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

Example 6 with Interval

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

the class SwiftTarget method serializeTojson.

//added by janyou -->
public String serializeTojson(ATN atn) {
    JsonObjectBuilder builder = Json.createObjectBuilder();
    builder.add("version", ATNDeserializer.SERIALIZED_VERSION);
    builder.add("uuid", ATNDeserializer.SERIALIZED_UUID.toString());
    // convert grammar type to ATN const to avoid dependence on ANTLRParser
    builder.add("grammarType", atn.grammarType.ordinal());
    builder.add("maxTokenType", atn.maxTokenType);
    //states
    int nedges = 0;
    Map<IntervalSet, Integer> setIndices = new HashMap<IntervalSet, Integer>();
    List<IntervalSet> sets = new ArrayList<IntervalSet>();
    JsonArrayBuilder statesBuilder = Json.createArrayBuilder();
    IntegerList nonGreedyStates = new IntegerList();
    IntegerList precedenceStates = new IntegerList();
    for (ATNState s : atn.states) {
        JsonObjectBuilder stateBuilder = Json.createObjectBuilder();
        if (s == null) {
            // might be optimized away
            statesBuilder.addNull();
            continue;
        }
        int stateType = s.getStateType();
        stateBuilder.add("stateType", stateType);
        //stateBuilder.add("stateNumber",s.stateNumber);
        stateBuilder.add("ruleIndex", s.ruleIndex);
        if (s instanceof DecisionState && ((DecisionState) s).nonGreedy) {
            nonGreedyStates.add(s.stateNumber);
        }
        if (s instanceof RuleStartState && ((RuleStartState) s).isLeftRecursiveRule) {
            precedenceStates.add(s.stateNumber);
        }
        if (s.getStateType() == ATNState.LOOP_END) {
            stateBuilder.add("detailStateNumber", ((LoopEndState) s).loopBackState.stateNumber);
        } else if (s instanceof BlockStartState) {
            stateBuilder.add("detailStateNumber", ((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;
                if (!setIndices.containsKey(st.set)) {
                    sets.add(st.set);
                    setIndices.put(st.set, sets.size() - 1);
                }
            }
        }
        statesBuilder.add(stateBuilder);
    }
    builder.add("states", statesBuilder);
    // non-greedy states
    JsonArrayBuilder nonGreedyStatesBuilder = Json.createArrayBuilder();
    for (int i = 0; i < nonGreedyStates.size(); i++) {
        nonGreedyStatesBuilder.add(nonGreedyStates.get(i));
    }
    builder.add("nonGreedyStates", nonGreedyStatesBuilder);
    // precedence states
    JsonArrayBuilder precedenceStatesBuilder = Json.createArrayBuilder();
    for (int i = 0; i < precedenceStates.size(); i++) {
        precedenceStatesBuilder.add(precedenceStates.get(i));
    }
    builder.add("precedenceStates", precedenceStatesBuilder);
    JsonArrayBuilder ruleToStartStateBuilder = Json.createArrayBuilder();
    int nrules = atn.ruleToStartState.length;
    for (int r = 0; r < nrules; r++) {
        JsonObjectBuilder stateBuilder = Json.createObjectBuilder();
        ATNState ruleStartState = atn.ruleToStartState[r];
        stateBuilder.add("stateNumber", ruleStartState.stateNumber);
        if (atn.grammarType == ATNType.LEXER) {
            //				if (atn.ruleToTokenType[r] == Token.EOF) {
            //					//data.add(Character.MAX_VALUE);
            //					stateBuilder.add("ruleToTokenType",-1);
            //				}
            //				else {
            //					//data.add(atn.ruleToTokenType[r]);
            stateBuilder.add("ruleToTokenType", atn.ruleToTokenType[r]);
        //				}
        }
        ruleToStartStateBuilder.add(stateBuilder);
    }
    builder.add("ruleToStartState", ruleToStartStateBuilder);
    JsonArrayBuilder modeToStartStateBuilder = Json.createArrayBuilder();
    int nmodes = atn.modeToStartState.size();
    if (nmodes > 0) {
        for (ATNState modeStartState : atn.modeToStartState) {
            modeToStartStateBuilder.add(modeStartState.stateNumber);
        }
    }
    builder.add("modeToStartState", modeToStartStateBuilder);
    JsonArrayBuilder nsetsBuilder = Json.createArrayBuilder();
    int nsets = sets.size();
    //data.add(nsets);
    builder.add("nsets", nsets);
    for (IntervalSet set : sets) {
        JsonObjectBuilder setBuilder = Json.createObjectBuilder();
        boolean containsEof = set.contains(Token.EOF);
        if (containsEof && set.getIntervals().get(0).b == Token.EOF) {
            //data.add(set.getIntervals().size() - 1);
            setBuilder.add("size", set.getIntervals().size() - 1);
        } else {
            //data.add(set.getIntervals().size());
            setBuilder.add("size", set.getIntervals().size());
        }
        setBuilder.add("containsEof", containsEof ? 1 : 0);
        JsonArrayBuilder IntervalsBuilder = Json.createArrayBuilder();
        for (Interval I : set.getIntervals()) {
            JsonObjectBuilder IntervalBuilder = Json.createObjectBuilder();
            if (I.a == Token.EOF) {
