yggdrasil/gcc
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Imported GNU Classpath 0.90

Browse Source 
Imported GNU Classpath 0.90
       * scripts/makemake.tcl: Set gnu/java/awt/peer/swing to ignore.
       * gnu/classpath/jdwp/VMFrame.java (SIZE): New constant.
       * java/lang/VMCompiler.java: Use gnu.java.security.hash.MD5.
       * java/lang/Math.java: New override file.
       * java/lang/Character.java: Merged from Classpath.
       (start, end): Now 'int's.
       (canonicalName): New field.
       (CANONICAL_NAME, NO_SPACES_NAME, CONSTANT_NAME): New constants.
       (UnicodeBlock): Added argument.
       (of): New overload.
       (forName): New method.
       Updated unicode blocks.
       (sets): Updated.
       * sources.am: Regenerated.
       * Makefile.in: Likewise.

From-SVN: r111942
This commit is contained in:
Mark Wielaard
2006-03-10 21:46:48 +00:00
parent 27079765d0
commit 8aa540d2f7
1367 changed files with 188789 additions and 22762 deletions
Download Patch File Download Diff File Expand all files Collapse all files
libjava/classpath/gnu/regexp/RE.java
+656 -122
View File
@@ -1,5 +1,5 @@
/* gnu/regexp/RE.java
Copyright (C) 1998-2001, 2004, 2005 Free Software Foundation, Inc.
Copyright (C) 2006 Free Software Foundation, Inc.
This file is part of GNU Classpath.
@@ -136,12 +136,13 @@ public class RE extends REToken {
/** Minimum length, in characters, of any possible match. */
private int minimumLength;
private int maximumLength;
/**
* Compilation flag. Do not differentiate case. Subsequent
* searches using this RE will be case insensitive.
*/
public static final int REG_ICASE = 2;
public static final int REG_ICASE = 0x02;
/**
* Compilation flag. The match-any-character operator (dot)
@@ -149,14 +150,14 @@ public class RE extends REToken {
* 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;
public static final int REG_DOT_NEWLINE = 0x04;
/**
* 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;
public static final int REG_MULTILINE = 0x08;
/**
* Execution flag.
@@ -185,14 +186,14 @@ public class RE extends REToken {
* // m4.toString(): "fool"<BR>
* </CODE>
*/
public static final int REG_NOTBOL = 16;
public static final int REG_NOTBOL = 0x10;
/**
* 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;
public static final int REG_NOTEOL = 0x20;
/**
* Execution flag.
@@ -206,7 +207,7 @@ public class RE extends REToken {
* the example under REG_NOTBOL. It also affects the use of the \&lt;
* and \b operators.
*/
public static final int REG_ANCHORINDEX = 64;
public static final int REG_ANCHORINDEX = 0x40;
/**
* Execution flag.
@@ -215,7 +216,24 @@ public class RE extends REToken {
* the corresponding subexpressions. For example, you may want to
* replace all matches of "one dollar" with "$1".
*/
public static final int REG_NO_INTERPOLATE = 128;
public static final int REG_NO_INTERPOLATE = 0x80;
/**
* Execution flag.
* Try to match the whole input string. An implicit match-end operator
* is added to this regexp.
*/
public static final int REG_TRY_ENTIRE_MATCH = 0x0100;
/**
* Execution flag.
* The substitute and substituteAll methods will treat the
* character '\' in the replacement as an escape to a literal
* character. In this case "\n", "\$", "\\", "\x40" and "\012"
* will become "n", "$", "\", "x40" and "012" respectively.
* This flag has no effect if REG_NO_INTERPOLATE is set on.
*/
public static final int REG_REPLACE_USE_BACKSLASHESCAPE = 0x0200;
/** Returns a string representing the version of the gnu.regexp package. */
public static final String version() {
@@ -273,12 +291,13 @@ public class RE extends REToken {
}
// internal constructor used for alternation
private RE(REToken first, REToken last,int subs, int subIndex, int minLength) {
private RE(REToken first, REToken last,int subs, int subIndex, int minLength, int maxLength) {
super(subIndex);
firstToken = first;
lastToken = last;
numSubs = subs;
minimumLength = minLength;
maximumLength = maxLength;
addToken(new RETokenEndSub(subIndex));
}
@@ -333,6 +352,11 @@ public class RE extends REToken {
char ch;
boolean quot = false;
// Saved syntax and flags.
