cflags,syntax,0,0);
} catch (REException e) {
throw new RuntimeException(e.getMessage());
}
}
}
}
getMatch() (see the
* description of that method for an explanation of why). Enumeration
* also saves a lot of overhead required when calling
* getMatch() multiple times.
*
* @author Wes Biggs
*/
public static class REMatchEnumeration implements Enumeration, Serializable {
private static final int YES = 1;
private static final int MAYBE = 0;
private static final int NO = -1;
private int more;
private REMatch match;
private RE expr;
private CharIndexed input;
private int eflags;
private int index;
// Package scope constructor is used by RE.getMatchEnumeration()
REMatchEnumeration(RE expr, CharIndexed input, int index, int eflags) {
more = MAYBE;
this.expr = expr;
this.input = input;
this.index = index;
this.eflags = eflags;
}
/** Returns true if there are more matches in the input text. */
public boolean hasMoreElements() {
return hasMoreMatches(null);
}
/** Returns true if there are more matches in the input text. */
public boolean hasMoreMatches() {
return hasMoreMatches(null);
}
/** Returns true if there are more matches in the input text.
* Saves the text leading up to the match (or to the end of the input)
* in the specified buffer.
*/
public boolean hasMoreMatches(StringBuffer buffer) {
if (more == MAYBE) {
match = expr.getMatchImpl(input,index,eflags,buffer);
if (match != null) {
input.move((match.end[0] > 0) ? match.end[0] : 1);
index = (match.end[0] > 0) ? match.end[0] + match.offset : index + 1;
more = YES;
} else more = NO;
}
return (more == YES);
}
/** Returns the next match in the input text. */
public Object nextElement() throws NoSuchElementException {
return nextMatch();
}
/**
* Returns the next match in the input text. This method is provided
* for convenience to avoid having to explicitly cast the return value
* to class REMatch.
*/
public REMatch nextMatch() throws NoSuchElementException {
if (hasMoreElements()) {
more = (input.isValid()) ? MAYBE : NO;
return match;
}
throw new NoSuchElementException();
}
}
/*
* gnu/regexp/RESyntax.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/**
* An RESyntax specifies the way a regular expression will be compiled.
* This class provides a number of predefined useful constants for
* emulating popular regular expression syntaxes. Additionally the
* user may construct his or her own syntax, using any combination of the
* syntax bit constants. The syntax is an optional argument to any of the
* matching methods on class RE.
*
* @author Wes Biggs
*/
public static final class RESyntax implements Serializable {
static final String DEFAULT_LINE_SEPARATOR = System.getProperty("line.separator");
private static final String SYNTAX_IS_FINAL = RE.getLocalizedMessage("syntax.final");
private BitSet bits;
// true for the constant defined syntaxes
private boolean isFinal = false;
private String lineSeparator = DEFAULT_LINE_SEPARATOR;
// Values for constants are bit indexes
/**
* Syntax bit. Backslash is an escape character in lists.
*/
public static final int RE_BACKSLASH_ESCAPE_IN_LISTS = 0;
/**
* Syntax bit. Use \? instead of ? and \+ instead of +.
*/
public static final int RE_BK_PLUS_QM = 1;
/**
* Syntax bit. POSIX character classes ([:...:]) in lists are allowed.
*/
public static final int RE_CHAR_CLASSES = 2;
/**
* Syntax bit. ^ and $ are special everywhere.
* Not implemented.
*/
public static final int RE_CONTEXT_INDEP_ANCHORS = 3;
/**
* Syntax bit. Repetition operators are only special in valid positions.
* Not implemented.
*/
public static final int RE_CONTEXT_INDEP_OPS = 4;
/**
* Syntax bit. Repetition and alternation operators are invalid
* at start and end of pattern and other places.
* Not implemented.
*/
public static final int RE_CONTEXT_INVALID_OPS = 5;
/**
* Syntax bit. Match-any-character operator (.) matches a newline.
