Adding upstream version 1:7.0.4.upstream/1%7.0.4upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

5 files changed, 1803 insertions, 0 deletions

diff --git a/i18npool/qa/cppunit/test_breakiterator.cxx b/i18npool/qa/cppunit/test_breakiterator.cxx
new file mode 100644
index 000000000..a41cd8218
--- /dev/null
+++ b/i18npool/qa/cppunit/test_breakiterator.cxx
@@ -0,0 +1,1065 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ */
+
+#include <com/sun/star/i18n/XBreakIterator.hpp>
+#include <com/sun/star/i18n/CharacterIteratorMode.hpp>
+#include <com/sun/star/i18n/ScriptType.hpp>
+#include <com/sun/star/i18n/WordType.hpp>
+#include <o3tl/cppunittraitshelper.hxx>
+#include <unotest/bootstrapfixturebase.hxx>
+
+#include <unicode/uvernum.h>
+
+#include <rtl/strbuf.hxx>
+
+#include <string.h>
+
+#include <stack>
+
+using namespace ::com::sun::star;
+
+class TestBreakIterator : public test::BootstrapFixtureBase
+{
+public:
+    virtual void setUp() override;
+    virtual void tearDown() override;
+
+    void testLineBreaking();
+    void testWordBoundaries();
+    void testGraphemeIteration();
+    void testWeak();
+    void testAsian();
+    void testThai();
+#if (U_ICU_VERSION_MAJOR_NUM > 51)
+    void testLao();
+#ifdef TODO
+    void testNorthernThai();
+    void testKhmer();
+#endif
+#endif
+    void testJapanese();
+    void testChinese();
+
+    CPPUNIT_TEST_SUITE(TestBreakIterator);
+    CPPUNIT_TEST(testLineBreaking);
+    CPPUNIT_TEST(testWordBoundaries);
+    CPPUNIT_TEST(testGraphemeIteration);
+    CPPUNIT_TEST(testWeak);
+    CPPUNIT_TEST(testAsian);
+    CPPUNIT_TEST(testThai);
+#if (U_ICU_VERSION_MAJOR_NUM > 51)
+    CPPUNIT_TEST(testLao);
+#ifdef TODO
+    CPPUNIT_TEST(testKhmer);
+    CPPUNIT_TEST(testNorthernThai);
+#endif
+#endif
+    CPPUNIT_TEST(testJapanese);
+    CPPUNIT_TEST(testChinese);
+    CPPUNIT_TEST_SUITE_END();
+
+private:
+    uno::Reference<i18n::XBreakIterator> m_xBreak;
+    void doTestJapanese(uno::Reference< i18n::XBreakIterator > const &xBreak);
+};
+
+void TestBreakIterator::testLineBreaking()
+{
+    i18n::LineBreakHyphenationOptions aHyphOptions;
+    i18n::LineBreakUserOptions aUserOptions;
+    lang::Locale aLocale;
+
+    //See https://bugs.libreoffice.org/show_bug.cgi?id=31271
+    {
+        OUString aTest("(some text here)");
+
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        {
+            //Here we want the line break to leave text here) on the next line
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, strlen("(some tex"), aLocale, 0, aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break at the start of the word", static_cast<sal_Int32>(6), aResult.breakIndex);
+        }
+
+        {
+            //Here we want the line break to leave "here)" on the next line
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, strlen("(some text here"), aLocale, 0, aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break at the start of the word", static_cast<sal_Int32>(11), aResult.breakIndex);
+        }
+    }
+
+    //See https://bugs.libreoffice.org/show_bug.cgi?id=49849
+    {
+        const sal_Unicode HEBREW1[] = { 0x05DE, 0x05D9, 0x05DC, 0x05D9, 0x5DD };
+        OUString aWord(HEBREW1, SAL_N_ELEMENTS(HEBREW1));
+        OUString aTest(aWord + " " + aWord);
+
+        aLocale.Language = "he";
+        aLocale.Country = "IL";
+
+        {
+            //Here we want the line break to happen at the whitespace
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, aTest.getLength()-1, aLocale, 0, aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break at the start of the word", aWord.getLength()+1, aResult.breakIndex);
+        }
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=17155
+    {
+        OUString const aTest("foo /bar/baz");
+
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        {
+            //Here we want the line break to leave /bar/ba clumped together on the next line
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, strlen("foo /bar/ba"), aLocale, 0,
+                aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break at the first slash", static_cast<sal_Int32>(4), aResult.breakIndex);
+        }
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=19716
+    {
+        OUString aTest("aaa]aaa");
+
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        {
+            //Here we want the line break to move the whole lot to the next line
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, aTest.getLength()-2, aLocale, 0,