                if (I.b == Token.EOF) {
                    continue;
                } else {
                    IntervalBuilder.add("a", 0);
                //data.add(0);
                }
            } else {
                IntervalBuilder.add("a", I.a);
            //data.add(I.a);
            }
            IntervalBuilder.add("b", I.b);
            IntervalsBuilder.add(IntervalBuilder);
        }
        setBuilder.add("Intervals", IntervalsBuilder);
        nsetsBuilder.add(setBuilder);
    }
    builder.add("IntervalSet", nsetsBuilder);
    //builder.add("nedges",nedges);
    JsonArrayBuilder allTransitionsBuilder = Json.createArrayBuilder();
    for (ATNState s : atn.states) {
        if (s == null) {
            // might be optimized away
            continue;
        }
        if (s.getStateType() == ATNState.RULE_STOP) {
            continue;
        }
        JsonArrayBuilder transitionsBuilder = Json.createArrayBuilder();
        for (int i = 0; i < s.getNumberOfTransitions(); i++) {
            JsonObjectBuilder transitionBuilder = Json.createObjectBuilder();
            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;
            }
            transitionBuilder.add("src", src);
            transitionBuilder.add("trg", trg);
            transitionBuilder.add("edgeType", edgeType);
            transitionBuilder.add("arg1", arg1);
            transitionBuilder.add("arg2", arg2);
            transitionBuilder.add("arg3", arg3);
            transitionsBuilder.add(transitionBuilder);
        }
        allTransitionsBuilder.add(transitionsBuilder);
    }
    builder.add("allTransitionsBuilder", allTransitionsBuilder);
    int ndecisions = atn.decisionToState.size();
    //data.add(ndecisions);
    JsonArrayBuilder decisionToStateBuilder = Json.createArrayBuilder();
    for (DecisionState decStartState : atn.decisionToState) {
        //data.add(decStartState.stateNumber);
        decisionToStateBuilder.add(decStartState.stateNumber);
    }
    builder.add("decisionToState", decisionToStateBuilder);
    //
    // LEXER ACTIONS
    //
    JsonArrayBuilder lexerActionsBuilder = Json.createArrayBuilder();
    if (atn.grammarType == ATNType.LEXER) {
        //data.add(atn.lexerActions.length);
        for (LexerAction action : atn.lexerActions) {
            JsonObjectBuilder lexerActionBuilder = Json.createObjectBuilder();
            lexerActionBuilder.add("actionType", action.getActionType().ordinal());
            //data.add(action.getActionType().ordinal());
            switch(action.getActionType()) {
                case CHANNEL:
                    int channel = ((LexerChannelAction) action).getChannel();
                    lexerActionBuilder.add("a", channel);
                    lexerActionBuilder.add("b", 0);
                    break;
                case CUSTOM:
                    int ruleIndex = ((LexerCustomAction) action).getRuleIndex();
                    int actionIndex = ((LexerCustomAction) action).getActionIndex();
                    lexerActionBuilder.add("a", ruleIndex);
                    lexerActionBuilder.add("b", actionIndex);
                    break;
                case MODE:
                    int mode = ((LexerModeAction) action).getMode();
                    lexerActionBuilder.add("a", mode);
                    lexerActionBuilder.add("b", 0);
                    break;
                case MORE:
                    lexerActionBuilder.add("a", 0);
                    lexerActionBuilder.add("b", 0);
                    break;
                case POP_MODE:
                    lexerActionBuilder.add("a", 0);
                    lexerActionBuilder.add("b", 0);
                    break;
                case PUSH_MODE:
                    mode = ((LexerPushModeAction) action).getMode();
                    lexerActionBuilder.add("a", mode);
                    lexerActionBuilder.add("b", 0);
                    break;
                case SKIP:
                    lexerActionBuilder.add("a", 0);
                    lexerActionBuilder.add("b", 0);
                    break;
                case TYPE:
                    int type = ((LexerTypeAction) action).getType();
                    lexerActionBuilder.add("a", type);
                    lexerActionBuilder.add("b", 0);
                    break;
                default:
                    String message = String.format(Locale.getDefault(), "The specified lexer action type %s is not valid.", action.getActionType());
                    throw new IllegalArgumentException(message);
            }
            lexerActionsBuilder.add(lexerActionBuilder);
        }
    }
    builder.add("lexerActions", lexerActionsBuilder);
    // 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 out of range.");
    //			}
    //
    //			int value = (data.get(i) + 2) & 0xFFFF;
    //			data.set(i, value);
    //		}
    JsonObject data = builder.build();
    //  System.out.print(data.toString());
    return data.toString();
}