*/
public static final int RE_DOT_NEWLINE = 6;
/**
* Syntax bit. Match-any-character operator (.) does not match a null.
*/
public static final int RE_DOT_NOT_NULL = 7;
/**
* Syntax bit. Intervals ({x}, {x,}, {x,y}) are allowed.
*/
public static final int RE_INTERVALS = 8;
/**
* Syntax bit. No alternation (|), match one-or-more (+), or
* match zero-or-one (?) operators.
*/
public static final int RE_LIMITED_OPS = 9;
/**
* Syntax bit. Newline is an alternation operator.
*/
public static final int RE_NEWLINE_ALT = 10; // impl.
/**
* Syntax bit. Intervals use { } instead of \{ \}
*/
public static final int RE_NO_BK_BRACES = 11;
/**
* Syntax bit. Grouping uses ( ) instead of \( \).
*/
public static final int RE_NO_BK_PARENS = 12;
/**
* Syntax bit. Backreferences not allowed.
*/
public static final int RE_NO_BK_REFS = 13;
/**
* Syntax bit. Alternation uses | instead of \|
*/
public static final int RE_NO_BK_VBAR = 14;
/**
* Syntax bit. Not implemented.
*/
public static final int RE_NO_EMPTY_RANGES = 15;
/**
* Syntax bit. An unmatched right parenthesis (')' or '\)', depending
* on RE_NO_BK_PARENS) will throw an exception when compiling.
*/
public static final int RE_UNMATCHED_RIGHT_PAREN_ORD = 16;
/**
* Syntax bit. Not implemented.
*/
public static final int RE_HAT_LISTS_NOT_NEWLINE = 17;
/**
* Syntax bit. Stingy matching is allowed (+?, *?, ??, {x,y}?).
*/
public static final int RE_STINGY_OPS = 18;
/**
* Syntax bit. Allow character class escapes (\d, \D, \s, \S, \w, \W).
*/
public static final int RE_CHAR_CLASS_ESCAPES = 19;
/**
* Syntax bit. Allow use of (?:xxx) grouping (subexpression is not saved).
*/
public static final int RE_PURE_GROUPING = 20;
/**
* Syntax bit. Allow use of (?=xxx) and (?!xxx) apply the subexpression
* to the text following the current position without consuming that text.
*/
public static final int RE_LOOKAHEAD = 21;
/**
* Syntax bit. Allow beginning- and end-of-string anchors (\A, \Z).
*/
public static final int RE_STRING_ANCHORS = 22;
/**
* Syntax bit. Allow embedded comments, (?#comment), as in Perl5.
*/
public static final int RE_COMMENTS = 23;
/**
* Syntax bit. Allow character class escapes within lists, as in Perl5.
*/
public static final int RE_CHAR_CLASS_ESC_IN_LISTS = 24;
private static final int BIT_TOTAL = 25;
/**
* Predefined syntax.
* Emulates regular expression support in the awk utility.
*/
public static final RESyntax RE_SYNTAX_AWK;
/**
* Predefined syntax.
* Emulates regular expression support in the ed utility.
*/
public static final RESyntax RE_SYNTAX_ED;
/**
* Predefined syntax.
* Emulates regular expression support in the egrep utility.
*/
public static final RESyntax RE_SYNTAX_EGREP;
/**
* Predefined syntax.
* Emulates regular expression support in the GNU Emacs editor.
*/
public static final RESyntax RE_SYNTAX_EMACS;
/**
* Predefined syntax.
* Emulates regular expression support in the grep utility.
*/
public static final RESyntax RE_SYNTAX_GREP;
/**
* Predefined syntax.
* Emulates regular expression support in the POSIX awk specification.
*/
public static final RESyntax RE_SYNTAX_POSIX_AWK;
/**
* Predefined syntax.
* Emulates POSIX basic regular expression support.
*/
public static final RESyntax RE_SYNTAX_POSIX_BASIC;
/**
* Predefined syntax.
* Emulates regular expression support in the POSIX egrep specification.
