Examples with IntervalSet org.antlr.v4.runtime.misc.IntervalSet used on opensource projects

Example 1 with IntervalSet

use of org.antlr.v4.runtime.misc.IntervalSet in project antlr4 by antlr.

the class ATNDeserializer method deserializeSets.

private int deserializeSets(char[] data, int p, List<IntervalSet> sets, UnicodeDeserializer unicodeDeserializer) {
    int nsets = toInt(data[p++]);
    for (int i = 0; i < nsets; i++) {
        int nintervals = toInt(data[p]);
        p++;
        IntervalSet set = new IntervalSet();
        sets.add(set);
        boolean containsEof = toInt(data[p++]) != 0;
        if (containsEof) {
            set.add(-1);
        }
        for (int j = 0; j < nintervals; j++) {
            int a = unicodeDeserializer.readUnicode(data, p);
            p += unicodeDeserializer.size();
            int b = unicodeDeserializer.readUnicode(data, p);
            p += unicodeDeserializer.size();
            set.add(a, b);
        }
    }
    return p;
}

Example 2 with IntervalSet

use of org.antlr.v4.runtime.misc.IntervalSet in project antlr4 by antlr.

the class ATNSerializer method serialize.

/** Serialize state descriptors, edge descriptors, and decision→state map
	 *  into list of ints:
	 *
	 * 		grammar-type, (ANTLRParser.LEXER, ...)
	 *  	max token type,
	 *  	num states,
	 *  	state-0-type ruleIndex, state-1-type ruleIndex, ... state-i-type ruleIndex optional-arg ...
	 *  	num rules,
	 *  	rule-1-start-state rule-1-args, rule-2-start-state  rule-2-args, ...
	 *  	(args are token type,actionIndex in lexer else 0,0)
	 *      num modes,
	 *      mode-0-start-state, mode-1-start-state, ... (parser has 0 modes)
	 *      num unicode-bmp-sets
	 *      bmp-set-0-interval-count intervals, bmp-set-1-interval-count intervals, ...
	 *      num unicode-smp-sets
	 *      smp-set-0-interval-count intervals, smp-set-1-interval-count intervals, ...
	 *	num total edges,
	 *      src, trg, edge-type, edge arg1, optional edge arg2 (present always), ...
	 *      num decisions,
	 *      decision-0-start-state, decision-1-start-state, ...
	 *
	 *  Convenient to pack into unsigned shorts to make as Java string.
	 */
public IntegerList serialize() {
    IntegerList data = new IntegerList();
    data.add(ATNDeserializer.SERIALIZED_VERSION);
    serializeUUID(data, ATNDeserializer.SERIALIZED_UUID);
    // convert grammar type to ATN const to avoid dependence on ANTLRParser
    data.add(atn.grammarType.ordinal());
    data.add(atn.maxTokenType);
    int nedges = 0;
    // Note that we use a LinkedHashMap as a set to
    // maintain insertion order while deduplicating
    // entries with the same key.
    Map<IntervalSet, Boolean> sets = new LinkedHashMap<>();
    // dump states, count edges and collect sets while doing so
    IntegerList nonGreedyStates = new IntegerList();
    IntegerList precedenceStates = new IntegerList();
    data.add(atn.states.size());
