regexodus.Term Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of jtransc-rt Show documentation
Show all versions of jtransc-rt Show documentation
JVM AOT compiler currently generating JavaScript, C++, Haxe, with initial focus on Kotlin and games.
/**
* Copyright (c) 2001, Sergey A. Samokhodkin
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without modification,
* are permitted provided that the following conditions are met:
*
* - Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
* - Redistributions in binary form
* must reproduce the above copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials provided with the distribution.
* - Neither the name of jregex nor the names of its contributors may be used
* to endorse or promote products derived from this software without specific prior
* written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
* INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
* OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY
* WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* @version 1.2_01
*/
package regexodus;
import com.jtransc.annotation.JTranscInvisible;
import regexodus.ds.IntBitSet;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashMap;
@JTranscInvisible
public class Term implements REFlags, Serializable {
private static final long serialVersionUID = 2528136757932720807L;
//runtime Term types
static final int CHAR = 0;
static final int BITSET = 1;
static final int BITSET2 = 2;
static final int ANY_CHAR = 4;
static final int ANY_CHAR_NE = 5;
static final int REG = 6;
static final int REG_I = 7;
static final int FIND = 8;
static final int FINDREG = 9;
static final int SUCCESS = 10;
/*optimization-transparent types*/
static final int BOUNDARY = 11;
static final int DIRECTION = 12;
static final int UBOUNDARY = 13;
static final int UDIRECTION = 14;
static final int GROUP_IN = 15;
static final int GROUP_OUT = 16;
static final int VOID = 17;
static final int START = 18;
static final int END = 19;
static final int END_EOL = 20;
static final int LINE_START = 21;
static final int LINE_END = 22;
static final int LAST_MATCH_END = 23;
static final int CNT_SET_0 = 24;
static final int CNT_INC = 25;
static final int CNT_GT_EQ = 26;
static final int READ_CNT_LT = 27;
static final int CRSTORE_CRINC = 28; //store on 'actual' search entry
static final int CR_SET_0 = 29;
static final int CR_LT = 30;
static final int CR_GT_EQ = 31;
static final int LITERAL_START = 60;
static final int LITERAL_END = 61;
/*optimization-nontransparent types*/
static final int BRANCH = 32;
static final int BRANCH_STORE_CNT = 33;
static final int BRANCH_STORE_CNT_AUX1 = 34;
static final int PLOOKAHEAD_IN = 35;
static final int PLOOKAHEAD_OUT = 36;
static final int NLOOKAHEAD_IN = 37;
static final int NLOOKAHEAD_OUT = 38;
static final int PLOOKBEHIND_IN = 39;
static final int PLOOKBEHIND_OUT = 40;
static final int NLOOKBEHIND_IN = 41;
static final int NLOOKBEHIND_OUT = 42;
static final int INDEPENDENT_IN = 43; //functionally the same as NLOOKAHEAD_IN
static final int INDEPENDENT_OUT = 44;
static final int REPEAT_0_INF = 45;
static final int REPEAT_MIN_INF = 46;
static final int REPEAT_MIN_MAX = 47;
static final int REPEAT_REG_MIN_INF = 48;
static final int REPEAT_REG_MIN_MAX = 49;
static final int BACKTRACK_0 = 50;
static final int BACKTRACK_MIN = 51;
static final int BACKTRACK_FIND_MIN = 52;
static final int BACKTRACK_FINDREG_MIN = 53;
static final int BACKTRACK_REG_MIN = 54;
static final int MEMREG_CONDITION = 55;
static final int LOOKAHEAD_CONDITION_IN = 56;
static final int LOOKAHEAD_CONDITION_OUT = 57;
static final int LOOKBEHIND_CONDITION_IN = 58;
static final int LOOKBEHIND_CONDITION_OUT = 59;
//optimization
static final int FIRST_TRANSPARENT = BOUNDARY;
static final int LAST_TRANSPARENT = CR_GT_EQ;
// compile-time: length of vars[] (see makeTree())
private static final int VARS_LENGTH = 4;
// compile-time variable indices:
private static final int MEMREG_COUNT = 0; //refers current memreg index
private static final int CNTREG_COUNT = 1; //refers current counters number
private static final int DEPTH = 2; //refers current depth: (((depth=3)))
private static final int LOOKAHEAD_COUNT = 3; //refers current memreg index
private static final int LIMITS_LENGTH = 3;
private static final int LIMITS_PARSE_RESULT_INDEX = 2;
private static final int LIMITS_OK = 1;
private static final int LIMITS_FAILURE = 2;
private static final int LITERAL_FLAG = 64;
//static CustomParser[] customParsers=new CustomParser[256];
// **** CONTROL FLOW ****
// next-to-execute and next-if-failed commands;
Term next, failNext;
// **** TYPES ****
int type = VOID;
boolean inverse;
// used with type=CHAR
char c;
// used with type=FIND
int distance;
boolean eat;
// used with type=BITSET(2);
IntBitSet bitset;
IntBitSet[] bitset2;
private boolean[] categoryBitset; //types(unicode categories)
boolean mode_insensitive;
boolean mode_reverse;
boolean mode_bracket;
// used for optimization with type=BITSET,BITSET2
int weight;
// **** MEMORISATION ****
// memory slot, used with type=REG,GROUP_IN,GROUP_OUT
int memreg = -1;
// **** COUNTERS ****
// max|min number of iterations
// used with CNT_GT_EQ ,REPEAT_* etc.;
int minCount, maxCount;
// used with REPEAT_*,REPEAT_REG_*;
Term target;
// a counter slot to increment & compare with maxCount (CNT_INC etc.);
int cntreg = 0;
// lookahead group id;
int lookaheadId;
// **** COMPILE HELPERS ****
Term prev;
Term in;
Term out;
Term out1;
protected Term first;
Term current;
//new!!