+                aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break at the start of the line, not at ]", static_cast<sal_Int32>(0), aResult.breakIndex);
+        }
+    }
+
+    //this is an example sequence from tdf92993-1.docx caught by the load crashtesting
+    {
+        const sal_Unicode WEIRD1[] = { 0xd83c, 0xdf56, 0xd83c, 0xdf57, 0xd83c, 0xdf46,
+                                       0xd83c, 0xdf64, 0x2668, 0xfe0f, 0xd83c, 0xdfc6};
+
+        OUString aTest(WEIRD1, SAL_N_ELEMENTS(WEIRD1));
+
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        {
+            //This must not assert/crash
+            (void)m_xBreak->getLineBreak(aTest, 0, aLocale, 0, aHyphOptions, aUserOptions);
+        }
+    }
+
+    //See https://bugs.documentfoundation.org/show_bug.cgi?id=96197
+    {
+        const sal_Unicode HANGUL[] = { 0xc560, 0xad6D, 0xac00, 0xc758, 0x0020, 0xac00,
+                                       0xc0ac, 0xb294};
+        OUString aTest(HANGUL, SAL_N_ELEMENTS(HANGUL));
+
+        aLocale.Language = "ko";
+        aLocale.Country = "KR";
+
+        {
+            i18n::LineBreakResults aResult = m_xBreak->getLineBreak(aTest, aTest.getLength()-2, aLocale, 0,
+                aHyphOptions, aUserOptions);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Expected a break don't split the Korean word!", static_cast<sal_Int32>(5), aResult.breakIndex);
+        }
+    }
+}
+
+//See https://bugs.libreoffice.org/show_bug.cgi?id=49629
+void TestBreakIterator::testWordBoundaries()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "en";
+    aLocale.Country = "US";
+
+    i18n::Boundary aBounds;
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=11993
+    {
+        OUString aTest("abcd ef  ghi??? KLM");
+
+        CPPUNIT_ASSERT(!m_xBreak->isBeginWord(aTest, 4, aLocale, i18n::WordType::DICTIONARY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 4, aLocale, i18n::WordType::DICTIONARY_WORD));
+        aBounds = m_xBreak->getWordBoundary(aTest, 4, aLocale, i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 4);
+
+        CPPUNIT_ASSERT(!m_xBreak->isBeginWord(aTest, 8, aLocale, i18n::WordType::DICTIONARY_WORD));
+        CPPUNIT_ASSERT(!m_xBreak->isEndWord(aTest, 8, aLocale, i18n::WordType::DICTIONARY_WORD));
+
+        //next word
+        aBounds = m_xBreak->getWordBoundary(aTest, 8, aLocale, i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 9 && aBounds.endPos == 12);
+
+        //previous word
+        aBounds = m_xBreak->getWordBoundary(aTest, 8, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 5 && aBounds.endPos == 7);
+
+        CPPUNIT_ASSERT(!m_xBreak->isBeginWord(aTest, 12, aLocale, i18n::WordType::DICTIONARY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 12, aLocale, i18n::WordType::DICTIONARY_WORD));
+        aBounds = m_xBreak->getWordBoundary(aTest, 12, aLocale, i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 9 && aBounds.endPos == 12);
+
+        CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, 16, aLocale, i18n::WordType::DICTIONARY_WORD));
+        CPPUNIT_ASSERT(!m_xBreak->isEndWord(aTest, 16, aLocale, i18n::WordType::DICTIONARY_WORD));
+        aBounds = m_xBreak->getWordBoundary(aTest, 16, aLocale, i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 16 && aBounds.endPos == 19);
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=21907
+    {
+        OUString aTest("b a?");
+
+        CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, 1, aLocale, i18n::WordType::ANY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, 2, aLocale, i18n::WordType::ANY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, 3, aLocale, i18n::WordType::ANY_WORD));
+
+        CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, 3, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES));
+
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 1, aLocale, i18n::WordType::ANY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 2, aLocale, i18n::WordType::ANY_WORD));
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 3, aLocale, i18n::WordType::ANY_WORD));
+
+        CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, 3, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES));
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=14904
+    {
+        const sal_Unicode TEST[] =
+        {
+            'W', 'o', 'r', 'k', 'i', 'n', 'g', ' ', 0x201C, 'W', 'o', 'r', 'd', 's',
+            ' ', 's', 't', 'a', 'r', 't', 'i', 'n', 'g', ' ', 'w', 'i', 't',
+            'h', ' ', 'q', 'u', 'o', 't', 'e', 's', 0x201D, ' ', 'W', 'o', 'r', 'k',
+            'i', 'n', 'g', ' ', 0x2018, 'B', 'r', 'o', 'k', 'e', 'n', 0x2019, ' ',
