Examples with Automaton org.apache.lucene.util.automaton.Automaton used on opensource projects

Example 51 with Automaton

use of org.apache.lucene.util.automaton.Automaton in project lucene-solr by apache.

the class MultiTermHighlighting method extractAutomata.

/**
   * Extracts MultiTermQueries that match the provided field predicate.
   * Returns equivalent automata that will match terms.
   */
public static CharacterRunAutomaton[] extractAutomata(Query query, Predicate<String> fieldMatcher, boolean lookInSpan, Function<Query, Collection<Query>> preRewriteFunc) {
    // TODO Lucene needs a Query visitor API!  LUCENE-3041
    List<CharacterRunAutomaton> list = new ArrayList<>();
    Collection<Query> customSubQueries = preRewriteFunc.apply(query);
    if (customSubQueries != null) {
        for (Query sub : customSubQueries) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (query instanceof BooleanQuery) {
        for (BooleanClause clause : (BooleanQuery) query) {
            if (!clause.isProhibited()) {
                list.addAll(Arrays.asList(extractAutomata(clause.getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
            }
        }
    } else if (query instanceof ConstantScoreQuery) {
        list.addAll(Arrays.asList(extractAutomata(((ConstantScoreQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof BoostQuery) {
        list.addAll(Arrays.asList(extractAutomata(((BoostQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof DisjunctionMaxQuery) {
        for (Query sub : ((DisjunctionMaxQuery) query).getDisjuncts()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanOrQuery) {
        for (Query sub : ((SpanOrQuery) query).getClauses()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanNearQuery) {
        for (Query sub : ((SpanNearQuery) query).getClauses()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanNotQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanNotQuery) query).getInclude(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanPositionCheckQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanPositionCheckQuery) query).getMatch(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanBoostQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanBoostQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanMultiTermQueryWrapper) {
        list.addAll(Arrays.asList(extractAutomata(((SpanMultiTermQueryWrapper<?>) query).getWrappedQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof PrefixQuery) {
        final PrefixQuery pq = (PrefixQuery) query;
        Term prefix = pq.getPrefix();
        if (fieldMatcher.test(prefix.field())) {
            list.add(new CharacterRunAutomaton(Operations.concatenate(Automata.makeString(prefix.text()), Automata.makeAnyString())) {

                @Override
                public String toString() {
                    return pq.toString();
                }
            });
        }
    } else if (query instanceof FuzzyQuery) {
        final FuzzyQuery fq = (FuzzyQuery) query;
        if (fieldMatcher.test(fq.getField())) {
            String utf16 = fq.getTerm().text();
            int[] termText = new int[utf16.codePointCount(0, utf16.length())];
            for (int cp, i = 0, j = 0; i < utf16.length(); i += Character.charCount(cp)) {
                termText[j++] = cp = utf16.codePointAt(i);
            }
            int termLength = termText.length;
            int prefixLength = Math.min(fq.getPrefixLength(), termLength);
            String suffix = UnicodeUtil.newString(termText, prefixLength, termText.length - prefixLength);
            LevenshteinAutomata builder = new LevenshteinAutomata(suffix, fq.getTranspositions());
            String prefix = UnicodeUtil.newString(termText, 0, prefixLength);
            Automaton automaton = builder.toAutomaton(fq.getMaxEdits(), prefix);
            list.add(new CharacterRunAutomaton(automaton) {

                @Override
                public String toString() {
                    return fq.toString();
                }
            });
        }
    } else if (query instanceof TermRangeQuery) {
        final TermRangeQuery tq = (TermRangeQuery) query;
        if (fieldMatcher.test(tq.getField())) {
            final CharsRef lowerBound;
            if (tq.getLowerTerm() == null) {
                lowerBound = null;
            } else {
                lowerBound = new CharsRef(tq.getLowerTerm().utf8ToString());
            }
            final CharsRef upperBound;
            if (tq.getUpperTerm() == null) {
                upperBound = null;
            } else {
                upperBound = new CharsRef(tq.getUpperTerm().utf8ToString());
            }
            final boolean includeLower = tq.includesLower();
            final boolean includeUpper = tq.includesUpper();
            final CharsRef scratch = new CharsRef();
            @SuppressWarnings("deprecation") final Comparator<CharsRef> comparator = CharsRef.getUTF16SortedAsUTF8Comparator();
            // this is *not* an automaton, but its very simple
            list.add(new CharacterRunAutomaton(Automata.makeEmpty()) {