Term branchOut;
//protected boolean newBranch=false,closed=false;
//protected boolean newBranch=false;
//for debugging
private static int instances;
private int instanceNum;
Term() {
//for debugging
instanceNum = instances;
instances++;
in = out = this;
}
Term(int type) {
this();
this.type = type;
}
static void makeTree(String s, int[] flags, Pattern re) throws PatternSyntaxException {
instances = 0;
char[] data = s.toCharArray();
makeTree(data, 0, data.length, flags, re);
}
private static void makeTree(char[] data, int offset, int end,
int[] flags, Pattern re) throws PatternSyntaxException {
// memreg,counter,depth,lookahead
int[] vars = {1, 0, 0, 0}; //don't use counters[0]
//collect iterators for subsequent optimization
ArrayList iterators = new ArrayList();
HashMap groupNames = new HashMap();
Pretokenizer t = new Pretokenizer(data, offset, end);
Term term = makeTree(t, data, vars, flags, new Group(), iterators, groupNames);
// convert closing outer bracket into success term
term.out.type = SUCCESS;
//throw out opening bracket
Term first = term.next;
// Optimisation:
Term optimized = first;
Optimizer opt = Optimizer.find(first);
if (opt != null) optimized = opt.makeFirst(first);
for (TermIterator i : iterators) {
i.optimize();
}
re.root = optimized;
re.root = first;
re.root0 = first;
re.memregs = vars[MEMREG_COUNT];
re.counters = vars[CNTREG_COUNT];
re.lookaheads = vars[LOOKAHEAD_COUNT];
re.namedGroupMap = groupNames;
}
private static Term makeTree(Pretokenizer t, char[] data, int[] vars,
int[] flags, Term term, ArrayList iterators, HashMap groupNames) throws PatternSyntaxException {
if (vars.length != VARS_LENGTH)
throw new IllegalArgumentException("vars.length should be " + VARS_LENGTH + ", not " + vars.length);
//Term term=new Term(isMemReg? vars[MEMREG_COUNT]: -1);
// use memreg 0 as insignificant
//Term term=new Group(isMemReg? vars[MEMREG_COUNT]: 0);
while (true) {
t.next();
term.append(t.tOffset, t.tOutside, data, vars, flags, iterators, groupNames);
switch (t.ttype) {
case Pretokenizer.FLAGS:
flags[0] = t.flags(flags[0]);
continue;
case Pretokenizer.CLASS_GROUP:
t.next();
Term clg = new Term();
CharacterClass.parseGroup(data, t.tOffset, t.tOutside, clg,
(flags[0] & IGNORE_CASE) > 0, (flags[0] & IGNORE_SPACES) > 0,
(flags[0] & UNICODE) > 0, (flags[0] & XML_SCHEMA) > 0);
term.append(clg);
continue;
case Pretokenizer.PLAIN_GROUP:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, new int[]{t.flags(flags[0])}, new Group(), iterators, groupNames));
break;
case Pretokenizer.NAMED_GROUP:
String gname = t.groupName;
int id;
if (Character.isDigit(gname.charAt(0))) {
try {
id = Integer.parseInt(gname);
} catch (NumberFormatException e) {
throw new PatternSyntaxException("group name starts with digit but is not a number");
}
if (groupNames.containsValue(id)) {
if (t.groupDeclared)
throw new PatternSyntaxException("group redeclaration: " + gname + "; use ({=id}...) for multiple group assignments");
}
if (vars[MEMREG_COUNT] <= id) vars[MEMREG_COUNT] = id + 1;
} else {
Integer no = groupNames.get(gname);
if (no == null) {
id = vars[MEMREG_COUNT]++;
groupNames.put(t.groupName, id);
} else {
if (t.groupDeclared)
throw new PatternSyntaxException("group redeclaration " + gname + "; use ({=name}...) for group reassignments");
id = no;
}
}
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Group(id), iterators, groupNames));
break;
case '(':
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Group(vars[MEMREG_COUNT]++), iterators, groupNames));
break;
case Pretokenizer.POS_LOOKAHEAD:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Lookahead(vars[LOOKAHEAD_COUNT]++, true), iterators, groupNames));
break;
case Pretokenizer.NEG_LOOKAHEAD:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Lookahead(vars[LOOKAHEAD_COUNT]++, false), iterators, groupNames));
break;
case Pretokenizer.POS_LOOKBEHIND:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Lookbehind(vars[LOOKAHEAD_COUNT]++, true), iterators, groupNames));
break;
case Pretokenizer.NEG_LOOKBEHIND:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new Lookbehind(vars[LOOKAHEAD_COUNT]++, false), iterators, groupNames));
break;
case Pretokenizer.INDEPENDENT_REGEX:
vars[DEPTH]++;
term.append(makeTree(t, data, vars, flags, new IndependentGroup(vars[LOOKAHEAD_COUNT]++), iterators, groupNames));
break;
case Pretokenizer.CONDITIONAL_GROUP:
vars[DEPTH]++;
t.next();
Term fork;
boolean positive = true;
switch (t.ttype) {
case Pretokenizer.NEG_LOOKAHEAD:
positive = false;
case Pretokenizer.POS_LOOKAHEAD:
vars[DEPTH]++;
Lookahead la = new Lookahead(vars[LOOKAHEAD_COUNT]++, positive);
makeTree(t, data, vars, flags, la, iterators, groupNames);
fork = new ConditionalExpr(la);
break;
case Pretokenizer.NEG_LOOKBEHIND:
positive = false;
case Pretokenizer.POS_LOOKBEHIND:
vars[DEPTH]++;
Lookbehind lb = new Lookbehind(vars[LOOKAHEAD_COUNT]++, positive);
makeTree(t, data, vars, flags, lb, iterators, groupNames);
fork = new ConditionalExpr(lb);
break;
case '(':
t.next();
if (t.ttype != ')') throw new PatternSyntaxException("malformed condition");
int memregNo;
if (Character.isDigit(data[t.tOffset])) memregNo = makeNumber(t.tOffset, t.tOutside, data);
else {
String gn = new String(data, t.tOffset, t.tOutside - t.tOffset);
Integer gno = groupNames.get(gn);
if (gno == null)
throw new PatternSyntaxException("unknown group name in conditional expr.: " + gn);
memregNo = gno;
}
fork = new ConditionalExpr(memregNo);
break;
default:
throw new PatternSyntaxException("malformed conditional expression: " + t.ttype + " '" + (char) t.ttype + "'");
}
term.append(makeTree(t, data, vars, flags, fork, iterators, groupNames));
break;
case '|':
term.newBranch();
break;
case Pretokenizer.END:
if (vars[DEPTH] > 0) throw new PatternSyntaxException("unbalanced parenthesis");
term.close();
return term;
case ')':
if (vars[DEPTH] <= 0) throw new PatternSyntaxException("unbalanced parenthesis");
term.close();
vars[DEPTH]--;
return term;
case Pretokenizer.COMMENT:
while (t.ttype != ')') t.next();
continue;
default:
throw new PatternSyntaxException("unknown token type: " + t.ttype);
}
}
}
private static int makeNumber(int off, int out, char[] data) {
int n = 0;
for (int i = off; i < out; i++) {
int d = data[i] - '0';
if (d < 0 || d > 9) return -1;
n *= 10;
n += d;
}
return n;
}
private void append(int offset, int end, char[] data,
int[] vars, int[] flags, ArrayList iterators, HashMap gmap) throws PatternSyntaxException {
int[] limits = new int[3];
int i = offset;
Term tmp, current = this.current;
while (i < end) {
char c = data[i];
boolean greedy = true;
if((flags[0] & LITERAL_FLAG) != LITERAL_FLAG) {
switch (c) {
//operations
case '*':
if (current == null) throw new PatternSyntaxException("missing term before *");
i++;
if (i < end && data[i] == '?') {
greedy = false;
i++;
}
tmp = greedy ? makeGreedyStar(vars, current, iterators) :
makeLazyStar(vars, current);
current = replaceCurrent(tmp);
break;
case '+':