    for (ATNState s : atn.states) {
        if (s == null) {
            // might be optimized away
            data.add(ATNState.INVALID_TYPE);
            continue;
        }
        int stateType = s.getStateType();
        if (s instanceof DecisionState && ((DecisionState) s).nonGreedy) {
            nonGreedyStates.add(s.stateNumber);
        }
        if (s instanceof RuleStartState && ((RuleStartState) s).isLeftRecursiveRule) {
            precedenceStates.add(s.stateNumber);
        }
        data.add(stateType);
        if (s.ruleIndex == -1) {
            data.add(Character.MAX_VALUE);
        } else {
            data.add(s.ruleIndex);
        }
        if (s.getStateType() == ATNState.LOOP_END) {
            data.add(((LoopEndState) s).loopBackState.stateNumber);
        } else if (s instanceof BlockStartState) {
            data.add(((BlockStartState) s).endState.stateNumber);
        }
        if (s.getStateType() != ATNState.RULE_STOP) {
            // the deserializer can trivially derive these edges, so there's no need to serialize them
            nedges += s.getNumberOfTransitions();
        }
        for (int i = 0; i < s.getNumberOfTransitions(); i++) {
            Transition t = s.transition(i);
            int edgeType = Transition.serializationTypes.get(t.getClass());
            if (edgeType == Transition.SET || edgeType == Transition.NOT_SET) {
                SetTransition st = (SetTransition) t;
                sets.put(st.set, true);
            }
        }
    }
    // non-greedy states
    data.add(nonGreedyStates.size());
    for (int i = 0; i < nonGreedyStates.size(); i++) {
        data.add(nonGreedyStates.get(i));
    }
    // precedence states
    data.add(precedenceStates.size());
    for (int i = 0; i < precedenceStates.size(); i++) {
        data.add(precedenceStates.get(i));
    }
    int nrules = atn.ruleToStartState.length;
    data.add(nrules);
    for (int r = 0; r < nrules; r++) {
        ATNState ruleStartState = atn.ruleToStartState[r];
        data.add(ruleStartState.stateNumber);
        if (atn.grammarType == ATNType.LEXER) {
            if (atn.ruleToTokenType[r] == Token.EOF) {
                data.add(Character.MAX_VALUE);
            } else {
                data.add(atn.ruleToTokenType[r]);
            }
        }
    }
    int nmodes = atn.modeToStartState.size();
    data.add(nmodes);
    if (nmodes > 0) {
        for (ATNState modeStartState : atn.modeToStartState) {
            data.add(modeStartState.stateNumber);
        }
    }
    List<IntervalSet> bmpSets = new ArrayList<>();
    List<IntervalSet> smpSets = new ArrayList<>();
    for (IntervalSet set : sets.keySet()) {
        if (set.getMaxElement() <= Character.MAX_VALUE) {
            bmpSets.add(set);
        } else {
            smpSets.add(set);
        }
    }
    serializeSets(data, bmpSets, new CodePointSerializer() {