                @Override
                public boolean run(char[] s, int offset, int length) {
                    scratch.chars = s;
                    scratch.offset = offset;
                    scratch.length = length;
                    if (lowerBound != null) {
                        int cmp = comparator.compare(scratch, lowerBound);
                        if (cmp < 0 || (!includeLower && cmp == 0)) {
                            return false;
                        }
                    }
                    if (upperBound != null) {
                        int cmp = comparator.compare(scratch, upperBound);
                        if (cmp > 0 || (!includeUpper && cmp == 0)) {
                            return false;
                        }
                    }
                    return true;
                }

                @Override
                public String toString() {
                    return tq.toString();
                }
            });
        }
    } else if (query instanceof AutomatonQuery) {
        final AutomatonQuery aq = (AutomatonQuery) query;
        if (fieldMatcher.test(aq.getField())) {
            list.add(new CharacterRunAutomaton(aq.getAutomaton()) {

                @Override
                public String toString() {
                    return aq.toString();
                }
            });
        }
    }
    return list.toArray(new CharacterRunAutomaton[list.size()]);
}

Example 52 with Automaton

use of org.apache.lucene.util.automaton.Automaton in project lucene-solr by apache.

the class TestSimplePatternSplitTokenizer method testNotDeterminized.

public void testNotDeterminized() throws Exception {
    Automaton a = new Automaton();
    int start = a.createState();
    int mid1 = a.createState();
    int mid2 = a.createState();
    int end = a.createState();
    a.setAccept(end, true);
    a.addTransition(start, mid1, 'a', 'z');
    a.addTransition(start, mid2, 'a', 'z');
    a.addTransition(mid1, end, 'b');
    a.addTransition(mid2, end, 'b');
    expectThrows(IllegalArgumentException.class, () -> {
        new SimplePatternSplitTokenizer(a);
    });
}

Example 53 with Automaton

use of org.apache.lucene.util.automaton.Automaton in project lucene-solr by apache.

the class TestGraphTokenizers method testOverlappedTokensLattice2.

public void testOverlappedTokensLattice2() throws Exception {
    final TokenStream ts = new CannedTokenStream(new Token[] { token("abc", 1, 1), token("xyz", 0, 3), token("def", 1, 1), token("ghi", 1, 1) });
    final Automaton a1 = s2a("xyz");
    final Automaton a2 = join("abc", "def", "ghi");
    assertSameLanguage(Operations.union(a1, a2), ts);
}

Example 54 with Automaton

use of org.apache.lucene.util.automaton.Automaton in project lucene-solr by apache.

the class TestGraphTokenizers method testSynOverHole2.

public void testSynOverHole2() throws Exception {
    final TokenStream ts = new CannedTokenStream(new Token[] { token("xyz", 1, 1), token("abc", 0, 3), token("def", 2, 1) });
    final Automaton expected = Operations.union(join(s2a("xyz"), SEP_A, HOLE_A, SEP_A, s2a("def")), s2a("abc"));
    assertSameLanguage(expected, ts);
}

Example 55 with Automaton

use of org.apache.lucene.util.automaton.Automaton in project lucene-solr by apache.

the class TestGraphTokenizers method assertSameLanguage.

private void assertSameLanguage(Automaton expected, Automaton actual) {
    Automaton expectedDet = Operations.determinize(Operations.removeDeadStates(expected), DEFAULT_MAX_DETERMINIZED_STATES);
    Automaton actualDet = Operations.determinize(Operations.removeDeadStates(actual), DEFAULT_MAX_DETERMINIZED_STATES);
    if (Operations.sameLanguage(expectedDet, actualDet) == false) {
        Set<String> expectedPaths = toPathStrings(expectedDet);
        Set<String> actualPaths = toPathStrings(actualDet);
        StringBuilder b = new StringBuilder();
        b.append("expected:\n");
        for (String path : expectedPaths) {
            b.append("  ");
            b.append(path);
            if (actualPaths.contains(path) == false) {
                b.append(" [missing!]");
            }
            b.append('\n');
        }
        b.append("actual:\n");
        for (String path : actualPaths) {
            b.append("  ");
            b.append(path);
            if (expectedPaths.contains(path) == false) {
                b.append(" [unexpected!]");
            }
            b.append('\n');
        }
        fail("accepted language is different:\n\n" + b.toString());
    }
}