+ // Now they're all cached, we shouldn't have to search.
+ for (final RowBase row : deIndex.rows) {
+ if (row instanceof TokenRow) {
+ lastTokenRow = (TokenRow) row;
+ }
+ // This will break if the Row cache isn't big enough.
+ assertEquals(lastTokenRow, row.getTokenRow(false));
+ }
+
+ raf.close();
+ }
+
+ public void testChemnitz() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(TEST_OUTPUTS + "de-en.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index deIndex = dict.indices.get(0);
+
+ assertSearchResult("Höschen", "Hos", deIndex.findInsertionPoint("Hos", new AtomicBoolean(false)));
+ assertSearchResult("Höschen", "hos", deIndex.findInsertionPoint("hos", new AtomicBoolean(false)));
+
+ raf.close();
+ }
+
+ public void testMultiSearch() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(TEST_OUTPUTS + "de-en.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index deIndex = dict.indices.get(0);
+
+ {
+ final List<RowBase> rows = deIndex.multiWordSearch("aaa aaab", Arrays.asList("aaa", "aaab"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ }
+
+ raf.close();
+ }
+
+ public void testMultiSearchIt() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "IT.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index index = dict.indices.get(0);
+
+ {
+ final List<RowBase> rows = index.multiWordSearch("fare centro",
+ Arrays.asList("fare", "centro"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertTrue(rows.get(0).toString().startsWith("fare centro@"));
+ }
+ }
+
+ public void testMultiSearchDeBig() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "DE-EN.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index enIndex = dict.indices.get(1);
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("train station", Arrays.asList("train", "station"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertTrue(rows.get(0).toString().startsWith("train station@"));
+ }
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("a train station", Arrays.asList("a", "train", "station"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertEquals("Bahnhofsuhr {{de-noun|g=f|plural=Bahnhofsuhren}}\tstation clock (at a train station)", rows.get(0).toString());
+ }
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("a station", Arrays.asList("a", "station"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertEquals("Abfahrthalle {en-noun}\tDeparture room of a station.", rows.get(0).toString());
+ }
+
+ {
+ // Should print: Giving up, too many words with prefix: p
+ final List<RowBase> rows = enIndex.multiWordSearch("p eat", Arrays.asList("p", "eat"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertTrue(rows.toString().contains("verschlingen; verputzen\tto dispatch (eat)"));
+ }
+
+ {
+ // Should print: Giving up, too many words with prefix: p
+ final List<RowBase> rows = enIndex.multiWordSearch("p p", Arrays.asList("p", "p"), new AtomicBoolean(false));
+ assertTrue(rows.size() >= 1000);
+ }
+
+ {
+ // Should print: Giving up, too many words with prefix: a
+ final List<RowBase> rows = enIndex.multiWordSearch("a a", Arrays.asList("a", "a"), new AtomicBoolean(false));
+ assertTrue(rows.size() >= 1000);
+ }
+
+ {
+ // Should print: Giving up, too many words with prefix: a
+ final List<RowBase> rows = enIndex.multiWordSearch("b ba", Arrays.asList("b", "ba"), new AtomicBoolean(false));
+ assertTrue(rows.size() >= 1000);
+ }
+
+ {
+ // Should print: Giving up, too many words with prefix: a
+ final List<RowBase> rows = enIndex.multiWordSearch("b ba", Arrays.asList("b", "ba"), new AtomicBoolean(false));
+ assertTrue(rows.size() >= 1000);
+ }
+
+ raf.close();
+ }
+
+ public void testMultiSearchBigAF() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "AF-EN.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index enIndex = dict.indices.get(1);
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("pig eats", Arrays.asList("pig", "eats"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertEquals("vark\tpig (someone who overeats or eats rapidly) (noun)", rows.get(0).toString());
+ }
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("pig eat", Arrays.asList("pig", "eat"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertEquals("vark\tpig (someone who overeats or eats rapidly) (noun)", rows.get(0).toString());
+ }
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("pi ea", Arrays.asList("pi", "ea"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertTrue(rows.toString().contains("vark\tpig (someone who overeats or eats rapidly) (noun)"));
+ }
+
+ {
+ final List<RowBase> rows = enIndex.multiWordSearch("p eat", Arrays.asList("p", "eat"), new AtomicBoolean(false));
+ System.out.println(CollectionUtil.join(rows, "\n "));
+ assertTrue(rows.toString(), rows.size() > 0);
+ assertTrue(rows.toString().contains("vark\tpig (someone who overeats or eats rapidly) (noun)"));
+ }
+
+
+ raf.close();
+ }
+
+
+ public void testExactSearch() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "EN-cmn.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index cmnIndex = dict.indices.get(1);
+
+ final Random random = new Random(10);
+
+ for (int i = 0; i < 1000; ++i) {
+ final int ii = random.nextInt(cmnIndex.sortedIndexEntries.size());
+ final IndexEntry indexEntry = cmnIndex.sortedIndexEntries.get(ii);
+ final IndexEntry found = cmnIndex.findExact(indexEntry.token);
+ assertNotNull(found);
+ assertEquals(indexEntry.token, found.token);
+ assertEquals(indexEntry, found); // Test of caching....
+ }
+
+ raf.close();
+ }
+
+ public void testThai() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "EN-TH.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index thIndex = dict.indices.get(1);
+
+ final IndexEntry entry = thIndex.findInsertionPoint("ดี", new AtomicBoolean(false));
+ assertEquals("di", entry.token);
+
+ raf.close();
+ }
+
+ public void testNorwegian() throws IOException {
+ final RandomAccessFile raf = new RandomAccessFile(OUTPUTS + "EN-NL.quickdic", "r");
+ final Dictionary dict = new Dictionary(raf.getChannel());
+ final Index nlIndex = dict.indices.get(1);
+
+ IndexEntry entry = nlIndex.findInsertionPoint("Xhosa", new AtomicBoolean(false));
+ assertEquals("Xhosa", entry.token);
+
+ entry = nlIndex.findInsertionPoint("Zyne", new AtomicBoolean(false));
+ assertEquals("Zyne", entry.token);
+
+ raf.close();
+ }