+            '?', 'S', 'p', 'a', 'n', 'i', 's', 'h', '?', ' ', 'd', 'o', 'e',
+            's', 'n', 0x2019, 't', ' ', 'w', 'o', 'r', 'k', '.', ' ', 'N', 'o',
+            't', ' ', 'e', 'v', 'e', 'n', ' ' , 0x00BF, 'r', 'e', 'a', 'l', '?', ' ',
+            'S', 'p', 'a', 'n', 'i', 's', 'h'
+        };
+        OUString aTest(TEST, SAL_N_ELEMENTS(TEST));
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 4, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 7);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 12, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 9 && aBounds.endPos == 14);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 40, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 37 && aBounds.endPos == 44);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 49, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 46 && aBounds.endPos == 52);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 58, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 55 && aBounds.endPos == 62);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 67, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 64 && aBounds.endPos == 71);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 90, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 88 && aBounds.endPos == 92);
+    }
+
+    //See https://bugs.libreoffice.org/show_bug.cgi?id=49629
+    sal_Unicode aBreakTests[] = { ' ', 1, 2, 3, 4, 5, 6, 7, 0x91, 0x92, 0x200B, 0xE8FF, 0xF8FF };
+    for (int mode = i18n::WordType::ANY_WORD; mode <= i18n::WordType::WORD_COUNT; ++mode)
+    {
+        //make sure that in all cases isBeginWord and isEndWord matches getWordBoundary
+        for (size_t i = 0; i < SAL_N_ELEMENTS(aBreakTests); ++i)
+        {
+#if (U_ICU_VERSION_MAJOR_NUM == 4) && (U_ICU_VERSION_MINOR_NUM <= 2)
+            //Note the breakiterator test is known to fail on older icu
+            //versions (4.2.1) for the 200B (ZWSP) Zero Width Space testcase.
+            if (aBreakTests[i] == 0x200B)
+                continue;
+#endif
+            OUString aTest = "Word" + OUStringChar(aBreakTests[i]) + "Word";
+            aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale, mode, true);
+            switch (mode)
+            {
+                case i18n::WordType::ANY_WORD:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 4);
+                    break;
+                case i18n::WordType::ANYWORD_IGNOREWHITESPACES:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 4);
+                    break;
+                case i18n::WordType::DICTIONARY_WORD:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 4);
+                    break;
+                case i18n::WordType::WORD_COUNT:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 4);
+                    break;
+            }
+
+            CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, aBounds.startPos, aLocale, mode));
+            CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, aBounds.endPos, aLocale, mode));
+        }
+    }
+
+    sal_Unicode aJoinTests[] = { 'X', 0x200C, 0x200D, 0x2060, 0xFEFF, 0xFFF9, 0xFFFA, 0xFFFB };
+    for (int mode = i18n::WordType::ANY_WORD; mode <= i18n::WordType::WORD_COUNT; ++mode)
+    {
+        //make sure that in all cases isBeginWord and isEndWord matches getWordBoundary
+        for (size_t i = 0; i < SAL_N_ELEMENTS(aJoinTests); ++i)
+        {
+            OUString aTest = "Word" + OUStringChar(aJoinTests[i]) + "Word";
+            aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale, mode, true);
+            switch (mode)
+            {
+                case i18n::WordType::ANY_WORD:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 9);
+                    break;
+                case i18n::WordType::ANYWORD_IGNOREWHITESPACES:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 9);
+                    break;
+                case i18n::WordType::DICTIONARY_WORD:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 9);
+                    break;
+                case i18n::WordType::WORD_COUNT:
+                    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 9);
+                    break;
+            }
+
+            CPPUNIT_ASSERT(m_xBreak->isBeginWord(aTest, aBounds.startPos, aLocale, mode));
+            CPPUNIT_ASSERT(m_xBreak->isEndWord(aTest, aBounds.endPos, aLocale, mode));
+        }
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=13494
+    {
+        const OUString aBase("xxAAxxBBxxCCxx");
+        const sal_Unicode aTests[] =
+        {