RESyntax savedSyntax = null;
int savedCflags = 0;
boolean flagsSaved = false;
while (index < pLength) {
// read the next character unit (including backslash escapes)
index = getCharUnit(pattern,index,unit,quot);
@@ -359,8 +383,9 @@ public class RE extends REToken {
&& !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);
RE theBranch = new RE(firstToken, lastToken, numSubs, subIndex, minimumLength, maximumLength);
minimumLength = 0;
maximumLength = 0;
if (branches == null) {
branches = new Vector();
}
@@ -402,102 +427,12 @@ public class RE extends REToken {
// [...] | [^...]
else if ((unit.ch == '[') && !(unit.bk || quot)) {
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
ParseCharClassResult result = parseCharClass(
subIndex, pattern, index, pLength, cflags, syntax, 0);
addToken(currentToken);
options.trimToSize();
currentToken = new RETokenOneOf(subIndex,options,negative);
currentToken = result.token;
index = result.index;
}
// SUBEXPRESSIONS
@@ -507,7 +442,10 @@ public class RE extends REToken {
boolean pure = false;
boolean comment = false;
boolean lookAhead = false;
boolean lookBehind = false;
boolean independent = false;
boolean negativelh = false;
boolean negativelb = false;
if ((index+1 < pLength) && (pattern[index] == '?')) {
switch (pattern[index+1]) {
case '!':
@@ -525,6 +463,114 @@ public class RE extends REToken {
index += 2;
}
break;
case '<':
// We assume that if the syntax supports look-ahead,
// it also supports look-behind.
if (syntax.get(RESyntax.RE_LOOKAHEAD)) {
index++;
switch (pattern[index +1]) {
case '!':
pure = true;
negativelb = true;
lookBehind = true;
index += 2;
break;
case '=':
pure = true;
lookBehind = true;
index += 2;
}
}
break;
case '>':
// We assume that if the syntax supports look-ahead,
// it also supports independent group.
if (syntax.get(RESyntax.RE_LOOKAHEAD)) {
pure = true;
independent = true;
index += 2;
}
break;
case 'i':
case 'd':
case 'm':
case 's':
// case 'u': not supported
// case 'x': not supported
case '-':
if (!syntax.get(RESyntax.RE_EMBEDDED_FLAGS)) break;
// Set or reset syntax flags.
int flagIndex = index + 1;
int endFlag = -1;
RESyntax newSyntax = new RESyntax(syntax);
int newCflags = cflags;
boolean negate = false;
while (flagIndex < pLength && endFlag < 0) {
switch(pattern[flagIndex]) {
case 'i':
if (negate)
newCflags &= ~REG_ICASE;
else
newCflags |= REG_ICASE;
flagIndex++;
break;
case 'd':
if (negate)
newSyntax.setLineSeparator(RESyntax.DEFAULT_LINE_SEPARATOR);
else
newSyntax.setLineSeparator("\n");
flagIndex++;
break;
case 'm':
if (negate)
newCflags &= ~REG_MULTILINE;
else
newCflags |= REG_MULTILINE;
flagIndex++;
break;
case 's':
if (negate)
newCflags &= ~REG_DOT_NEWLINE;
else
newCflags |= REG_DOT_NEWLINE;
flagIndex++;
break;
// case 'u': not supported
// case 'x': not supported
case '-':
negate = true;
flagIndex++;
break;
case ':':
case ')':
endFlag = pattern[flagIndex];
break;
default:
throw new REException(getLocalizedMessage("repeat.no.token"), REException.REG_BADRPT, index);
}
}
if (endFlag == ')') {
syntax = newSyntax;
cflags = newCflags;
insens = ((cflags & REG_ICASE) > 0);
// This can be treated as though it were a comment.
comment = true;
index = flagIndex - 1;
break;
}
if (endFlag == ':') {
savedSyntax = syntax;
savedCflags = cflags;
flagsSaved = true;
syntax = newSyntax;
cflags = newCflags;
insens = ((cflags & REG_ICASE) > 0);
index = flagIndex -1;
// Fall through to the next case.
}
else {
throw new REException(getLocalizedMessage("unmatched.paren"), REException.REG_ESUBREG,index);
}
case ':':
if (syntax.get(RESyntax.RE_PURE_GROUPING)) {
pure = true;
@@ -607,15 +653,28 @@ public class RE extends REToken {
numSubs++;
}
int useIndex = (pure || lookAhead) ? 0 : nextSub + numSubs;
int useIndex = (pure || lookAhead || lookBehind || independent) ?