if (current == null) throw new PatternSyntaxException("missing term before +");
i++;
if (i < end && data[i] == '?') {
greedy = false;
i++;
}
tmp = greedy ? makeGreedyPlus(vars, current, iterators) :
makeLazyPlus(vars, current);
current = replaceCurrent(tmp);
break;
case '?':
if (current == null) throw new PatternSyntaxException("missing term before ?");
i++;
if (i < end && data[i] == '?') {
greedy = false;
i++;
}
tmp = greedy ? makeGreedyQMark(vars, current) :
makeLazyQMark(vars, current);
current = replaceCurrent(tmp);
break;
case '{':
limits[0] = 0;
limits[1] = -1;
int le = parseLimits(i + 1, end, data, limits);
if (limits[LIMITS_PARSE_RESULT_INDEX] == LIMITS_OK) { //parse ok
if (current == null) throw new PatternSyntaxException("missing term before {}");
i = le;
if (i < end && data[i] == '?') {
greedy = false;
i++;
}
tmp = greedy ? makeGreedyLimits(vars, current, limits, iterators) :
makeLazyLimits(vars, current, limits);
current = replaceCurrent(tmp);
break;
} else { //unicode class or named backreference
if (data[i + 1] == '\\') { //'{\name}' - backreference
int p = i + 2;
if (p == end) throw new PatternSyntaxException("'group_id' expected");
char cp = data[p];
boolean mi = false, mb = false, mr = false;
while (Category.Z.contains(cp) || Category.Po.contains(cp)) {
p++;
if (p == end) throw new PatternSyntaxException("'group_id' expected");
switch (cp) {
case '@':
mi = !mi;
break;
case '/':
mr = !mr;
break;
case ':':
mb = !mb;
break;
}
cp = data[p];
}
BackReference br = new BackReference(-1, mi || (flags[0] & IGNORE_CASE) > 0, mr, mb);
i = parseGroupId(data, p, end, br, gmap, '}');
current = append(br);
continue;
} else {
Term t = new Term();
i = CharacterClass.parseName(data, i, end, t, false, (flags[0] & IGNORE_SPACES) > 0);
current = append(t);
continue;
}
}
case '\\':
if (i + 4 < end && data[i + 1] == 'k' && data[i + 2] == '<') { //'\k' - backreference
int p = i + 3;
if (p == end) throw new PatternSyntaxException("'group_id' expected");
char cp = data[p];
boolean mi = false, mb = false, mr = false;
while (Category.Z.contains(cp) || Category.Po.contains(cp)) {
p++;
if (p == end) throw new PatternSyntaxException("'group_id' expected");
switch (cp) {
case '@':
mi = !mi;
break;
case '/':
mr = !mr;
break;
case ':':
mb = !mb;
break;
}
cp = data[p];
}
BackReference br = new BackReference(-1, mi || (flags[0] & IGNORE_CASE) > 0, mr, mb);
i = parseGroupId(data, p, end, br, gmap, '>');
current = append(br);
continue;
}
case ' ':
case '\t':
case '\r':
case '\n':
if ((flags[0] & IGNORE_SPACES) > 0) {
i++;
continue;
}
//else go on as default
//symbolic items
default:
tmp = new Term();
i = parseTerm(data, i, end, tmp, flags[0]);
if (tmp.type == LITERAL_START) {
flags[0] |= LITERAL_FLAG;
continue;
} else if (tmp.type == LITERAL_END) {
flags[0] &= ~LITERAL_FLAG;
continue;
}
if (tmp.type == END && i < end) {
throw new PatternSyntaxException("'$' is not a last term in the group: <" + new String(data, offset, end - offset) + ">");
}
//"\A"
//if(tmp.type==START && i>(offset+1)){
// throw new PatternSyntaxException("'^' is not a first term in the group: <"+new String(data,offset,end-offset)+">");
//}
current = append(tmp);
break;
}
}
else {
tmp = new Term();
i = parseTerm(data, i, end, tmp, flags[0]);
if (tmp.type == LITERAL_START) {
flags[0] |= LITERAL_FLAG;
continue;
} else if (tmp.type == LITERAL_END) {
flags[0] &= ~LITERAL_FLAG;
continue;
}
if (tmp.type == END && i < end) {
throw new PatternSyntaxException("'$' is not a last term in the group: <" + new String(data, offset, end - offset) + ">");
}
current = append(tmp);
}
}
}
/*
static boolean isIdentifierPart()
{
}*/
private static int parseGroupId(char[] data, int i, int end, Term term, HashMap gmap, char closer) throws PatternSyntaxException {
int id;
int nstart = i;
if (Character.isDigit(data[i])) {
while (Character.isDigit(data[i])) {
i++;
if (i == end) throw new PatternSyntaxException("group_id expected");
}
id = makeNumber(nstart, i, data);
} else {
while (Category.IdentifierPart.contains(data[i])) {
i++;
if (i == end) throw new PatternSyntaxException("group_id expected");
}
String s = new String(data, nstart, i - nstart);
Integer no = gmap.get(s);
if (no == null) throw new PatternSyntaxException("backreference to unknown group: " + s);
id = no;
}
while (Category.Z.contains(data[i])) {
i++;
if (i == end) throw new PatternSyntaxException("'" + closer + "' expected");
}
int c = data[i++];
if (c != closer) throw new PatternSyntaxException("'" + closer + "' expected");
term.memreg = id;
return i;
}
Term append(Term term) throws PatternSyntaxException {
//Term prev=this.prev;
Term current = this.current;
if (current == null) {
in.next = term;
term.prev = in;
this.current = term;
return term;
}
link(current, term);
//this.prev=current;
this.current = term;
return term;
}
Term replaceCurrent(Term term) throws PatternSyntaxException {
//Term prev=this.prev;
Term prev = current.prev;
if (prev != null) {
Term in = this.in;
if (prev == in) {
//in.next=term;
//term.prev=in;
in.next = term.in;
term.in.prev = in;
} else link(prev, term);
}
this.current = term;
return term;
}
private void newBranch() throws PatternSyntaxException {
close();
startNewBranch();
}
void close() throws PatternSyntaxException {
/*
Term prev=this.prev;
if(prev!=null){
Term current=this.current;
if(current!=null){
link(prev,current);
prev=current;
this.current=null;
}
link(prev,out);
this.prev=null;
}
*/
Term current = this.current;
if (current != null) linkd(current, out);
else in.next = out;
}
private static void link(Term term, Term next) {
linkd(term, next.in);
next.prev = term;
}
private static void linkd(Term term, Term next) {
Term prev_out = term.out;
if (prev_out != null) {
prev_out.next = next;
}
Term prev_out1 = term.out1;
if (prev_out1 != null) {
prev_out1.next = next;
}
Term prev_branch = term.branchOut;
if (prev_branch != null) {
prev_branch.failNext = next;
}
}
void startNewBranch() throws PatternSyntaxException {
Term tmp = in.next;
Term b = new Branch();
in.next = b;
b.next = tmp;
b.in = null;
b.out = null;
b.out1 = null;
b.branchOut = b;
current = b;
}
private static Term makeGreedyStar(int[] vars, Term term, ArrayList iterators) throws PatternSyntaxException {
//vars[STACK_SIZE]++;
switch (term.type) {
case GROUP_IN: {
Term b = new Branch();
b.next = term.in;
term.out.next = b;
b.in = b;
b.out = null;
b.out1 = null;
b.branchOut = b;
return b;
}
default: {
return new TermIterator(term, 0, -1, iterators);
}
}
}
private static Term makeLazyStar(int[] vars, Term term) {
//vars[STACK_SIZE]++;
switch (term.type) {
case GROUP_IN: {
Term b = new Branch();
b.failNext = term.in;
term.out.next = b;
b.in = b;
b.out = b;
b.out1 = null;
b.branchOut = null;
return b;
}
default: {
Term b = new Branch();
b.failNext = term;
term.next = b;
b.in = b;
b.out = b;
b.out1 = null;
b.branchOut = null;
return b;
}
}
}
private static Term makeGreedyPlus(int[] vars, Term term, ArrayList iterators) throws PatternSyntaxException {
//vars[STACK_SIZE]++;
switch (term.type) {
case INDEPENDENT_IN://?