Example 7 with Interval

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

the class ATNOptimizer method optimizeSets.

private static void optimizeSets(Grammar g, ATN atn) {
    if (g.isParser()) {
        // parser codegen doesn't currently support SetTransition
        return;
    }
    int removedStates = 0;
    List<DecisionState> decisions = atn.decisionToState;
    for (DecisionState decision : decisions) {
        if (decision.ruleIndex >= 0) {
            Rule rule = g.getRule(decision.ruleIndex);
            if (Character.isLowerCase(rule.name.charAt(0))) {
                // parser codegen doesn't currently support SetTransition
                continue;
            }
        }
        IntervalSet setTransitions = new IntervalSet();
        for (int i = 0; i < decision.getNumberOfTransitions(); i++) {
            Transition epsTransition = decision.transition(i);
            if (!(epsTransition instanceof EpsilonTransition)) {
                continue;
            }
            if (epsTransition.target.getNumberOfTransitions() != 1) {
                continue;
            }
            Transition transition = epsTransition.target.transition(0);
            if (!(transition.target instanceof BlockEndState)) {
                continue;
            }
            if (transition instanceof NotSetTransition) {
                // TODO: not yet implemented
                continue;
            }
            if (transition instanceof AtomTransition || transition instanceof RangeTransition || transition instanceof SetTransition) {
                setTransitions.add(i);
            }
        }
        // due to min alt resolution policies, can only collapse sequential alts
        for (int i = setTransitions.getIntervals().size() - 1; i >= 0; i--) {
            Interval interval = setTransitions.getIntervals().get(i);
            if (interval.length() <= 1) {
                continue;
            }
            ATNState blockEndState = decision.transition(interval.a).target.transition(0).target;
            IntervalSet matchSet = new IntervalSet();
            for (int j = interval.a; j <= interval.b; j++) {
                Transition matchTransition = decision.transition(j).target.transition(0);
                if (matchTransition instanceof NotSetTransition) {
                    throw new UnsupportedOperationException("Not yet implemented.");
                }
                IntervalSet set = matchTransition.label();
                List<Interval> intervals = set.getIntervals();
                int n = intervals.size();
                for (int k = 0; k < n; k++) {
                    Interval setInterval = intervals.get(k);
                    int a = setInterval.a;
                    int b = setInterval.b;
                    if (a != -1 && b != -1) {
                        for (int v = a; v <= b; v++) {
                            if (matchSet.contains(v)) {
                                // TODO: Token is missing (i.e. position in source will not be displayed).
                                g.tool.errMgr.grammarError(ErrorType.CHARACTERS_COLLISION_IN_SET, g.fileName, null, CharSupport.getANTLRCharLiteralForChar(v), CharSupport.getIntervalSetEscapedString(matchSet));
                                break;
                            }
                        }
                    }
                }
                matchSet.addAll(set);
            }
            Transition newTransition;
            if (matchSet.getIntervals().size() == 1) {
                if (matchSet.size() == 1) {
                    newTransition = CodePointTransitions.createWithCodePoint(blockEndState, matchSet.getMinElement());
                } else {
                    Interval matchInterval = matchSet.getIntervals().get(0);
                    newTransition = CodePointTransitions.createWithCodePointRange(blockEndState, matchInterval.a, matchInterval.b);
                }
            } else {
                newTransition = new SetTransition(blockEndState, matchSet);
            }
            decision.transition(interval.a).target.setTransition(0, newTransition);
            for (int j = interval.a + 1; j <= interval.b; j++) {
                Transition removed = decision.removeTransition(interval.a + 1);
                atn.removeState(removed.target);
                removedStates++;
            }
        }
    }
//		System.out.println("ATN optimizer removed " + removedStates + " states by collapsing sets.");
}