+            '\'', ';', ',', '.', '!', '@', '#', '%', '&', '*',
+            '(', ')', '_', '-', '{', '}', '[', ']', '\"', '/',
+            '\\', '?', '~', '$', '+', '^', '=', '<', '>', '|'
+        };
+
+        const sal_Int32 aDoublePositions[] = {0, 2, 4, 6, 8, 10, 12, 14};
+        for (size_t j = 0; j < SAL_N_ELEMENTS(aTests); ++j)
+        {
+            OUString aTest = aBase.replace('x', aTests[j]);
+            sal_Int32 nPos = -1;
+            size_t i = 0;
+            do
+            {
+                CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aDoublePositions));
+                nPos = m_xBreak->nextWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                CPPUNIT_ASSERT_EQUAL(aDoublePositions[i], nPos);
+                ++i;
+            }
+            while (nPos < aTest.getLength());
+            nPos = aTest.getLength();
+            i = SAL_N_ELEMENTS(aDoublePositions)-1;
+            do
+            {
+                nPos = m_xBreak->previousWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                --i;
+                CPPUNIT_ASSERT_EQUAL(aDoublePositions[i], nPos);
+            }
+            while (nPos > 0);
+        }
+
+        const sal_Int32 aSinglePositions[] = {0, 1, 3, 4, 6, 7, 9, 10};
+        for (size_t j = 1; j < SAL_N_ELEMENTS(aTests); ++j)
+        {
+            OUString aTest = aBase.replaceAll("xx", OUStringChar(aTests[j]));
+            sal_Int32 nPos = -1;
+            size_t i = 0;
+            do
+            {
+                CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aSinglePositions));
+                nPos = m_xBreak->nextWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                CPPUNIT_ASSERT_EQUAL(aSinglePositions[i], nPos);
+                ++i;
+            }
+            while (nPos < aTest.getLength());
+            nPos = aTest.getLength();
+            i = SAL_N_ELEMENTS(aSinglePositions)-1;
+            do
+            {
+                nPos = m_xBreak->previousWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                --i;
+                CPPUNIT_ASSERT_EQUAL(aSinglePositions[i], nPos);
+            }
+            while (nPos > 0);
+        }
+
+        const sal_Int32 aSingleQuotePositions[] = {0, 1, 9, 10};
+        CPPUNIT_ASSERT_EQUAL(u'\'', aTests[0]);
+        {
+            OUString aTest = aBase.replaceAll("xx", OUStringChar(aTests[0]));
+            sal_Int32 nPos = -1;
+            size_t i = 0;
+            do
+            {
+                CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aSingleQuotePositions));
+                nPos = m_xBreak->nextWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                CPPUNIT_ASSERT_EQUAL(aSingleQuotePositions[i], nPos);
+                ++i;
+            }
+            while (nPos < aTest.getLength());
+            nPos = aTest.getLength();
+            i = SAL_N_ELEMENTS(aSingleQuotePositions)-1;
+            do
+            {
+                nPos = m_xBreak->previousWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+                --i;
+                CPPUNIT_ASSERT_EQUAL(aSingleQuotePositions[i], nPos);
+            }
+            while (nPos > 0);
+        }
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=13451
+    {
+        aLocale.Language = "ca";
+        aLocale.Country = "ES";
+
+        OUString aTest("mirar-se comprar-vos donem-nos les mans aneu-vos-en!");
+
+        sal_Int32 nPos = 0;
+        sal_Int32 aExpected[] = {8, 20, 30, 34, 39, 51, 52};
+        size_t i = 0;
+        do
+        {
+            CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aExpected));
+            nPos = m_xBreak->getWordBoundary(aTest, nPos, aLocale,
+                i18n::WordType::DICTIONARY_WORD, true).endPos;
+            CPPUNIT_ASSERT_EQUAL(aExpected[i], nPos);
+            ++i;
+        }
+        while (nPos++ < aTest.getLength());
+        CPPUNIT_ASSERT_EQUAL(SAL_N_ELEMENTS(aExpected), i);
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=85411
+    for (int j = 0; j < 3; ++j)
+    {
+        switch (j)
+        {
+            case 0:
+                aLocale.Language = "en";
+                aLocale.Country = "US";
+                break;
+            case 1:
+                aLocale.Language = "ca";
+                aLocale.Country = "ES";
+                break;
+            case 2:
+                aLocale.Language = "fi";
+                aLocale.Country = "FI";
+                break;
+            default:
+                CPPUNIT_ASSERT(false);
+                break;
+        }