case GROUP_IN: {
Term b = new Branch();
b.next = term.in;
term.out.next = b;
b.in = term.in;
b.out = null;
b.out1 = null;
b.branchOut = b;
return b;
}
default: {
return new TermIterator(term, 1, -1, iterators);
}
}
}
private static Term makeLazyPlus(int[] vars, Term term) {
//vars[STACK_SIZE]++;
switch (term.type) {
case GROUP_IN: {
Term b = new Branch();
term.out.next = b;
b.failNext = term.in;
b.in = term.in;
b.out = b;
b.out1 = null;
b.branchOut = null;
return b;
}
case REG:
default: {
Term b = new Branch();
term.next = b;
b.failNext = term;
b.in = term;
b.out = b;
b.out1 = null;
b.branchOut = null;
return b;
}
}
}
private static Term makeGreedyQMark(int[] vars, Term term) {
//vars[STACK_SIZE]++;
switch (term.type) {
case GROUP_IN: {
Term b = new Branch();
b.next = term.in;
b.in = b;
b.out = term.out;
b.out1 = null;
b.branchOut = b;
return b;
}
case REG:
default: {
Term b = new Branch();
b.next = term;
b.in = b;
b.out = term;
b.out1 = null;
b.branchOut = b;
return b;
}
}
}
private static Term makeLazyQMark(int[] vars, Term term) {
//vars[STACK_SIZE]++;
switch (term.type) {
case GROUP_IN: {
Term b = new Branch();
b.failNext = term.in;
b.in = b;
b.out = b;
b.out1 = term.out;
b.branchOut = null;
return b;
}
case REG:
default: {
Term b = new Branch();
b.failNext = term;
b.in = b;
b.out = b;
b.out1 = term;
b.branchOut = null;
return b;
}
}
}
private static Term makeGreedyLimits(int[] vars, Term term, int[] limits, ArrayList iterators) throws PatternSyntaxException {
//vars[STACK_SIZE]++;
int m = limits[0];
int n = limits[1];
switch (term.type) {
case GROUP_IN: {
int cntreg = vars[CNTREG_COUNT]++;
Term reset = new Term(CR_SET_0);
reset.cntreg = cntreg;
Term b = new Term(BRANCH);
Term inc = new Term(CRSTORE_CRINC);
inc.cntreg = cntreg;
reset.next = b;
if (n >= 0) {
Term lt = new Term(CR_LT);
lt.cntreg = cntreg;
lt.maxCount = n;
b.next = lt;
lt.next = term.in;
} else {
b.next = term.in;
}
term.out.next = inc;
inc.next = b;
if (m >= 0) {
Term gt = new Term(CR_GT_EQ);
gt.cntreg = cntreg;
gt.maxCount = m;
b.failNext = gt;
reset.in = reset;
reset.out = gt;
reset.out1 = null;
reset.branchOut = null;
} else {
reset.in = reset;
reset.out = null;
reset.out1 = null;
reset.branchOut = b;
}
return reset;
}
default: {
return new TermIterator(term, limits[0], limits[1], iterators);
}
}
}
private static Term makeLazyLimits(int[] vars, Term term, int[] limits) {
//vars[STACK_SIZE]++;
int m = limits[0];
int n = limits[1];
switch (term.type) {
case GROUP_IN: {
int cntreg = vars[CNTREG_COUNT]++;
Term reset = new Term(CR_SET_0);
reset.cntreg = cntreg;
Term b = new Term(BRANCH);
Term inc = new Term(CRSTORE_CRINC);
inc.cntreg = cntreg;
reset.next = b;
if (n >= 0) {
Term lt = new Term(CR_LT);
lt.cntreg = cntreg;
lt.maxCount = n;
b.failNext = lt;
lt.next = term.in;
} else {
b.failNext = term.in;
}
term.out.next = inc;
inc.next = b;
if (m >= 0) {
Term gt = new Term(CR_GT_EQ);
gt.cntreg = cntreg;
gt.maxCount = m;
b.next = gt;
reset.in = reset;
reset.out = gt;
reset.out1 = null;
reset.branchOut = null;
return reset;
} else {
reset.in = reset;
reset.out = b;
reset.out1 = null;
reset.branchOut = null;
return reset;
}
}
case REG:
default: {
Term reset = new Term(CNT_SET_0);
Term b = new Branch(BRANCH_STORE_CNT);
Term inc = new Term(CNT_INC);
reset.next = b;
if (n >= 0) {
Term lt = new Term(READ_CNT_LT);
lt.maxCount = n;
b.failNext = lt;
lt.next = term;
term.next = inc;
inc.next = b;
} else {
b.next = term;
term.next = inc;
inc.next = term;
}
if (m >= 0) {
Term gt = new Term(CNT_GT_EQ);
gt.maxCount = m;
b.next = gt;
reset.in = reset;
reset.out = gt;
reset.out1 = null;
reset.branchOut = null;
return reset;
} else {
reset.in = reset;
reset.out = b;
reset.out1 = null;
reset.branchOut = null;
return reset;
}
}
}
}
private int parseTerm(char[] data, int i, int out, Term term,
int flags) throws PatternSyntaxException {
char c = data[i++];
boolean inv = false;
if((flags & LITERAL_FLAG) == LITERAL_FLAG)
{
switch (c)
{
case '\\':
if(i < out + 1 && data[i] == 'E')
{
term.type = LITERAL_END;
return i + 1;
}
default:
term.type = CHAR;
if ((flags & IGNORE_CASE) == 0) {
term.c = c;
} else {
term.c = Category.caseFold(c);
}
return i;
}
}
switch (c) {
case '[':
return CharacterClass.parseClass(data, i, out, term, (flags & IGNORE_CASE) > 0, (flags & IGNORE_SPACES) > 0, (flags & UNICODE) > 0, (flags & XML_SCHEMA) > 0);
case '.':
term.type = (flags & DOTALL) > 0 ? ANY_CHAR : ANY_CHAR_NE;
break;
case '$':
//term.type=mods[MULTILINE_IND]? LINE_END: END; //??
term.type = (flags & MULTILINE) > 0 ? LINE_END : END_EOL;
break;
case '^':
term.type = (flags & MULTILINE) > 0 ? LINE_START : START;
break;
case '\\':
if (i >= out) throw new PatternSyntaxException("Escape without a character");
c = data[i++];
switch (c) {
case 'f':
c = '\f'; // form feed
break;
case 'n':
c = '\n'; // new line
break;
case 'r':
c = '\r'; // carriage return
break;
case 't':
c = '\t'; // tab
break;
case 'u':
if(i < out - 3)
c = (char) ((CharacterClass.toHexDigit(data[i++]) << 12) +
(CharacterClass.toHexDigit(data[i++]) << 8) +
(CharacterClass.toHexDigit(data[i++]) << 4) +
CharacterClass.toHexDigit(data[i++]));
else {
c = '\0';
i = out;
}
break;
case 'x': { // hex 2-digit number -> char
int hex = 0;
char d;
if ((d = data[i++]) == '{') {
while (i < out && (d = data[i++]) != '}') {
hex = (hex << 4) + CharacterClass.toHexDigit(d);
if (hex > 0xffff || i == out)
throw new PatternSyntaxException("\\x{}");
}
} else {
hex = (CharacterClass.toHexDigit(d) << 4) +
CharacterClass.toHexDigit(data[i++]);
}
c = (char) hex;
break;
}
case '0':
case 'o': // oct arbitrary-digit number -> char
int oct = 0;
for (; i < out; ) {
char d = data[i++];
if (d >= '0' && d <= '7') {
oct *= 8;
oct += d - '0';
if (oct > 0xffff) {
oct -= d - '0';
oct /= 8;
break;
}
} else break;
}
c = (char) oct;
break;
case 'm': // decimal number -> char
int dec = 0;
for (; i < out; ) {
char d = data[i++];
if (d >= '0' && d <= '9') {
dec *= 10;
dec += d - '0';
if (dec > 0xffff){
dec -= d - '0';
dec /= 10;
break;
}
} else break;
}
c = (char) dec;
break;
case 'c': // ctrl-char
c = (char) (data[i++] & 0x1f);
break;
case 'D': // non-digit
inv = true;
// go on
case 'd': // digit
CharacterClass.makeDigit(term, inv, (flags & UNICODE) > 0);
return i;
case 'S': // non-space
inv = true;
// go on
case 's': // space
CharacterClass.makeSpace(term, inv, (flags & UNICODE) > 0);
return i;
case 'W': // non-letter
inv = true;
// go on
case 'w': // letter
CharacterClass.makeWordChar(term, inv, (flags & UNICODE) > 0);
return i;
case 'H':
inv = true;
case 'h':
CharacterClass.makeHSpace(term, inv, (flags & UNICODE) > 0);
return i;
case 'V':
inv = true;