Example 8 with Interval

use of org.antlr.v4.runtime.misc.Interval in project compiler by boalang.

the class BoaCompiler method main.

public static void main(final String[] args) throws IOException {
    CommandLine cl = processCommandLineOptions(args);
    if (cl == null)
        return;
    final ArrayList<File> inputFiles = BoaCompiler.inputFiles;
    // get the name of the generated class
    final String className = getGeneratedClass(cl);
    // get the filename of the jar we will be writing
    final String jarName;
    if (cl.hasOption('o'))
        jarName = cl.getOptionValue('o');
    else
        jarName = className + ".jar";
    // make the output directory
    File outputRoot = null;
    if (cl.hasOption("cd")) {
        outputRoot = new File(cl.getOptionValue("cd"));
    } else {
        outputRoot = new File(new File(System.getProperty("java.io.tmpdir")), UUID.randomUUID().toString());
    }
    final File outputSrcDir = new File(outputRoot, "boa");
    if (!outputSrcDir.mkdirs())
        throw new IOException("unable to mkdir " + outputSrcDir);
    // find custom libs to load
    final List<URL> libs = new ArrayList<URL>();
    if (cl.hasOption('l'))
        for (final String lib : cl.getOptionValues('l')) libs.add(new File(lib).toURI().toURL());
    final File outputFile = new File(outputSrcDir, className + ".java");
    final BufferedOutputStream o = new BufferedOutputStream(new FileOutputStream(outputFile));
    try {
        final List<String> jobnames = new ArrayList<String>();
        final List<String> jobs = new ArrayList<String>();
        final List<Integer> seeds = new ArrayList<Integer>();
        boolean isSimple = true;
        final List<Program> visitorPrograms = new ArrayList<Program>();
        SymbolTable.initialize(libs);
        final int maxVisitors;
        if (cl.hasOption('v'))
            maxVisitors = Integer.parseInt(cl.getOptionValue('v'));
        else
            maxVisitors = Integer.MAX_VALUE;
        for (int i = 0; i < inputFiles.size(); i++) {
            final File f = inputFiles.get(i);
            try {
                final BoaLexer lexer = new BoaLexer(new ANTLRFileStream(f.getAbsolutePath()));
                // use the whole input string to seed the RNG
                seeds.add(lexer._input.getText(new Interval(0, lexer._input.size())).hashCode());
                lexer.removeErrorListeners();
                lexer.addErrorListener(new LexerErrorListener());
                final CommonTokenStream tokens = new CommonTokenStream(lexer);
                final BoaParser parser = new BoaParser(tokens);
                parser.removeErrorListeners();
                parser.addErrorListener(new BaseErrorListener() {