+
+        const sal_Unicode TEST[] =
+        {
+            'I', 0x200B, 'w', 'a', 'n', 't', 0x200B, 't', 'o', 0x200B, 'g', 'o'
+        };
+        OUString aTest(TEST, SAL_N_ELEMENTS(TEST));
+
+        sal_Int32 nPos = 0;
+        sal_Int32 aExpected[] = {1, 6, 9, 12};
+        size_t i = 0;
+        do
+        {
+            CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aExpected));
+            nPos = m_xBreak->getWordBoundary(aTest, nPos, aLocale,
+                i18n::WordType::DICTIONARY_WORD, true).endPos;
+            CPPUNIT_ASSERT_EQUAL(aExpected[i], nPos);
+            ++i;
+        }
+        while (nPos++ < aTest.getLength());
+        CPPUNIT_ASSERT_EQUAL(SAL_N_ELEMENTS(aExpected), i);
+    }
+
+    //https://bz.apache.org/ooo/show_bug.cgi?id=21290
+    for (int j = 0; j < 2; ++j)
+    {
+        switch (j)
+        {
+            case 0:
+                aLocale.Language = "en";
+                aLocale.Country = "US";
+                break;
+            case 1:
+                aLocale.Language = "grc";
+                aLocale.Country.clear();
+                break;
+            default:
+                CPPUNIT_ASSERT(false);
+                break;
+        }
+
+        const sal_Unicode TEST[] =
+        {
+            0x1F0C, 0x03BD, 0x03B4, 0x03C1, 0x03B1, 0x0020, 0x1F00,
+            0x03C1, 0x03BD, 0x1F7B, 0x03BC, 0x03B5, 0x03BD, 0x03BF,
+            0x03C2, 0x0020, 0x1F00, 0x03BB, 0x03BB, 0x0020, 0x1F24,
+            0x03C3, 0x03B8, 0x03B9, 0x03BF, 0x03BD
+        };
+        OUString aTest(TEST, SAL_N_ELEMENTS(TEST));
+
+        sal_Int32 nPos = 0;
+        sal_Int32 aExpected[] = {5, 15, 19, 26};
+        size_t i = 0;
+        do
+        {
+            CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aExpected));
+            nPos = m_xBreak->getWordBoundary(aTest, nPos, aLocale,
+                i18n::WordType::DICTIONARY_WORD, true).endPos;
+            CPPUNIT_ASSERT_EQUAL(aExpected[i], nPos);
+            ++i;
+        }
+        while (nPos++ < aTest.getLength());
+        CPPUNIT_ASSERT_EQUAL(SAL_N_ELEMENTS(aExpected), i);
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=58513
+    //See https://bugs.libreoffice.org/show_bug.cgi?id=55707
+    {
+        aLocale.Language = "fi";
+        aLocale.Country = "FI";
+
+        OUString aTest("Kuorma-auto kaakkois- ja Keski-Suomi USA:n 90:n %:n");
+
+        {
+            sal_Int32 nPos = 0;
+            sal_Int32 aExpected[] = {11, 21, 24, 36, 42, 47, 51};
+            size_t i = 0;
+            do
+            {
+                CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aExpected));
+                nPos = m_xBreak->getWordBoundary(aTest, nPos, aLocale,
+                    i18n::WordType::WORD_COUNT, true).endPos;
+                CPPUNIT_ASSERT_EQUAL(aExpected[i], nPos);
+                ++i;
+            }
+            while (nPos++ < aTest.getLength());
+            CPPUNIT_ASSERT_EQUAL(SAL_N_ELEMENTS(aExpected), i);
+        }
+
+        {
+            sal_Int32 nPos = 0;
+            sal_Int32 aExpected[] = {0, 11, 12, 20, 22, 24, 25, 36, 37,
+                                    40, 41, 42, 43, 45, 46, 47, 50, 51};
+            size_t i = 0;
+            do
+            {
+                CPPUNIT_ASSERT(i < SAL_N_ELEMENTS(aExpected));
+                aBounds = m_xBreak->getWordBoundary(aTest, nPos, aLocale,
+                    i18n::WordType::DICTIONARY_WORD, true);
+                CPPUNIT_ASSERT_EQUAL(aExpected[i], aBounds.startPos);
+                ++i;
+                CPPUNIT_ASSERT_EQUAL(aExpected[i], aBounds.endPos);
+                ++i;
+                nPos = aBounds.endPos;
+            }
+            while (nPos++ < aTest.getLength());
+            CPPUNIT_ASSERT_EQUAL(SAL_N_ELEMENTS(aExpected), i);
+        }
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=107843
+    {
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        const sal_Unicode TEST[] =
+        {
+            'r', 'u', 0xFB00, 'l', 'e', ' ', 0xFB01, 's', 'h'
+        };
+        OUString aTest(TEST, SAL_N_ELEMENTS(TEST));
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 1, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 5);
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 7, aLocale, i18n::WordType::DICTIONARY_WORD, false);
+        CPPUNIT_ASSERT(aBounds.startPos == 6 && aBounds.endPos == 9);
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=113785
+    {
+        aLocale.Language = "en";
+        aLocale.Country = "US";
+
+        const sal_Unicode TEST[] =
+        {