*/
public static final RESyntax RE_SYNTAX_POSIX_EGREP;
/**
* Predefined syntax.
* Emulates POSIX extended regular expression support.
*/
public static final RESyntax RE_SYNTAX_POSIX_EXTENDED;
/**
* Predefined syntax.
* Emulates POSIX basic minimal regular expressions.
*/
public static final RESyntax RE_SYNTAX_POSIX_MINIMAL_BASIC;
/**
* Predefined syntax.
* Emulates POSIX extended minimal regular expressions.
*/
public static final RESyntax RE_SYNTAX_POSIX_MINIMAL_EXTENDED;
/**
* Predefined syntax.
* Emulates regular expression support in the sed utility.
*/
public static final RESyntax RE_SYNTAX_SED;
/**
* Predefined syntax.
* Emulates regular expression support in Larry Wall's perl, version 4,
*/
public static final RESyntax RE_SYNTAX_PERL4;
/**
* Predefined syntax.
* Emulates regular expression support in Larry Wall's perl, version 4,
* using single line mode (/s modifier).
*/
public static final RESyntax RE_SYNTAX_PERL4_S; // single line mode (/s)
/**
* Predefined syntax.
* Emulates regular expression support in Larry Wall's perl, version 5.
*/
public static final RESyntax RE_SYNTAX_PERL5;
/**
* Predefined syntax.
* Emulates regular expression support in Larry Wall's perl, version 5,
* using single line mode (/s modifier).
*/
public static final RESyntax RE_SYNTAX_PERL5_S;
static {
// Define syntaxes
RE_SYNTAX_EMACS = new RESyntax().makeFinal();
RESyntax RE_SYNTAX_POSIX_COMMON = new RESyntax()
.set(RE_CHAR_CLASSES)
.set(RE_DOT_NEWLINE)
.set(RE_DOT_NOT_NULL)
.set(RE_INTERVALS)
.set(RE_NO_EMPTY_RANGES)
.makeFinal();
RE_SYNTAX_POSIX_BASIC = new RESyntax(RE_SYNTAX_POSIX_COMMON)
.set(RE_BK_PLUS_QM)
.makeFinal();
RE_SYNTAX_POSIX_EXTENDED = new RESyntax(RE_SYNTAX_POSIX_COMMON)
.set(RE_CONTEXT_INDEP_ANCHORS)
.set(RE_CONTEXT_INDEP_OPS)
.set(RE_NO_BK_BRACES)
.set(RE_NO_BK_PARENS)
.set(RE_NO_BK_VBAR)
.set(RE_UNMATCHED_RIGHT_PAREN_ORD)
.makeFinal();
RE_SYNTAX_AWK = new RESyntax()
.set(RE_BACKSLASH_ESCAPE_IN_LISTS)
.set(RE_DOT_NOT_NULL)
.set(RE_NO_BK_PARENS)
.set(RE_NO_BK_REFS)
.set(RE_NO_BK_VBAR)
.set(RE_NO_EMPTY_RANGES)
.set(RE_UNMATCHED_RIGHT_PAREN_ORD)
.makeFinal();
RE_SYNTAX_POSIX_AWK = new RESyntax(RE_SYNTAX_POSIX_EXTENDED)
.set(RE_BACKSLASH_ESCAPE_IN_LISTS)
.makeFinal();
RE_SYNTAX_GREP = new RESyntax()
.set(RE_BK_PLUS_QM)
.set(RE_CHAR_CLASSES)
.set(RE_HAT_LISTS_NOT_NEWLINE)
.set(RE_INTERVALS)
.set(RE_NEWLINE_ALT)
.makeFinal();
RE_SYNTAX_EGREP = new RESyntax()
.set(RE_CHAR_CLASSES)
.set(RE_CONTEXT_INDEP_ANCHORS)
.set(RE_CONTEXT_INDEP_OPS)
.set(RE_HAT_LISTS_NOT_NEWLINE)
.set(RE_NEWLINE_ALT)