        @Override
        public void serializeCodePoint(IntegerList data, int cp) {
            data.add(cp);
        }
    });
    serializeSets(data, smpSets, new CodePointSerializer() {

        @Override
        public void serializeCodePoint(IntegerList data, int cp) {
            serializeInt(data, cp);
        }
    });
    Map<IntervalSet, Integer> setIndices = new HashMap<>();
    int setIndex = 0;
    for (IntervalSet bmpSet : bmpSets) {
        setIndices.put(bmpSet, setIndex++);
    }
    for (IntervalSet smpSet : smpSets) {
        setIndices.put(smpSet, setIndex++);
    }
    data.add(nedges);
    for (ATNState s : atn.states) {
        if (s == null) {
            // might be optimized away
            continue;
        }
        if (s.getStateType() == ATNState.RULE_STOP) {
            continue;
        }
        for (int i = 0; i < s.getNumberOfTransitions(); i++) {
            Transition t = s.transition(i);
            if (atn.states.get(t.target.stateNumber) == null) {
                throw new IllegalStateException("Cannot serialize a transition to a removed state.");
            }
            int src = s.stateNumber;
            int trg = t.target.stateNumber;
            int edgeType = Transition.serializationTypes.get(t.getClass());
            int arg1 = 0;
            int arg2 = 0;
            int arg3 = 0;
            switch(edgeType) {
                case Transition.RULE:
                    trg = ((RuleTransition) t).followState.stateNumber;
                    arg1 = ((RuleTransition) t).target.stateNumber;
                    arg2 = ((RuleTransition) t).ruleIndex;
                    arg3 = ((RuleTransition) t).precedence;
                    break;
                case Transition.PRECEDENCE:
                    PrecedencePredicateTransition ppt = (PrecedencePredicateTransition) t;
                    arg1 = ppt.precedence;
                    break;
                case Transition.PREDICATE:
                    PredicateTransition pt = (PredicateTransition) t;
                    arg1 = pt.ruleIndex;
                    arg2 = pt.predIndex;
                    arg3 = pt.isCtxDependent ? 1 : 0;
                    break;
                case Transition.RANGE:
                    arg1 = ((RangeTransition) t).from;
                    arg2 = ((RangeTransition) t).to;
                    if (arg1 == Token.EOF) {
                        arg1 = 0;
                        arg3 = 1;
                    }
                    break;
                case Transition.ATOM:
                    arg1 = ((AtomTransition) t).label;
                    if (arg1 == Token.EOF) {
                        arg1 = 0;
                        arg3 = 1;
                    }
                    break;
                case Transition.ACTION:
                    ActionTransition at = (ActionTransition) t;
                    arg1 = at.ruleIndex;
                    arg2 = at.actionIndex;
                    if (arg2 == -1) {
                        arg2 = 0xFFFF;
                    }
                    arg3 = at.isCtxDependent ? 1 : 0;
                    break;
                case Transition.SET:
                    arg1 = setIndices.get(((SetTransition) t).set);
                    break;
                case Transition.NOT_SET:
                    arg1 = setIndices.get(((SetTransition) t).set);
                    break;
                case Transition.WILDCARD:
                    break;
            }
            data.add(src);
            data.add(trg);
            data.add(edgeType);
            data.add(arg1);
            data.add(arg2);
            data.add(arg3);
        }
    }
    int ndecisions = atn.decisionToState.size();
    data.add(ndecisions);
    for (DecisionState decStartState : atn.decisionToState) {
        data.add(decStartState.stateNumber);
    }
    //
    if (atn.grammarType == ATNType.LEXER) {
        data.add(atn.lexerActions.length);
        for (LexerAction action : atn.lexerActions) {
            data.add(action.getActionType().ordinal());
            switch(action.getActionType()) {
                case CHANNEL:
                    int channel = ((LexerChannelAction) action).getChannel();
                    data.add(channel != -1 ? channel : 0xFFFF);
                    data.add(0);
                    break;
                case CUSTOM:
                    int ruleIndex = ((LexerCustomAction) action).getRuleIndex();
                    int actionIndex = ((LexerCustomAction) action).getActionIndex();
                    data.add(ruleIndex != -1 ? ruleIndex : 0xFFFF);
                    data.add(actionIndex != -1 ? actionIndex : 0xFFFF);
                    break;
                case MODE:
                    int mode = ((LexerModeAction) action).getMode();
                    data.add(mode != -1 ? mode : 0xFFFF);
                    data.add(0);
                    break;
                case MORE:
                    data.add(0);
                    data.add(0);
                    break;
                case POP_MODE:
                    data.add(0);
                    data.add(0);
                    break;
                case PUSH_MODE:
                    mode = ((LexerPushModeAction) action).getMode();
                    data.add(mode != -1 ? mode : 0xFFFF);
                    data.add(0);
                    break;
                case SKIP:
                    data.add(0);
                    data.add(0);
                    break;
                case TYPE:
                    int type = ((LexerTypeAction) action).getType();
                    data.add(type != -1 ? type : 0xFFFF);
                    data.add(0);
                    break;
                default:
                    String message = String.format(Locale.getDefault(), "The specified lexer action type %s is not valid.", action.getActionType());
                    throw new IllegalArgumentException(message);
            }
        }
    }
    // don't adjust the first value since that's the version number
    for (int i = 1; i < data.size(); i++) {
        if (data.get(i) < Character.MIN_VALUE || data.get(i) > Character.MAX_VALUE) {
            throw new UnsupportedOperationException("Serialized ATN data element " + data.get(i) + " element " + i + " out of range " + (int) Character.MIN_VALUE + ".." + (int) Character.MAX_VALUE);
        }
        int value = (data.get(i) + 2) & 0xFFFF;
        data.set(i, value);
    }
    return data;
}

Example 3 with IntervalSet

use of org.antlr.v4.runtime.misc.IntervalSet in project antlr4 by antlr.

the class LL1Analyzer method getDecisionLookahead.