+            'a', 0x2013, 'b', 0x2014, 'c'
+        };
+        OUString aTest(TEST, SAL_N_ELEMENTS(TEST));
+
+        aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale, i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 1);
+
+        aBounds = m_xBreak->nextWord(aTest, 0, aLocale, i18n::WordType::DICTIONARY_WORD);
+        CPPUNIT_ASSERT(aBounds.startPos == 2 && aBounds.endPos == 3);
+
+        aBounds = m_xBreak->nextWord(aTest, aBounds.endPos, aLocale, i18n::WordType::DICTIONARY_WORD);
+        CPPUNIT_ASSERT(aBounds.startPos == 4 && aBounds.endPos == 5);
+    }
+}
+
+//See https://bugs.libreoffice.org/show_bug.cgi?id=40292
+//See https://bz.apache.org/ooo/show_bug.cgi?id=80412
+//See https://bz.apache.org/ooo/show_bug.cgi?id=111152
+//See https://bz.apache.org/ooo/show_bug.cgi?id=50172
+void TestBreakIterator::testGraphemeIteration()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "bn";
+    aLocale.Country = "IN";
+
+    {
+        const sal_Unicode BA_HALANT_LA[] = { 0x09AC, 0x09CD, 0x09AF };
+        OUString aTest(BA_HALANT_LA, SAL_N_ELEMENTS(BA_HALANT_LA));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos;
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(BA_HALANT_LA)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(BA_HALANT_LA), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+
+    {
+        const sal_Unicode HA_HALANT_NA_VOWELSIGNI[] = { 0x09B9, 0x09CD, 0x09A3, 0x09BF };
+        OUString aTest(HA_HALANT_NA_VOWELSIGNI, SAL_N_ELEMENTS(HA_HALANT_NA_VOWELSIGNI));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos;
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(HA_HALANT_NA_VOWELSIGNI)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(HA_HALANT_NA_VOWELSIGNI), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+
+    {
+        const sal_Unicode TA_HALANT_MA_HALANT_YA  [] = { 0x09A4, 0x09CD, 0x09AE, 0x09CD, 0x09AF };
+        OUString aTest(TA_HALANT_MA_HALANT_YA, SAL_N_ELEMENTS(TA_HALANT_MA_HALANT_YA));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos;
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(TA_HALANT_MA_HALANT_YA)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(TA_HALANT_MA_HALANT_YA), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+
+    aLocale.Language = "ta";
+    aLocale.Country = "IN";
+
+    {
+        const sal_Unicode KA_VIRAMA_SSA[] = { 0x0B95, 0x0BCD, 0x0BB7 };
+        OUString aTest(KA_VIRAMA_SSA, SAL_N_ELEMENTS(KA_VIRAMA_SSA));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos = 0;
+
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(KA_VIRAMA_SSA)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(KA_VIRAMA_SSA), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+
+    {
+        const sal_Unicode KA_VOWELSIGNU[] = { 0x0B95, 0x0BC1 };
+        OUString aTest(KA_VOWELSIGNU, SAL_N_ELEMENTS(KA_VOWELSIGNU));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos = 0;
+
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(KA_VOWELSIGNU)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(KA_VOWELSIGNU), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+
+    {
+        const sal_Unicode CA_VOWELSIGNI_TA_VIRAMA_TA_VOWELSIGNI_RA_VOWELSIGNAI[] =
+            { 0x0B9A, 0x0BBF, 0x0BA4, 0x0BCD, 0x0BA4, 0x0BBF, 0x0BB0, 0x0BC8 };
+        OUString aTest(CA_VOWELSIGNI_TA_VIRAMA_TA_VOWELSIGNI_RA_VOWELSIGNAI,
+            SAL_N_ELEMENTS(CA_VOWELSIGNI_TA_VIRAMA_TA_VOWELSIGNI_RA_VOWELSIGNAI));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos=0;
+
+        for (sal_Int32 i = 0; i < 4; ++i)
+        {
+            sal_Int32 nOldPos = nPos;
+            nPos = m_xBreak->nextCharacters(aTest, nPos, aLocale,
+                i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip 2 units", nOldPos+2, nPos);
+        }
+
+        for (sal_Int32 i = 0; i < 4; ++i)
+        {
+            sal_Int32 nOldPos = nPos;
+            nPos = m_xBreak->previousCharacters(aTest, nPos, aLocale,
+                i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+            CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip 2 units", nOldPos-2, nPos);
+        }
+    }
+
+    {
+        const sal_Unicode ALEF_QAMATS [] = { 0x05D0, 0x05B8 };
+        OUString aText(ALEF_QAMATS, SAL_N_ELEMENTS(ALEF_QAMATS));
+
+        sal_Int32 nGraphemeCount = 0;
+
+        sal_Int32 nCurPos = 0;
+        while (nCurPos < aText.getLength())
+        {
+            sal_Int32 nCount2 = 1;
+            nCurPos = m_xBreak->nextCharacters(aText, nCurPos, lang::Locale(),
+                i18n::CharacterIteratorMode::SKIPCELL, nCount2, nCount2);