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);
}
else if (lookBehind) {
currentToken = new RETokenLookBehind(currentToken,negativelb);
}
else if (independent) {
currentToken = new RETokenIndependent(currentToken);
}
index = nextIndex;
if (flagsSaved) {
syntax = savedSyntax;
cflags = savedCflags;
insens = ((cflags & REG_ICASE) > 0);
flagsSaved = false;
}
} // not a comment
} // subexpression
@@ -715,14 +774,45 @@ public class RE extends REToken {
else
currentToken = setRepeated(currentToken,0,1,index);
}
// OCTAL CHARACTER
// \0377
else if (unit.bk && (unit.ch == '0') && syntax.get(RESyntax.RE_OCTAL_CHAR)) {
CharExpression ce = getCharExpression(pattern, index - 2, pLength, syntax);
if (ce == null)
throw new REException("invalid octal character", REException.REG_ESCAPE, index);
index = index - 2 + ce.len;
addToken(currentToken);
currentToken = new RETokenChar(subIndex,ce.ch,insens);
}
// BACKREFERENCE OPERATOR
// \1 \2 ... \9
// \1 \2 ... \9 and \10 \11 \12 ...
// not available if RE_NO_BK_REFS is set
// Perl recognizes \10, \11, and so on only if enough number of
// parentheses have opened before it, otherwise they are treated
// as aliases of \010, \011, ... (octal characters). In case of
// Sun's JDK, octal character expression must always begin with \0.
// We will do as JDK does. But FIXME, take a look at "(a)(b)\29".
// JDK treats \2 as a back reference to the 2nd group because
// there are only two groups. But in our poor implementation,
// we cannot help but treat \29 as a back reference to the 29th group.
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);
int numBegin = index - 1;
int numEnd = pLength;
for (int i = index; i < pLength; i++) {
if (! Character.isDigit(pattern[i])) {
numEnd = i;
break;
}
}
int num = parseInt(pattern, numBegin, numEnd-numBegin, 10);
currentToken = new RETokenBackRef(subIndex,num,insens);
index = numEnd;
}
// START OF STRING OPERATOR
@@ -844,6 +934,32 @@ public class RE extends REToken {
currentToken = new RETokenEnd(subIndex,null);
}
// HEX CHARACTER, UNICODE CHARACTER
// \x1B, \u1234
else if ((unit.bk && (unit.ch == 'x') && syntax.get(RESyntax.RE_HEX_CHAR)) ||
(unit.bk && (unit.ch == 'u') && syntax.get(RESyntax.RE_UNICODE_CHAR))) {
CharExpression ce = getCharExpression(pattern, index - 2, pLength, syntax);
if (ce == null)
throw new REException("invalid hex character", REException.REG_ESCAPE, index);
index = index - 2 + ce.len;
addToken(currentToken);
currentToken = new RETokenChar(subIndex,ce.ch,insens);
}
// NAMED PROPERTY
// \p{prop}, \P{prop}
else if ((unit.bk && (unit.ch == 'p') && syntax.get(RESyntax.RE_NAMED_PROPERTY)) ||
(unit.bk && (unit.ch == 'P') && syntax.get(RESyntax.RE_NAMED_PROPERTY))) {
NamedProperty np = getNamedProperty(pattern, index - 2, pLength);
if (np == null)
throw new REException("invalid escape sequence", REException.REG_ESCAPE, index);
index = index - 2 + np.len;
addToken(currentToken);
currentToken = getRETokenNamedProperty(subIndex,np,insens,index);
}
// NON-SPECIAL CHARACTER (or escape to make literal)
// c | \* for example
@@ -857,9 +973,10 @@ public class RE extends REToken {
addToken(currentToken);
if (branches != null) {
branches.addElement(new RE(firstToken,lastToken,numSubs,subIndex,minimumLength));
branches.addElement(new RE(firstToken,lastToken,numSubs,subIndex,minimumLength, maximumLength));
branches.trimToSize(); // compact the Vector
minimumLength = 0;
maximumLength = 0;
firstToken = lastToken = null;
addToken(new RETokenOneOf(subIndex,branches,false));
}
@@ -867,6 +984,199 @@ public class RE extends REToken {
}
private static class ParseCharClassResult {
RETokenOneOf token;
int index;
boolean returnAtAndOperator = false;
}
/**
* Parse [...] or [^...] and make an RETokenOneOf instance.
* @param subIndex subIndex to be given to the created RETokenOneOf instance.
* @param pattern Input array of characters to be parsed.
* @param index Index pointing to the character next to the beginning '['.