.set(RE_NO_BK_PARENS)
.set(RE_NO_BK_VBAR)
.makeFinal();
RE_SYNTAX_POSIX_EGREP = new RESyntax(RE_SYNTAX_EGREP)
.set(RE_INTERVALS)
.set(RE_NO_BK_BRACES)
.makeFinal();
/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */
RE_SYNTAX_ED = new RESyntax(RE_SYNTAX_POSIX_BASIC)
.makeFinal();
RE_SYNTAX_SED = new RESyntax(RE_SYNTAX_POSIX_BASIC)
.makeFinal();
RE_SYNTAX_POSIX_MINIMAL_BASIC = new RESyntax(RE_SYNTAX_POSIX_COMMON)
.set(RE_LIMITED_OPS)
.makeFinal();
/* Differs from RE_SYNTAX_POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */
RE_SYNTAX_POSIX_MINIMAL_EXTENDED = new RESyntax(RE_SYNTAX_POSIX_COMMON)
.set(RE_CONTEXT_INDEP_ANCHORS)
.set(RE_CONTEXT_INVALID_OPS)
.set(RE_NO_BK_BRACES)
.set(RE_NO_BK_PARENS)
.set(RE_NO_BK_REFS)
.set(RE_NO_BK_VBAR)
.set(RE_UNMATCHED_RIGHT_PAREN_ORD)
.makeFinal();
/* There is no official Perl spec, but here's a "best guess" */
RE_SYNTAX_PERL4 = new RESyntax()
.set(RE_BACKSLASH_ESCAPE_IN_LISTS)
.set(RE_CONTEXT_INDEP_ANCHORS)
.set(RE_CONTEXT_INDEP_OPS) // except for '{', apparently
.set(RE_INTERVALS)
.set(RE_NO_BK_BRACES)
.set(RE_NO_BK_PARENS)
.set(RE_NO_BK_VBAR)
.set(RE_NO_EMPTY_RANGES)
.set(RE_CHAR_CLASS_ESCAPES) // \d,\D,\w,\W,\s,\S
.makeFinal();
RE_SYNTAX_PERL4_S = new RESyntax(RE_SYNTAX_PERL4)
.set(RE_DOT_NEWLINE)
.makeFinal();
RE_SYNTAX_PERL5 = new RESyntax(RE_SYNTAX_PERL4)
.set(RE_PURE_GROUPING) // (?:)
.set(RE_STINGY_OPS) // *?,??,+?,{}?
.set(RE_LOOKAHEAD) // (?=)(?!)
.set(RE_STRING_ANCHORS) // \A,\Z
.set(RE_CHAR_CLASS_ESC_IN_LISTS)// \d,\D,\w,\W,\s,\S within []
.set(RE_COMMENTS) // (?#)
.makeFinal();
RE_SYNTAX_PERL5_S = new RESyntax(RE_SYNTAX_PERL5)
.set(RE_DOT_NEWLINE)
.makeFinal();
}
/**
* Construct a new syntax object with all bits turned off.
* This is equivalent to RE_SYNTAX_EMACS.
*/
public RESyntax() {
bits = new BitSet(BIT_TOTAL);
}
/**
* Called internally when constructing predefined syntaxes
* so their interpretation cannot vary. Conceivably useful
* for your syntaxes as well. Causes IllegalAccessError to
* be thrown if any attempt to modify the syntax is made.
*
* @return this object for convenient chaining
*/
public RESyntax makeFinal() {
isFinal = true;
return this;
}
/**
* Construct a new syntax object with all bits set the same
* as the other syntax.
*/
public RESyntax(RESyntax other) {
bits = (BitSet) other.bits.clone();
}
/**
* Check if a given bit is set in this syntax.
*/
public boolean get(int index) {
return bits.get(index);
}
/**
* Set a given bit in this syntax.
*
* @param index the constant (RESyntax.RE_xxx) bit to set.