+            ++nGraphemeCount;
+        }
+
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should be considered 1 grapheme", static_cast<sal_Int32>(1), nGraphemeCount);
+    }
+
+    aLocale.Language = "hi";
+    aLocale.Country = "IN";
+
+    {
+        const sal_Unicode SHA_VOWELSIGNII[] = { 0x936, 0x940 };
+        OUString aTest(SHA_VOWELSIGNII, SAL_N_ELEMENTS(SHA_VOWELSIGNII));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos = 0;
+
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(SAL_N_ELEMENTS(SHA_VOWELSIGNII)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(SHA_VOWELSIGNII), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full grapheme", static_cast<sal_Int32>(0), nPos);
+    }
+}
+
+//A test to ensure that certain ranges and codepoints that are categorized as
+//weak remain as weak, so that existing docs that depend on this don't silently
+//change font for those weak chars
+void TestBreakIterator::testWeak()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "en";
+    aLocale.Country = "US";
+
+    {
+        const sal_Unicode WEAKS[] =
+        {
+            0x0001, 0x0002,
+            0x0020, 0x00A0,
+            0x0300, 0x036F, //Combining Diacritical Marks
+            0x1AB0, 0x1AFF, //Combining Diacritical Marks Extended
+            0x1DC0, 0x1DFF, //Combining Diacritical Marks Supplement
+            0x20D0, 0x20FF, //Combining Diacritical Marks for Symbols
+            0x2150, 0x215F, //Number Forms, fractions
+            0x2160, 0x2180, //Number Forms, roman numerals
+            0x2200, 0x22FF, //Mathematical Operators
+            0x27C0, 0x27EF, //Miscellaneous Mathematical Symbols-A
+            0x2980, 0x29FF, //Miscellaneous Mathematical Symbols-B
+            0x2A00, 0x2AFF, //Supplemental Mathematical Operators
+            0x2100, 0x214F, //Letterlike Symbols
+            0x2308, 0x230B, //Miscellaneous technical
+            0x25A0, 0x25FF, //Geometric Shapes
+            0x2B30, 0x2B4C  //Miscellaneous Symbols and Arrows
+        };
+        OUString aWeaks(WEAKS, SAL_N_ELEMENTS(WEAKS));
+
+        for (sal_Int32 i = 0; i < aWeaks.getLength(); ++i)
+        {
+            sal_Int16 nScript = m_xBreak->getScriptType(aWeaks, i);
+            OString aMsg =
+                "Char 0x" +
+                OString::number(static_cast<sal_Int32>(aWeaks[i]), 16) +
+                " should have been weak";
+            CPPUNIT_ASSERT_EQUAL_MESSAGE(aMsg.getStr(),
+                i18n::ScriptType::WEAK, nScript);
+        }
+    }
+}
+
+//A test to ensure that certain ranges and codepoints that are categorized as
+//asian remain as asian, so that existing docs that depend on this don't silently
+//change font for those asian chars.
+//See https://bugs.libreoffice.org/show_bug.cgi?id=38095
+void TestBreakIterator::testAsian()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "en";
+    aLocale.Country = "US";
+
+    {
+        const sal_Unicode ASIANS[] =
+        {
+            //some typical CJK chars
+            0x4E00, 0x62FF,
+            //The full HalfWidth and FullWidth block has historically been
+            //designated as taking the CJK font :-(
+            //HalfWidth and FullWidth forms of ASCII 0-9, categorized under
+            //UAX24 as "Common" i.e. by that logic WEAK
+            0xFF10, 0xFF19,
+            //HalfWidth and FullWidth forms of ASCII A-z, categorized under
+            //UAX25 as "Latin", i.e. by that logic LATIN
+            0xFF21, 0xFF5A
+        };
+        OUString aAsians(ASIANS, SAL_N_ELEMENTS(ASIANS));
+
+        for (sal_Int32 i = 0; i < aAsians.getLength(); ++i)
+        {
+            sal_Int16 nScript = m_xBreak->getScriptType(aAsians, i);
+            OString aMsg =
+                "Char 0x" +
+                OString::number(static_cast<sal_Int32>(aAsians[i]), 16) +
+                " should have been asian";
+            CPPUNIT_ASSERT_EQUAL_MESSAGE(aMsg.getStr(),
+                i18n::ScriptType::ASIAN, nScript);
+        }
+    }
+}
+
+#if (U_ICU_VERSION_MAJOR_NUM > 51)
+//A test to ensure that our Lao word boundary detection is useful
+void TestBreakIterator::testLao()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "lo";
+    aLocale.Country = "LA";
+
+    const sal_Unicode LAO[] = { 0x0e8d, 0x0eb4, 0x0e99, 0x0e94, 0x0eb5, 0x0e95, 0x0ec9, 0x0ead, 0x0e99, 0x0eae, 0x0eb1, 0x0e9a };
+    OUString aTest(LAO, SAL_N_ELEMENTS(LAO));
+    i18n::Boundary aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale,
+        i18n::WordType::DICTIONARY_WORD, true);
+