case 'v':
CharacterClass.makeVSpace(term, inv, (flags & UNICODE) > 0);
return i;
case 'B': // non-(word boundary)
inv = true;
// go on
case 'b': // word boundary
CharacterClass.makeWordBoundary(term, inv, (flags & UNICODE) > 0);
return i;
case '<': // word start
CharacterClass.makeWordStart(term, (flags & UNICODE) > 0);
return i;
case '>': // word end
CharacterClass.makeWordEnd(term, (flags & UNICODE) > 0);
return i;
case 'A': // text beginning
term.type = START;
return i;
case 'Z': // text end
term.type = END_EOL;
return i;
case 'z': // text end
term.type = END;
return i;
case 'G': // end of last match
term.type = LAST_MATCH_END;
return i;
case 'P': // \\P{..}
inv = true;
case 'p': // \\p{..}
i = CharacterClass.parseName(data, i, out, term, inv, (flags & IGNORE_SPACES) > 0);
return i;
case 'Q':
term.type = LITERAL_START;
return i;
default:
if (c >= '1' && c <= '9') {
int n = c - '0';
while ((i < out) && (c = data[i]) >= '0' && c <= '9') {
n = (n * 10) + c - '0';
i++;
}
term.type = (flags & IGNORE_CASE) > 0 ? REG_I : REG;
term.memreg = n;
return i;
}
/*
if(c<256){
CustomParser termp=customParsers[c];
if(termp!=null){
i=termp.parse(i,data,term);
return i;
}
}
*/
}
term.type = CHAR;
term.c = c;
break;
default:
if ((flags & IGNORE_CASE) == 0) {
term.type = CHAR;
term.c = c;
} else {
term.type = CHAR;
term.c = Category.caseFold(c);
//CharacterClass.makeICase(term, c);
}
break;
}
return i;
}
// one of {n},{n,},{,n},{n1,n2}
private static int parseLimits(int i, int end, char[] data, int[] limits) throws PatternSyntaxException {
if (limits.length != LIMITS_LENGTH)
throw new IllegalArgumentException("limits.length=" + limits.length + ", should be " + LIMITS_LENGTH);
limits[LIMITS_PARSE_RESULT_INDEX] = LIMITS_OK;
int ind = 0;
int v = 0;
char c;
while (i < end) {
c = data[i++];
switch (c) {
case ' ':
continue;
case ',':
if (ind > 0) throw new PatternSyntaxException("illegal construction: {.. , , ..}");
limits[ind++] = v;
v = -1;
continue;
case '}':
limits[ind] = v;
if (ind == 0) limits[1] = v;
return i;
default:
if (c > '9' || c < '0') {
//throw new PatternSyntaxException("illegal symbol in iterator: '{"+c+"}'");
limits[LIMITS_PARSE_RESULT_INDEX] = LIMITS_FAILURE;
return i;
}
if (v < 0) v = 0;
v = v * 10 + (c - '0');
}
}
throw new PatternSyntaxException("malformed quantifier");
}
static String termLookup(int t)
{
switch (t)
{
case CHAR: return "CHAR";
case BITSET: return "BITSET";
case BITSET2: return "BITSET2";
case ANY_CHAR: return "ANY_CHAR";
case ANY_CHAR_NE: return "ANY_CHAR_NE";
case REG: return "REG";
case REG_I: return "REG_I";
case FIND: return "FIND";
case FINDREG: return "FINDREG";
case SUCCESS: return "SUCCESS";
case BOUNDARY: return "BOUNDARY";
case DIRECTION: return "DIRECTION";
case UBOUNDARY: return "UBOUNDARY";
case UDIRECTION: return "UDIRECTION";
case GROUP_IN: return "GROUP_IN";
case GROUP_OUT: return "GROUP_OUT";
case VOID: return "VOID";
case START: return "START";
case END: return "END";
case END_EOL: return "END_EOL";
case LINE_START: return "LINE_START";
case LINE_END: return "LINE_END";
case LAST_MATCH_END: return "LAST_MATCH_END";
case CNT_SET_0: return "CNT_SET_0";
case CNT_INC: return "CNT_INC";
case CNT_GT_EQ: return "CNT_GT_EQ";
case READ_CNT_LT: return "READ_CNT_LT";
case CRSTORE_CRINC: return "CRSTORE_CRINC";
case CR_SET_0: return "CR_SET_0";
case CR_LT: return "CR_LT";
case CR_GT_EQ: return "CR_GT_EQ";
case BRANCH: return "BRANCH";
case BRANCH_STORE_CNT: return "BRANCH_STORE_CNT";
case BRANCH_STORE_CNT_AUX1: return "BRANCH_STORE_CNT_AUX1";
case PLOOKAHEAD_IN: return "PLOOKAHEAD_IN";
case PLOOKAHEAD_OUT: return "PLOOKAHEAD_OUT";
case NLOOKAHEAD_IN: return "NLOOKAHEAD_IN";
case NLOOKAHEAD_OUT: return "NLOOKAHEAD_OUT";
case PLOOKBEHIND_IN: return "PLOOKBEHIND_IN";
case PLOOKBEHIND_OUT: return "PLOOKBEHIND_OUT";
case NLOOKBEHIND_IN: return "NLOOKBEHIND_IN";
case NLOOKBEHIND_OUT: return "NLOOKBEHIND_OUT";
case INDEPENDENT_IN: return "INDEPENDENT_IN";
case INDEPENDENT_OUT: return "INDEPENDENT_OUT";
case REPEAT_0_INF: return "REPEAT_0_INF";
case REPEAT_MIN_INF: return "REPEAT_MIN_INF";
case REPEAT_MIN_MAX: return "REPEAT_MIN_MAX";
case REPEAT_REG_MIN_INF: return "REPEAT_REG_MIN_INF";
case REPEAT_REG_MIN_MAX: return "REPEAT_REG_MIN_MAX";
case BACKTRACK_0: return "BACKTRACK_0";
case BACKTRACK_MIN: return "BACKTRACK_MIN";
case BACKTRACK_FIND_MIN: return "BACKTRACK_FIND_MIN";
case BACKTRACK_FINDREG_MIN: return "BACKTRACK_FINDREG_MIN";
case BACKTRACK_REG_MIN: return "BACKTRACK_REG_MIN";
case MEMREG_CONDITION: return "MEMREG_CONDITION";
case LOOKAHEAD_CONDITION_IN: return "LOOKAHEAD_CONDITION_IN";
case LOOKAHEAD_CONDITION_OUT: return "LOOKAHEAD_CONDITION_OUT";
case LOOKBEHIND_CONDITION_IN: return "LOOKBEHIND_CONDITION_IN";
case LOOKBEHIND_CONDITION_OUT: return "LOOKBEHIND_CONDITION_OUT";
default: return "UNKNOWN_TERM";
}
}
public String toString() {
StringBuilder b = new StringBuilder(100);
//b.append(hashCode());
b.append(instanceNum);
b.append(' ');
b.append(termLookup(type));
b.append(": ");
if (inverse) b.append('^');
switch (type) {
case VOID:
b.append("[]");
b.append(" , ");
break;
case CHAR:
b.append(CharacterClass.stringValue(c));
b.append(" , ");
break;
case ANY_CHAR:
b.append("dotall, ");
break;
case ANY_CHAR_NE:
b.append("dot-eols, ");
break;
case BITSET:
b.append('[');
b.append(CharacterClass.stringValue0(bitset));
b.append(']');
b.append(" , weight=");
b.append(weight);
b.append(" , ");
break;
case BITSET2:
b.append('[');
b.append(CharacterClass.stringValue2(bitset2));
b.append(']');
b.append(" , weight2=");
b.append(weight);
b.append(" , ");
break;
case START:
b.append("abs.start");
break;
case END:
b.append("abs.end");
break;
case END_EOL:
b.append("abs.end-eol");
break;
case LINE_START:
b.append("line start");
break;
case LINE_END:
b.append("line end");
break;
case LAST_MATCH_END:
if (inverse) b.append("non-");
b.append("BOUNDARY");
break;
case BOUNDARY:
if (inverse) b.append("non-");
b.append("BOUNDARY");
break;
case UBOUNDARY:
if (inverse) b.append("non-");
b.append("UBOUNDARY");
break;
case DIRECTION:
b.append("DIRECTION");
break;
case UDIRECTION:
b.append("UDIRECTION");
break;
case FINDREG:
b.append('%');
case FIND:
b.append(">>>{");
b.append(target);
b.append("}, <<");
b.append(distance);
if (eat) {
b.append(",eat");
}
b.append(", ");
break;
case REPEAT_0_INF:
b.append("rpt{");
b.append(target);
b.append(",0,inf}");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case REPEAT_MIN_INF:
b.append("rpt{");
b.append(target);
b.append(",");
b.append(minCount);
b.append(",inf}");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case REPEAT_MIN_MAX:
b.append("rpt{");