* @param pLength Limit of the input array.
* @param cflags Compilation flags used to parse the pattern.
* @param pflags Flags that affect the behavior of this method.
* @param syntax Syntax used to parse the pattern.
*/
private static ParseCharClassResult parseCharClass(int subIndex,
char[] pattern, int index,
int pLength, int cflags, RESyntax syntax, int pflags)
throws REException {
boolean insens = ((cflags & REG_ICASE) > 0);
Vector options = new Vector();
Vector addition = new Vector();
boolean additionAndAppeared = false;
final int RETURN_AT_AND = 0x01;
boolean returnAtAndOperator = ((pflags & RETURN_AT_AND) != 0);
boolean negative = false;
char ch;
char lastChar = 0;
boolean lastCharIsSet = false;
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; lastCharIsSet = true;
if (++index == pLength) throw new REException(getLocalizedMessage("class.no.end"),REException.REG_EBRACK,index);
}
while ((ch = pattern[index++]) != ']') {
if ((ch == '-') && (lastCharIsSet)) {
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 {
if ((ch == '\\') && syntax.get(RESyntax.RE_BACKSLASH_ESCAPE_IN_LISTS)) {
CharExpression ce = getCharExpression(pattern, index, pLength, syntax);
if (ce == null)
throw new REException("invalid escape sequence", REException.REG_ESCAPE, index);
ch = ce.ch;
index = index + ce.len - 1;
}
options.addElement(new RETokenRange(subIndex,lastChar,ch,insens));
lastChar = 0; lastCharIsSet = false;
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;
boolean asciiEscIsSet = false;
NamedProperty np = null;
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;
}
}
if (("pP".indexOf(pattern[index]) != -1) && syntax.get(RESyntax.RE_NAMED_PROPERTY)) {
np = getNamedProperty(pattern, index - 1, pLength);
if (np == null)
throw new REException("invalid escape sequence", REException.REG_ESCAPE, index);
index = index - 1 + np.len - 1;
}
else {
CharExpression ce = getCharExpression(pattern, index - 1, pLength, syntax);
if (ce == null)
throw new REException("invalid escape sequence", REException.REG_ESCAPE, index);
asciiEsc = ce.ch; asciiEscIsSet = true;
index = index - 1 + ce.len - 1;
}
if (lastCharIsSet) options.addElement(new RETokenChar(subIndex,lastChar,insens));
if (posixID != -1) {
options.addElement(new RETokenPOSIX(subIndex,posixID,insens,negate));
} else if (np != null) {
options.addElement(getRETokenNamedProperty(subIndex,np,insens,index));
} else if (asciiEscIsSet) {
lastChar = asciiEsc; lastCharIsSet = true;
} else {
lastChar = pattern[index]; lastCharIsSet = true;
}
++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 ((ch == '[') && (syntax.get(RESyntax.RE_NESTED_CHARCLASS))) {
ParseCharClassResult result = parseCharClass(
subIndex, pattern, index, pLength, cflags, syntax, 0);
addition.addElement(result.token);
addition.addElement("|");
index = result.index;
} else if ((ch == '&') &&
(syntax.get(RESyntax.RE_NESTED_CHARCLASS)) &&
(index < pLength) && (pattern[index] == '&')) {
if (returnAtAndOperator) {
ParseCharClassResult result = new ParseCharClassResult();
options.trimToSize();
if (additionAndAppeared) addition.addElement("&");
if (addition.size() == 0) addition = null;
result.token = new RETokenOneOf(subIndex,
options, addition, negative);
result.index = index - 1;
result.returnAtAndOperator = true;
return result;
}
// The precedence of the operator "&&" is the lowest.
// So we postpone adding "&" until other elements
// are added. And we insert Boolean.FALSE at the
// beginning of the list of tokens following "&&".
// So, "&&[a-b][k-m]" will be stored in the Vecter
// addition in this order:
// Boolean.FALSE, [a-b], "|", [k-m], "|", "&"
if (additionAndAppeared) addition.addElement("&");
addition.addElement(Boolean.FALSE);
additionAndAppeared = true;
// The part on which "&&" operates may be either
// (1) explicitly enclosed by []
// or
// (2) not enclosed by [] and terminated by the
// next "&&" or the end of the character list.
// Let the preceding else if block do the case (1).
// We must do something in case of (2).
if ((index + 1 < pLength) && (pattern[index + 1] != '[')) {
ParseCharClassResult result = parseCharClass(
subIndex, pattern, index+1, pLength, cflags, syntax,
RETURN_AT_AND);
addition.addElement(result.token);
addition.addElement("|");
// If the method returned at the next "&&", it is OK.