                    @Override
                    public void syntaxError(Recognizer<?, ?> recognizer, Object offendingSymbol, int line, int charPositionInLine, String msg, RecognitionException e) throws ParseCancellationException {
                        throw new ParseCancellationException(e);
                    }
                });
                final BoaErrorListener parserErrorListener = new ParserErrorListener();
                final Start p = parse(tokens, parser, parserErrorListener);
                if (cl.hasOption("ast"))
                    new ASTPrintingVisitor().start(p);
                final String jobName = "" + i;
                try {
                    if (!parserErrorListener.hasError) {
                        new TypeCheckingVisitor().start(p, new SymbolTable());
                        final TaskClassifyingVisitor simpleVisitor = new TaskClassifyingVisitor();
                        simpleVisitor.start(p);
                        LOG.info(f.getName() + ": task complexity: " + (!simpleVisitor.isComplex() ? "simple" : "complex"));
                        isSimple &= !simpleVisitor.isComplex();
                        new InheritedAttributeTransformer().start(p);
                        new LocalAggregationTransformer().start(p);
                        // also let jobs have own methods if visitor merging is disabled
                        if (!simpleVisitor.isComplex() || maxVisitors < 2 || inputFiles.size() == 1) {
                            new VisitorOptimizingTransformer().start(p);
                            if (cl.hasOption("pp"))
                                new PrettyPrintVisitor().start(p);
                            if (cl.hasOption("ast2"))
                                new ASTPrintingVisitor().start(p);
                            final CodeGeneratingVisitor cg = new CodeGeneratingVisitor(jobName);
                            cg.start(p);
                            jobs.add(cg.getCode());
                            jobnames.add(jobName);
                        } else // if a job has visitors, fuse them all together into a single program
                        {
                            p.getProgram().jobName = jobName;
                            visitorPrograms.add(p.getProgram());
                        }
                    }
                } catch (final TypeCheckException e) {
                    parserErrorListener.error("typecheck", lexer, null, e.n.beginLine, e.n.beginColumn, e.n2.endColumn - e.n.beginColumn + 1, e.getMessage(), e);
                }
            } catch (final Exception e) {
                System.err.print(f.getName() + ": compilation failed: ");
                e.printStackTrace();
            }
        }
        if (!visitorPrograms.isEmpty())
            try {
                for (final Program p : new VisitorMergingTransformer().mergePrograms(visitorPrograms, maxVisitors)) {
                    new VisitorOptimizingTransformer().start(p);
                    if (cl.hasOption("pp"))
                        new PrettyPrintVisitor().start(p);
                    if (cl.hasOption("ast2"))
                        new ASTPrintingVisitor().start(p);
                    final CodeGeneratingVisitor cg = new CodeGeneratingVisitor(p.jobName);
                    cg.start(p);
                    jobs.add(cg.getCode());
                    jobnames.add(p.jobName);
                }
            } catch (final Exception e) {
                System.err.println("error fusing visitors - falling back: " + e);
                e.printStackTrace();
                for (final Program p : visitorPrograms) {
                    new VisitorOptimizingTransformer().start(p);
                    if (cl.hasOption("pp"))
                        new PrettyPrintVisitor().start(p);
                    if (cl.hasOption("ast2"))
                        new ASTPrintingVisitor().start(p);
                    final CodeGeneratingVisitor cg = new CodeGeneratingVisitor(p.jobName);
                    cg.start(p);
                    jobs.add(cg.getCode());
                    jobnames.add(p.jobName);
                }
            }
        if (jobs.size() == 0)
            throw new RuntimeException("no files compiled without error");
        final ST st = AbstractCodeGeneratingVisitor.stg.getInstanceOf("Program");
        st.add("name", className);
        st.add("numreducers", inputFiles.size());
        st.add("jobs", jobs);
        st.add("jobnames", jobnames);
        st.add("combineTables", CodeGeneratingVisitor.combineAggregatorStrings);
        st.add("reduceTables", CodeGeneratingVisitor.reduceAggregatorStrings);
        st.add("splitsize", isSimple ? 64 * 1024 * 1024 : 10 * 1024 * 1024);
        st.add("seeds", seeds);
        if (DefaultProperties.localDataPath != null) {
            st.add("isLocal", true);
        }
        o.write(st.render().getBytes());
    } finally {
        o.close();
    }
    compileGeneratedSrc(cl, jarName, outputRoot, outputFile);
}

Example 9 with Interval

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

the class ATNSerializer method serializeSets.

private static void serializeSets(IntegerList data, Collection<IntervalSet> sets, CodePointSerializer codePointSerializer) {
    int nSets = sets.size();
    data.add(nSets);
    for (IntervalSet set : sets) {
        boolean containsEof = set.contains(Token.EOF);
        if (containsEof && set.getIntervals().get(0).b == Token.EOF) {
            data.add(set.getIntervals().size() - 1);
        } else {
            data.add(set.getIntervals().size());
        }
        data.add(containsEof ? 1 : 0);
        for (Interval I : set.getIntervals()) {
            if (I.a == Token.EOF) {
                if (I.b == Token.EOF) {
                    continue;
                } else {
                    codePointSerializer.serializeCodePoint(data, 0);
                }
            } else {
                codePointSerializer.serializeCodePoint(data, I.a);
            }
            codePointSerializer.serializeCodePoint(data, I.b);
        }
    }
}

Example 10 with Interval

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

the class ParserATNSimulator method reportContextSensitivity.

protected void reportContextSensitivity(DFA dfa, int prediction, ATNConfigSet configs, int startIndex, int stopIndex) {
    if (debug || retry_debug) {
        Interval interval = Interval.of(startIndex, stopIndex);
        System.out.println("reportContextSensitivity decision=" + dfa.decision + ":" + configs + ", input=" + parser.getTokenStream().getText(interval));
    }
    if (parser != null)
        parser.getErrorListenerDispatch().reportContextSensitivity(parser, dfa, startIndex, stopIndex, prediction, configs);
}