b.append(target);
b.append(",");
b.append(minCount);
b.append(",");
b.append(maxCount);
b.append("}");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case REPEAT_REG_MIN_INF:
b.append("rpt{$");
b.append(memreg);
b.append(',');
b.append(minCount);
b.append(",inf}");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case REPEAT_REG_MIN_MAX:
b.append("rpt{$");
b.append(memreg);
b.append(',');
b.append(minCount);
b.append(',');
b.append(maxCount);
b.append("}");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case BACKTRACK_0:
b.append("back(0)");
break;
case BACKTRACK_MIN:
b.append("back(");
b.append(minCount);
b.append(")");
break;
case BACKTRACK_REG_MIN:
b.append("back");
b.append("_$");
b.append(memreg);
b.append("(");
b.append(minCount);
b.append(")");
break;
case GROUP_IN:
b.append('(');
if (memreg > 0) b.append(memreg);
b.append('-');
b.append(" , ");
break;
case GROUP_OUT:
b.append('-');
if (memreg > 0) b.append(memreg);
b.append(')');
b.append(" , ");
break;
case PLOOKAHEAD_IN:
b.append('(');
b.append("=");
b.append(lookaheadId);
b.append(" , ");
break;
case PLOOKAHEAD_OUT:
b.append('=');
b.append(lookaheadId);
b.append(')');
b.append(" , ");
break;
case NLOOKAHEAD_IN:
b.append("(!");
b.append(lookaheadId);
b.append(" , ");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case NLOOKAHEAD_OUT:
b.append('!');
b.append(lookaheadId);
b.append(')');
b.append(" , ");
break;
case PLOOKBEHIND_IN:
b.append('(');
b.append("<=");
b.append(lookaheadId);
b.append(" , dist=");
b.append(distance);
b.append(" , ");
break;
case PLOOKBEHIND_OUT:
b.append("<=");
b.append(lookaheadId);
b.append(')');
b.append(" , ");
break;
case NLOOKBEHIND_IN:
b.append("(");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case NLOOKBEHIND_OUT:
b.append("");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case LOOKAHEAD_CONDITION_IN:
b.append("(cond");
b.append(lookaheadId);
b.append(((Lookahead) this).isPositive ? '=' : '!');
b.append(" , ");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case LOOKAHEAD_CONDITION_OUT:
b.append("cond");
b.append(lookaheadId);
b.append(")");
if (failNext != null) {
b.append(", =>");
b.append(failNext.instanceNum);
b.append(", ");
}
break;
case REG:
b.append("$");
b.append(memreg);
b.append(", ");
break;
case SUCCESS:
b.append("END");
break;
case BRANCH_STORE_CNT_AUX1:
b.append("(aux1)");
case BRANCH_STORE_CNT:
b.append("(cnt)");
case BRANCH:
b.append("=>");
if (failNext != null) b.append(failNext.instanceNum);
else b.append("null");
b.append(" , ");
break;
default:
b.append('[');
switch (type) {
case CNT_SET_0:
b.append("cnt=0");
break;
case CNT_INC:
b.append("cnt++");
break;
case CNT_GT_EQ:
b.append("cnt>=").append(maxCount);
break;
case READ_CNT_LT:
b.append("->cnt<").append(maxCount);
break;
case CRSTORE_CRINC:
b.append("M(").append(memreg).append(")->,Cr(").append(cntreg).append(")->,Cr(").append(cntreg).append(")++");
break;
case CR_SET_0:
b.append("Cr(").append(cntreg).append(")=0");
break;
case CR_LT:
b.append("Cr(").append(cntreg).append(")<").append(maxCount);
break;
case CR_GT_EQ:
b.append("Cr(").append(cntreg).append(")>=").append(maxCount);
break;
default:
b.append("unknown type: ").append(type);
}
b.append("] , ");
}
if (next != null) {
b.append("->");
b.append(next.instanceNum);
b.append(", ");
}
//b.append("\r\n");
return b.toString();
}
public String toStringAll() {
return toStringAll(new ArrayList());
}
private String toStringAll(ArrayList v) {
v.add(instanceNum);
String s = toString();
if (next != null) {
if (!v.contains(next.instanceNum)) {
s += "\r\n";
s += next.toStringAll(v);
}
}
if (failNext != null) {
if (!v.contains(failNext.instanceNum)) {
s += "\r\n";
s += failNext.toStringAll(v);
}
}
return s;
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
Term term = (Term) o;
if (type != term.type) return false;
if (inverse != term.inverse) return false;
if (c != term.c) return false;
if (distance != term.distance) return false;
if (eat != term.eat) return false;
if (weight != term.weight) return false;
if (memreg != term.memreg) return false;
if (minCount != term.minCount) return false;
if (maxCount != term.maxCount) return false;
if (cntreg != term.cntreg) return false;
if (lookaheadId != term.lookaheadId) return false;
if (next != null ? !next.equals(term.next) : term.next != null) return false;
if (bitset != null ? !bitset.equals(term.bitset) : term.bitset != null) return false;
// Probably incorrect - comparing Object[] arrays with Arrays.equals
return Arrays.equals(bitset2, term.bitset2) && Arrays.equals(categoryBitset, term.categoryBitset);
//if (!Arrays.equals(brackets, term.brackets)) return false;
/*
if (failNext != null ? !failNext.equals(term.failNext) : term.failNext != null) return false;
if (target != null ? !target.equals(term.target) : term.target != null) return false;
if (prev != null ? !prev.equals(term.prev) : term.prev != null) return false;
if (in != null ? !in.equals(term.in) : term.in != null) return false;
if (out != null ? !out.equals(term.out) : term.out != null) return false;
if (out1 != null ? !out1.equals(term.out1) : term.out1 != null) return false;
if (first != null ? !first.equals(term.first) : term.first != null) return false;
if (current != null ? !current.equals(term.current) : term.current != null) return false;
return branchOut != null ? branchOut.equals(term.branchOut) : term.branchOut == null;
*/
}
@Override
public int hashCode() {
int result = next != null ? next.hashCode() : 0;
result = 31 * result + type;
result = 31 * result + (inverse ? 1 : 0);
result = 31 * result + (int) c;
result = 31 * result + distance;
result = 31 * result + (eat ? 1 : 0);
result = 31 * result + (bitset != null ? bitset.hashCode() : 0);
result = 31 * result + Arrays.hashCode(bitset2);
result = 31 * result + Arrays.hashCode(categoryBitset);
result = 31 * result + weight;
result = 31 * result + memreg;
result = 31 * result + minCount;
result = 31 * result + maxCount;
result = 31 * result + cntreg;
result = 31 * result + lookaheadId;
/*
result = 31 * result + (failNext != null ? failNext.hashCode() : 0);
result = 31 * result + (target != null ? (this == target ? 73 : target.hashCode()) : 0);
result = 31 * result + (prev != null ? (this == prev ? 73 : prev.hashCode()) : 0);
result = 31 * result + (in != null ? (this == in ? 73 : in.hashCode()) : 0);
result = 31 * result + (out != null ? (this == out ? 73 : out.hashCode()) : 0);
result = 31 * result + (out1 != null ? (this == out1 ? 73 : out1.hashCode()) : 0);
result = 31 * result + (first != null ? (this == first ? 73 : first.hashCode()) : 0);