* @return a reference to this object for easy chaining.
*/
public RESyntax set(int index) {
if (isFinal) throw new IllegalAccessError(SYNTAX_IS_FINAL);
bits.set(index);
return this;
}
/**
* Clear a given bit in this syntax.
*
* @param index the constant (RESyntax.RE_xxx) bit to clear.
* @return a reference to this object for easy chaining.
*/
public RESyntax clear(int index) {
if (isFinal) throw new IllegalAccessError(SYNTAX_IS_FINAL);
bits.clear(index);
return this;
}
/**
* Changes the line separator string for regular expressions
* created using this RESyntax. The default separator is the
* value returned by the system property "line.separator", which
* should be correct when reading platform-specific files from a
* filesystem. However, many programs may collect input from
* sources where the line separator is differently specified (for
* example, in the applet environment, the text box widget
* interprets line breaks as single-character newlines,
* regardless of the host platform.
*
* Note that setting the line separator to a character or
* characters that have specific meaning within the current syntax
* can cause unexpected chronosynclastic infundibula.
*
* @return this object for convenient chaining
*/
public RESyntax setLineSeparator(String aSeparator) {
if (isFinal) throw new IllegalAccessError(SYNTAX_IS_FINAL);
lineSeparator = aSeparator;
return this;
}
/**
* Returns the currently active line separator string. The default
* is the platform-dependent system property "line.separator".
*/
public String getLineSeparator() {
return lineSeparator;
}
}
/*
* gnu/regexp/REToken.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
abstract static class REToken implements Serializable {
protected REToken next = null;
protected REToken uncle = null;
protected int subIndex;
protected REToken(int subIndex) {
this.subIndex = subIndex;
}
int getMinimumLength() {
return 0;
}
void setUncle(REToken anUncle) {
uncle = anUncle;
}
/** Returns true if the match succeeded, false if it failed. */
abstract boolean match(CharIndexed input, REMatch mymatch);
/** Returns true if the rest of the tokens match, false if they fail. */
protected boolean next(CharIndexed input, REMatch mymatch) {
if (next == null) {
if (uncle == null) {
return true;
} else {
return uncle.match(input, mymatch);
}
} else {
return next.match(input, mymatch);
}
}
boolean chain(REToken token) {
next = token;
return true; // Token was accepted
}
abstract void dump(StringBuffer os);
void dumpAll(StringBuffer os) {
dump(os);
if (next != null) next.dumpAll(os);
}
}
/*
* gnu/regexp/RETokenAny.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
final static class RETokenAny extends REToken {
/** True if '.' can match a newline (RE_DOT_NEWLINE) */
private boolean newline;
/** True if '.' can't match a null (RE_DOT_NOT_NULL) */
private boolean matchNull;
RETokenAny(int subIndex, boolean newline, boolean matchNull) {
super(subIndex);
this.newline = newline;
this.matchNull = matchNull;
}
int getMinimumLength() {
return 1;
}
boolean match(CharIndexed input, REMatch mymatch) {
char ch = input.charAt(mymatch.index);
if ((ch == CharIndexed.OUT_OF_BOUNDS)
|| (!newline && (ch == '\n'))
|| (matchNull && (ch == 0))) {
return false;
}
++mymatch.index;
return next(input, mymatch);
}
void dump(StringBuffer os) {
os.append('.');
}
}
/*
* gnu/regexp/RETokenBackRef.java
* Copyright (C) 1998-2001 Wes Biggs
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
final static class RETokenBackRef extends REToken {
private int num;
private boolean insens;
RETokenBackRef(int subIndex, int num, boolean insens) {
super(subIndex);
this.num = num;
this.insens = insens;
}
// should implement getMinimumLength() -- any ideas?
boolean match(CharIndexed input, REMatch mymatch) {
int b,e;
b = mymatch.start[num];
e = mymatch.end[num];
if ((b==-1)||(e==-1)) return false; // this shouldn't happen, but...
for (int i=b; i