// Otherwise we have eaten the mark of the end of this
// character list "]". In this case we must give back
// the end mark.
index = (result.returnAtAndOperator ?
result.index: result.index - 1);
}
} else {
if (lastCharIsSet) options.addElement(new RETokenChar(subIndex,lastChar,insens));
lastChar = ch; lastCharIsSet = true;
}
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 (lastCharIsSet) options.addElement(new RETokenChar(subIndex,lastChar,insens));
ParseCharClassResult result = new ParseCharClassResult();
// Create a new RETokenOneOf
options.trimToSize();
if (additionAndAppeared) addition.addElement("&");
if (addition.size() == 0) addition = null;
result.token = new RETokenOneOf(subIndex,options, addition, negative);
result.index = index;
return result;
}
private static int getCharUnit(char[] input, int index, CharUnit unit, boolean quot) throws REException {
unit.ch = input[index++];
unit.bk = (unit.ch == '\\'
@@ -878,6 +1188,176 @@ public class RE extends REToken {
return index;
}
private static int parseInt(char[] input, int pos, int len, int radix) {
int ret = 0;
for (int i = pos; i < pos + len; i++) {
ret = ret * radix + Character.digit(input[i], radix);
}
return ret;
}
/**
* This class represents various expressions for a character.
* "a" : 'a' itself.
* "\0123" : Octal char 0123
* "\x1b" : Hex char 0x1b
* "\u1234" : Unicode char \u1234
*/
private static class CharExpression {
/** character represented by this expression */
char ch;
/** String expression */
String expr;
/** length of this expression */
int len;
public String toString() { return expr; }
}
private static CharExpression getCharExpression(char[] input, int pos, int lim,
RESyntax syntax) {
CharExpression ce = new CharExpression();
char c = input[pos];
if (c == '\\') {
if (pos + 1 >= lim) return null;
c = input[pos + 1];
switch(c) {
case 't':
ce.ch = '\t';
ce.len = 2;
break;
case 'n':
ce.ch = '\n';
ce.len = 2;
break;
case 'r':
ce.ch = '\r';
ce.len = 2;
break;
case 'x':
case 'u':
if ((c == 'x' && syntax.get(RESyntax.RE_HEX_CHAR)) ||
(c == 'u' && syntax.get(RESyntax.RE_UNICODE_CHAR))) {
int l = 0;
int expectedLength = (c == 'x' ? 2 : 4);
for (int i = pos + 2; i < pos + 2 + expectedLength; i++) {
if (i >= lim) break;
if (!((input[i] >= '0' && input[i] <= '9') ||
(input[i] >= 'A' && input[i] <= 'F') ||
(input[i] >= 'a' && input[i] <= 'f')))
break;
l++;
}
if (l != expectedLength) return null;
ce.ch = (char)(parseInt(input, pos + 2, l, 16));
ce.len = l + 2;
}
else {
ce.ch = c;
ce.len = 2;
}
break;
case '0':
if (syntax.get(RESyntax.RE_OCTAL_CHAR)) {
int l = 0;
for (int i = pos + 2; i < pos + 2 + 3; i++) {
if (i >= lim) break;
if (input[i] < '0' || input[i] > '7') break;
l++;
}
if (l == 3 && input[pos + 2] > '3') l--;
if (l <= 0) return null;
ce.ch = (char)(parseInt(input, pos + 2, l, 8));
ce.len = l + 2;
}
else {
ce.ch = c;
ce.len = 2;
}
break;
default:
ce.ch = c;
ce.len = 2;
break;
}
}
else {
ce.ch = input[pos];
ce.len = 1;
}
ce.expr = new String(input, pos, ce.len);
return ce;
}
/**
* This class represents a substring in a pattern string expressing
* a named property.