result = 31 * result + (current != null ? (this == current ? 73 : current.hashCode()) : 0);
result = 31 * result + (branchOut != null ? (this == branchOut ? 73 : branchOut.hashCode()) : 0);
*/
return result;
}
}
@JTranscInvisible
class Pretokenizer implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
private static final int START = 1;
static final int END = 2;
static final int PLAIN_GROUP = 3;
static final int POS_LOOKAHEAD = 4;
static final int NEG_LOOKAHEAD = 5;
static final int POS_LOOKBEHIND = 6;
static final int NEG_LOOKBEHIND = 7;
static final int INDEPENDENT_REGEX = 8;
static final int COMMENT = 9;
static final int CONDITIONAL_GROUP = 10;
static final int FLAGS = 11;
static final int CLASS_GROUP = 12;
static final int NAMED_GROUP = 13;
int tOffset;
int tOutside;
private int skip;
private int offset;
private int end;
int c;
int ttype = START;
private char[] data;
//results
private int flags;
private boolean flagsChanged;
String groupName;
boolean groupDeclared;
Pretokenizer(char[] data, int offset, int end) {
if (offset < 0 || end > data.length)
throw new IndexOutOfBoundsException("offset=" + offset + ", end=" + end + ", length=" + data.length);
this.offset = offset;
this.end = end;
this.tOffset = offset;
this.tOutside = offset;
this.data = data;
}
int flags(int def) {
return flagsChanged ? flags : def;
}
void next() throws PatternSyntaxException {
int tOffset = this.tOutside;
int skip = this.skip;
tOffset += skip;
flagsChanged = false;
int end = this.end;
char[] data = this.data;
boolean esc = false;
char ender = '}';
for (int i = tOffset; i < end; i++) {
char c = data[i];
if (esc) {
if(c == 'Q')
{
for (; i < end; i++) {
char c1 = data[i];
if(c1 == '\\') {
if (i + 1 < end && data[i + 1] == 'E') {
i++;
esc = false;
break;
}
}
}
}
else {
esc = false;
}
continue;
}
switch (c) {
case '\\':
esc = true;
continue;
case '|':
case ')':
ttype = c;
this.tOffset = tOffset;
this.tOutside = i;
this.skip = 1;
return;
case '(':
if (((i + 2) < end) && (data[i + 1] == '?')) {
char c1 = data[i + 2];
switch (c1) {
case ':':
ttype = PLAIN_GROUP;
skip = 3; // "(?:" - skip 3 chars
break;
case '=':
ttype = POS_LOOKAHEAD;
skip = 3; // "(?="
break;
case '!':
ttype = NEG_LOOKAHEAD;
skip = 3; // "(?!"
break;
case '<':
switch (c1 = data[i + 3]) {
case '=':
ttype = POS_LOOKBEHIND;
skip = 4; // "(?<="
break;
case '!':
ttype = NEG_LOOKBEHIND;
skip = 4; // "(?'
int nstart, nend;
nstart = p;
if(Category.N.contains(c1))
throw new PatternSyntaxException("number at the start of a named group");
while (Category.IdentifierPart.contains(c1)) {
c1 = data[++p];
skip++;
if (p == end) throw new PatternSyntaxException("malformed named group");
}
nend = p;
if (c1 != '>')
throw new PatternSyntaxException("'>' expected at " + (p - i) + " in " + new String(data, i, end - i));
this.groupName = new String(data, nstart, nend - nstart);
this.groupDeclared = true;
ttype = NAMED_GROUP;
break;
//throw new PatternSyntaxException("invalid character after '(?<' : " + c1);
}
break;
case '>':
ttype = INDEPENDENT_REGEX;
skip = 3; // "(?>"
break;
case '#':
ttype = COMMENT;
skip = 3; // ="(?#".length, the makeTree() skips the rest by itself
break;
case '(':
ttype = CONDITIONAL_GROUP;
skip = 2; //"(?"+"(..." - skip "(?" (2 chars) and parse condition as a group
break;
case '[':
ttype = CLASS_GROUP;
skip = 2; // "(?"+"[..]+...-...&...)" - skip 2 chars and parse a class group
break;
default:
int mOff, mLen;
mLoop:
for (int p = i + 2; p < end; p++) {
char c2 = data[p];
switch (c2) {
case '-':
case 'i':
case 'm':
case 's':
case 'x':
case 'u':
case 'X':
continue mLoop;
case ':':
mOff = i + 2;
mLen = p - mOff;
if (mLen > 0) {
flags = Pattern.parseFlags(data, mOff, mLen);
flagsChanged = true;
}
ttype = PLAIN_GROUP;
skip = mLen + 3; // "(?imsx:" mLen=4; skip= "(?".len + ":".len + mLen = 2+1+4=7
break mLoop;
case ')':
flags = Pattern.parseFlags(data, mOff = (i + 2), mLen = (p - mOff));
flagsChanged = true;
ttype = FLAGS;
skip = mLen + 3; // "(?imsx)" mLen=4, skip="(?".len+")".len+mLen=2+1+4=7
break mLoop;
default:
throw new PatternSyntaxException("wrong char after \"(?\": " + c2);
}
}
break;
}
} else if (((i + 2) < end) && (data[i + 1] == '{')) { //parse named group: ({name}....),({=name}....)
int p = i + 2;
skip = 3; //'({' + '}'
int nstart, nend;
boolean isDecl;
c = data[p];
while (Category.Z.contains(c)) {
c = data[++p];
skip++;
if (p == end) throw new PatternSyntaxException("malformed named group");
}
if (c == '=') {
isDecl = false;
c = data[++p];
skip++;
if (p == end) throw new PatternSyntaxException("malformed named group");
} else isDecl = true;
nstart = p;
while (Category.IdentifierPart.contains(c)) {
c = data[++p];
skip++;
if (p == end) throw new PatternSyntaxException("malformed named group");
}
nend = p;
while (Category.Z.contains(c)) {
c = data[++p];
skip++;
if (p == end) throw new PatternSyntaxException("malformed named group");
}
if (c != '}')
throw new PatternSyntaxException("'}' expected at " + (p - i) + " in " + new String(data, i, end - i));
this.groupName = new String(data, nstart, nend - nstart);
this.groupDeclared = isDecl;
ttype = NAMED_GROUP;
} else {
ttype = '(';
skip = 1;
}
this.tOffset = tOffset;
this.tOutside = i;
this.skip = skip;
return;
case '[':
loop:
for (; ; i++) {
if (i == end) throw new PatternSyntaxException("malformed character class");
char c1 = data[i];
switch (c1) {
case '\\':
i++;
continue;
case ']':
break loop;
}
}
}
}
ttype = END;
this.tOffset = tOffset;
this.tOutside = end;
}
}
@JTranscInvisible
class Branch extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
Branch() {
type = BRANCH;
}
Branch(int type) {
switch (type) {
case BRANCH:
case BRANCH_STORE_CNT:
case BRANCH_STORE_CNT_AUX1:
this.type = type;
break;
default:
throw new IllegalArgumentException("not a branch type: " + type);
}
}
}
@JTranscInvisible
class BackReference extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
BackReference(int no, boolean icase, boolean reverse, boolean bracket) {
super(icase ? REG_I : REG);
mode_reverse = reverse;
mode_bracket = bracket;
mode_insensitive = icase;
memreg = no;
}
}
@JTranscInvisible
class Group extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
Group() {
this(0);
}
Group(int memreg) {
type = GROUP_IN;
this.memreg = memreg;
//used in append()
current = null;
in = this;
prev = null;
out = new Term();
out.type = GROUP_OUT;
out.memreg = memreg;
}
}
@JTranscInvisible
class ConditionalExpr extends Group implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
private Term node;
private boolean newBranchStarted = false;
private boolean linkAsBranch = true;
ConditionalExpr(Lookahead la) {
super(0);
/*
* This all is rather tricky.