/**
	 * Calculates the SLL(1) expected lookahead set for each outgoing transition
	 * of an {@link ATNState}. The returned array has one element for each
	 * outgoing transition in {@code s}. If the closure from transition
	 * <em>i</em> leads to a semantic predicate before matching a symbol, the
	 * element at index <em>i</em> of the result will be {@code null}.
	 *
	 * @param s the ATN state
	 * @return the expected symbols for each outgoing transition of {@code s}.
	 */
public IntervalSet[] getDecisionLookahead(ATNState s) {
    //		System.out.println("LOOK("+s.stateNumber+")");
    if (s == null) {
        return null;
    }
    IntervalSet[] look = new IntervalSet[s.getNumberOfTransitions()];
    for (int alt = 0; alt < s.getNumberOfTransitions(); alt++) {
        look[alt] = new IntervalSet();
        Set<ATNConfig> lookBusy = new HashSet<ATNConfig>();
        // fail to get lookahead upon pred
        boolean seeThruPreds = false;
        _LOOK(s.transition(alt).target, null, PredictionContext.EMPTY, look[alt], lookBusy, new BitSet(), seeThruPreds, false);
        // or we had a predicate when we !seeThruPreds
        if (look[alt].size() == 0 || look[alt].contains(HIT_PRED)) {
            look[alt] = null;
        }
    }
    return look;
}

Example 4 with IntervalSet

use of org.antlr.v4.runtime.misc.IntervalSet in project antlr4 by antlr.

the class LL1Analyzer method LOOK.

/**
	 * Compute set of tokens that can follow {@code s} in the ATN in the
	 * specified {@code ctx}.
	 *
	 * <p>If {@code ctx} is {@code null} and the end of the rule containing
	 * {@code s} is reached, {@link Token#EPSILON} is added to the result set.
	 * If {@code ctx} is not {@code null} and the end of the outermost rule is
	 * reached, {@link Token#EOF} is added to the result set.</p>
	 *
	 * @param s the ATN state
	 * @param stopState the ATN state to stop at. This can be a
	 * {@link BlockEndState} to detect epsilon paths through a closure.
	 * @param ctx the complete parser context, or {@code null} if the context
	 * should be ignored
	 *
	 * @return The set of tokens that can follow {@code s} in the ATN in the
	 * specified {@code ctx}.
	 */
public IntervalSet LOOK(ATNState s, ATNState stopState, RuleContext ctx) {
    IntervalSet r = new IntervalSet();
    // ignore preds; get all lookahead
    boolean seeThruPreds = true;
    PredictionContext lookContext = ctx != null ? PredictionContext.fromRuleContext(s.atn, ctx) : null;
    _LOOK(s, stopState, lookContext, r, new HashSet<ATNConfig>(), new BitSet(), seeThruPreds, true);
    return r;
}

Example 5 with IntervalSet

use of org.antlr.v4.runtime.misc.IntervalSet in project antlr4 by antlr.

the class DefaultErrorStrategy method reportUnwantedToken.

/**
	 * This method is called to report a syntax error which requires the removal
	 * of a token from the input stream. At the time this method is called, the
	 * erroneous symbol is current {@code LT(1)} symbol and has not yet been
	 * removed from the input stream. When this method returns,
	 * {@code recognizer} is in error recovery mode.
	 *
	 * <p>This method is called when {@link #singleTokenDeletion} identifies
	 * single-token deletion as a viable recovery strategy for a mismatched
	 * input error.</p>
	 *
	 * <p>The default implementation simply returns if the handler is already in
	 * error recovery mode. Otherwise, it calls {@link #beginErrorCondition} to
	 * enter error recovery mode, followed by calling
	 * {@link Parser#notifyErrorListeners}.</p>
	 *
	 * @param recognizer the parser instance
	 */
protected void reportUnwantedToken(Parser recognizer) {
    if (inErrorRecoveryMode(recognizer)) {
        return;
    }
    beginErrorCondition(recognizer);
    Token t = recognizer.getCurrentToken();
    String tokenName = getTokenErrorDisplay(t);
    IntervalSet expecting = getExpectedTokens(recognizer);
    String msg = "extraneous input " + tokenName + " expecting " + expecting.toString(recognizer.getVocabulary());
    recognizer.notifyErrorListeners(t, msg, null);
}