* "\pA" : Property named "A"
* "\p{prop}" : Property named "prop"
* "\PA" : Property named "A" (Negated)
* "\P{prop}" : Property named "prop" (Negated)
*/
private static class NamedProperty {
/** Property name */
String name;
/** Negated or not */
boolean negate;
/** length of this expression */
int len;
}
private static NamedProperty getNamedProperty(char[] input, int pos, int lim) {
NamedProperty np = new NamedProperty();
char c = input[pos];
if (c == '\\') {
if (++pos >= lim) return null;
c = input[pos++];
switch(c) {
case 'p':
np.negate = false;
break;
case 'P':
np.negate = true;
break;
default:
return null;
}
c = input[pos++];
if (c == '{') {
int p = -1;
for (int i = pos; i < lim; i++) {
if (input[i] == '}') {
p = i;
break;
}
}
if (p < 0) return null;
int len = p - pos;
np.name = new String(input, pos, len);
np.len = len + 4;
}
else {
np.name = new String(input, pos - 1, 1);
np.len = 3;
}
return np;
}
else return null;
}
private static RETokenNamedProperty getRETokenNamedProperty(
int subIndex, NamedProperty np, boolean insens, int index)
throws REException {
try {
return new RETokenNamedProperty(subIndex, np.name, insens, np.negate);
}
catch (REException e) {
REException ree;
ree = new REException(e.getMessage(), REException.REG_ESCAPE, index);
ree.initCause(e);
throw ree;
}
}
/**
* Checks if the regular expression matches the input in its entirety.
*
@@ -958,6 +1438,10 @@ public class RE extends REToken {
return minimumLength;
}
public int getMaximumLength() {
return maximumLength;
}
/**
* Returns an array of all matches found in the input.
*
@@ -1025,7 +1509,9 @@ public class RE extends REToken {
/* Implements abstract method REToken.match() */
boolean match(CharIndexed input, REMatch mymatch) {
if (firstToken == null) return next(input, mymatch);
if (firstToken == null) {
return next(input, mymatch);
}
// Note the start of this subexpression
mymatch.start[subIndex] = mymatch.index;
@@ -1089,23 +1575,34 @@ public class RE extends REToken {
}
REMatch getMatchImpl(CharIndexed input, int anchor, int eflags, StringBuffer buffer) {
boolean tryEntireMatch = ((eflags & REG_TRY_ENTIRE_MATCH) != 0);
RE re = (tryEntireMatch ? (RE) this.clone() : this);
if (tryEntireMatch) {
re.chain(new RETokenEnd(0, null));
}
// 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;
if (re.match(input, mymatch)) {
REMatch best = mymatch;
// We assume that the match that coms first is the best.
// And the following "The longer, the better" rule has
// been commented out. The longest is not neccesarily
// the best. For example, "a" out of "aaa" is the best
// match for /a+?/.
/*
// Find best match of them all to observe leftmost longest
while ((mymatch = mymatch.next) != null) {
if (mymatch.index > longest.index) {
longest = mymatch;
if (mymatch.index > best.index) {
best = mymatch;
}
}
longest.end[0] = longest.index;
longest.finish(input);
return longest;
*/
best.end[0] = best.index;
best.finish(input);
return best;
}
}
mymatch.clear(++anchor);
@@ -1216,8 +1713,7 @@ public class RE extends REToken {
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) );
buffer.append(getReplacement(replace, m, eflags));
if (input.move(m.end[0])) {
do {
buffer.append(input.charAt(0));
@@ -1278,8 +1774,7 @@ public class RE extends REToken {
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) );
buffer.append(getReplacement(replace, m, eflags));
index = m.getEndIndex();
if (m.end[0] == 0) {
char ch = input.charAt(0);
@@ -1294,11 +1789,50 @@ public class RE extends REToken {
}
return buffer.toString();
}
public static String getReplacement(String replace, REMatch m, int eflags) {
if ((eflags & REG_NO_INTERPOLATE) > 0)
return replace;
else {
if ((eflags & REG_REPLACE_USE_BACKSLASHESCAPE) > 0) {
StringBuffer sb = new StringBuffer();
int l = replace.length();
for (int i = 0; i < l; i++) {
char c = replace.charAt(i);
switch(c) {
case '\\':
i++;
// Let StringIndexOutOfBoundsException be thrown.
sb.append(replace.charAt(i));
break;
case '$':
int i1 = i + 1;
while (i1 < replace.length() &&
Character.isDigit(replace.charAt(i1))) i1++;
sb.append(m.substituteInto(replace.substring(i, i1)));
i = i1 - 1;
break;
default:
sb.append(c);
}
}
return sb.toString();
}
else
return m.substituteInto(replace);
}
}
/* Helper function for constructor */
private void addToken(REToken next) {
if (next == null) return;
minimumLength += next.getMinimumLength();
int nmax = next.getMaximumLength();
if (nmax < Integer.MAX_VALUE && maximumLength < Integer.MAX_VALUE)
maximumLength += nmax;
else
maximumLength = Integer.MAX_VALUE;
if (firstToken == null) {
lastToken = firstToken = next;
} else {