* See how this types are handled in Matcher.
* The shortcoming is that we strongly rely upon
* the internal structure of Lookahead.
*/
la.in.type = LOOKAHEAD_CONDITION_IN;
la.out.type = LOOKAHEAD_CONDITION_OUT;
if (la.isPositive) {
node = la.in;
linkAsBranch = true;
//empty 2'nd branch
node.failNext = out;
} else {
node = la.out;
linkAsBranch = false;
//empty 2'nd branch
node.next = out;
}
//node.prev=in;
//in.next=node;
la.prev = in;
in.next = la;
current = la;
//current=node;
}
ConditionalExpr(Lookbehind lb) {
super(0);
/*
* This all is rather tricky.
* See how this types are handled in Matcher.
* The shortcoming is that we strongly rely upon
* the internal structure of Lookahead.
*/
lb.in.type = LOOKBEHIND_CONDITION_IN;
lb.out.type = LOOKBEHIND_CONDITION_OUT;
if (lb.isPositive) {
node = lb.in;
linkAsBranch = true;
//empty 2'nd branch
node.failNext = out;
} else {
node = lb.out;
linkAsBranch = false;
//empty 2'nd branch
node.next = out;
}
lb.prev = in;
in.next = lb;
current = lb;
//current=node;
}
ConditionalExpr(int memreg) {
super(0);
Term condition = new Term(MEMREG_CONDITION);
condition.memreg = memreg;
condition.out = condition;
condition.out1 = null;
condition.branchOut = null;
//default branch
condition.failNext = out;
node = current = condition;
linkAsBranch = true;
condition.prev = in;
in.next = condition;
current = condition;
}
protected void startNewBranch() throws PatternSyntaxException {
if (newBranchStarted) throw new PatternSyntaxException("attempt to set a 3'd choice in a conditional expr.");
Term node = this.node;
node.out1 = null;
if (linkAsBranch) {
node.out = null;
node.branchOut = node;
} else {
node.out = node;
node.branchOut = null;
}
newBranchStarted = true;
current = node;
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
if (!super.equals(o)) return false;
ConditionalExpr that = (ConditionalExpr) o;
return newBranchStarted == that.newBranchStarted && linkAsBranch == that.linkAsBranch && (node != null ? node.equals(that.node) : that.node == null);
}
@Override
public int hashCode() {
int result = super.hashCode();
result = 31 * result + (node != null ? node.hashCode() : 0);
result = 31 * result + (newBranchStarted ? 1 : 0);
result = 31 * result + (linkAsBranch ? 1 : 0);
return result;
}
}
@JTranscInvisible
class IndependentGroup extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
IndependentGroup(int id) {
super(0);
in = this;
out = new Term();
type = INDEPENDENT_IN;
out.type = INDEPENDENT_OUT;
lookaheadId = out.lookaheadId = id;
}
}
@JTranscInvisible
class Lookahead extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
final boolean isPositive;
Lookahead(int id, boolean isPositive) {
this.isPositive = isPositive;
in = this;
out = new Term();
if (isPositive) {
type = PLOOKAHEAD_IN;
out.type = PLOOKAHEAD_OUT;
} else {
type = NLOOKAHEAD_IN;
out.type = NLOOKAHEAD_OUT;
branchOut = this;
}
lookaheadId = id;
out.lookaheadId = id;
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
if (!super.equals(o)) return false;
Lookahead lookahead = (Lookahead) o;
return isPositive == lookahead.isPositive;
}
@Override
public int hashCode() {
int result = super.hashCode();
result = 31 * result + (isPositive ? 1 : 0);
return result;
}
}
@JTranscInvisible
class Lookbehind extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
final boolean isPositive;
private int prevDistance = -1;
Lookbehind(int id, boolean isPositive) {
distance = 0;
this.isPositive = isPositive;
in = this;
out = new Term();
if (isPositive) {
type = PLOOKBEHIND_IN;
out.type = PLOOKBEHIND_OUT;
} else {
type = NLOOKBEHIND_IN;
out.type = NLOOKBEHIND_OUT;
branchOut = this;
}
lookaheadId = id;
out.lookaheadId = id;
}
protected Term append(Term t) throws PatternSyntaxException {
distance += length(t);
return super.append(t);
}
protected Term replaceCurrent(Term t) throws PatternSyntaxException {
distance += length(t) - length(current);
return super.replaceCurrent(t);
}
private static int length(Term t) throws PatternSyntaxException {
int type = t.type;
switch (type) {
case CHAR:
case BITSET:
case BITSET2:
case ANY_CHAR:
case ANY_CHAR_NE:
return 1;
case BOUNDARY:
case DIRECTION:
case UBOUNDARY:
case UDIRECTION:
return 0;
default:
if (type >= FIRST_TRANSPARENT && type <= LAST_TRANSPARENT) return 0;
throw new PatternSyntaxException("variable length element within a lookbehind assertion");
}
}
protected void startNewBranch() throws PatternSyntaxException {
prevDistance = distance;
distance = 0;
super.startNewBranch();
}
protected void close() throws PatternSyntaxException {
int pd = prevDistance;
if (pd >= 0) {
if (distance != pd)
throw new PatternSyntaxException("non-equal branch lengths within a lookbehind assertion");
}
super.close();
}
@Override
public boolean equals(Object o) {
if (this == o) return true;
if (o == null || getClass() != o.getClass()) return false;
if (!super.equals(o)) return false;
Lookbehind that = (Lookbehind) o;
return isPositive == that.isPositive && prevDistance == that.prevDistance;
}
@Override
public int hashCode() {
int result = super.hashCode();
result = 31 * result + (isPositive ? 1 : 0);
result = 31 * result + prevDistance;
return result;
}
}
@JTranscInvisible
class TermIterator extends Term implements Serializable {
private static final long serialVersionUID = 2528136757932720807L;
TermIterator(Term term, int min, int max, ArrayList collection) throws PatternSyntaxException {
collection.add(this);
switch (term.type) {
case CHAR:
case ANY_CHAR:
case ANY_CHAR_NE:
case BITSET:
case BITSET2: {
target = term;
Term back = new Term();
if (min <= 0 && max < 0) {
type = REPEAT_0_INF;
back.type = BACKTRACK_0;
} else if (min > 0 && max < 0) {
type = REPEAT_MIN_INF;
back.type = BACKTRACK_MIN;
minCount = back.minCount = min;
} else {
type = REPEAT_MIN_MAX;
back.type = BACKTRACK_MIN;
minCount = back.minCount = min;
maxCount = max;
}
failNext = back;
in = this;
out = this;
out1 = back;
branchOut = null;
return;
}
case REG: {
target = term;
memreg = term.memreg;
Term back = new Term();
if (max < 0) {
type = REPEAT_REG_MIN_INF;
back.type = BACKTRACK_REG_MIN;
minCount = back.minCount = min;
} else {
type = REPEAT_REG_MIN_MAX;
back.type = BACKTRACK_REG_MIN;
minCount = back.minCount = min;
maxCount = max;
}
failNext = back;
in = this;
out = this;
out1 = back;
branchOut = null;
return;
}
default:
throw new PatternSyntaxException("can't iterate this type: " + term.type);
}
}
void optimize() {
//BACKTRACK_MIN_REG_FIND
Term back = failNext;
Optimizer opt = Optimizer.find(back.next);
if (opt == null) return;
failNext = opt.makeBacktrack(back);
}
}