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(0), aBounds.startPos);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(5), aBounds.endPos);
+
+    aBounds = m_xBreak->getWordBoundary(aTest, aBounds.endPos, aLocale,
+        i18n::WordType::DICTIONARY_WORD, true);
+
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(5), aBounds.startPos);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(9), aBounds.endPos);
+}
+#endif
+
+//A test to ensure that our thai word boundary detection is useful
+void TestBreakIterator::testThai()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "th";
+    aLocale.Country = "TH";
+
+    //See http://lists.freedesktop.org/archives/libreoffice/2012-February/025959.html
+    {
+        const sal_Unicode THAI[] = { 0x0E01, 0x0E38, 0x0E2B, 0x0E25, 0x0E32, 0x0E1A };
+        OUString aTest(THAI, SAL_N_ELEMENTS(THAI));
+        i18n::Boundary aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale,
+            i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT_MESSAGE("Should skip full word",
+            aBounds.startPos == 0 && aBounds.endPos == aTest.getLength());
+    }
+
+    //See https://bz.apache.org/ooo/show_bug.cgi?id=29548
+    //make sure forwards and back are consistent
+    {
+        const sal_Unicode THAI[] =
+        {
+            0x0E2D, 0x0E38, 0x0E17, 0x0E22, 0x0E32, 0x0E19, 0x0E41,
+            0x0E2B, 0x0E48, 0x0E07, 0x0E0A, 0x0E32, 0x0E15, 0x0E34,
+            0x0E19, 0x0E49, 0x0E33, 0x0E2B, 0x0E19, 0x0E32, 0x0E27,
+            0x0E2D, 0x0E38, 0x0E17, 0x0E22, 0x0E32, 0x0E19, 0x0E41,
+            0x0E2B, 0x0E48, 0x0E07, 0x0E0A, 0x0E32, 0x0E15, 0x0E34,
+            0x0E19, 0x0E49, 0x0E33, 0x0E2B, 0x0E19, 0x0E32, 0x0E27
+        };
+        OUString aTest(THAI, SAL_N_ELEMENTS(THAI));
+
+        std::stack<sal_Int32> aPositions;
+        sal_Int32 nPos = -1;
+        do
+        {
+            nPos = m_xBreak->nextWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+            aPositions.push(nPos);
+        }
+        while (nPos < aTest.getLength());
+        nPos = aTest.getLength();
+        CPPUNIT_ASSERT(!aPositions.empty());
+        aPositions.pop();
+        do
+        {
+            CPPUNIT_ASSERT(!aPositions.empty());
+            nPos = m_xBreak->previousWord(aTest, nPos, aLocale, i18n::WordType::ANYWORD_IGNOREWHITESPACES).startPos;
+            CPPUNIT_ASSERT_EQUAL(aPositions.top(), nPos);
+            aPositions.pop();
+        }
+        while (nPos > 0);
+    }
+
+    // tdf#113694
+    {
+        const sal_Unicode NON_BMP[] = { 0xD800, 0xDC00 };
+        OUString aTest(NON_BMP, SAL_N_ELEMENTS(NON_BMP));
+
+        sal_Int32 nDone=0;
+        sal_Int32 nPos;
+
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full surrogate pair", static_cast<sal_Int32>(SAL_N_ELEMENTS(NON_BMP)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(NON_BMP), aLocale,
+            i18n::CharacterIteratorMode::SKIPCELL, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full surrogate pair", static_cast<sal_Int32>(0), nPos);
+
+        nPos = m_xBreak->nextCharacters(aTest, 0, aLocale,
+            i18n::CharacterIteratorMode::SKIPCHARACTER, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full surrogate pair", static_cast<sal_Int32>(SAL_N_ELEMENTS(NON_BMP)), nPos);
+        nPos = m_xBreak->previousCharacters(aTest, SAL_N_ELEMENTS(NON_BMP), aLocale,
+            i18n::CharacterIteratorMode::SKIPCHARACTER, 1, nDone);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should skip full surrogate pair", static_cast<sal_Int32>(0), nPos);
+    }
+}
+
+#ifdef TODO
+void TestBreakIterator::testNorthernThai()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "nod";
+    aLocale.Country = "TH";
+
+    const sal_Unicode NORTHERN_THAI1[] = { 0x0E01, 0x0E38, 0x0E4A, 0x0E2B, 0x0E25, 0x0E32, 0x0E1A };
+    OUString aTest(NORTHERN_THAI1, SAL_N_ELEMENTS(NORTHERN_THAI1));
+    i18n::Boundary aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale,
+        i18n::WordType::DICTIONARY_WORD, true);
+    CPPUNIT_ASSERT_MESSAGE("Should skip full word",
+        aBounds.startPos == 0 && aBounds.endPos == aTest.getLength());
+}
+
+// Not sure if any version earlier than 49 did have Khmer word boundary
+// dictionaries, 4.6 does not.
+
+// As of icu 54, word boundary detection for Khmer is still considered
+// insufficient, so icu khmer stuff is disabled
+
+//A test to ensure that our khmer word boundary detection is useful
+//https://bugs.libreoffice.org/show_bug.cgi?id=52020
+void TestBreakIterator::testKhmer()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "km";
+    aLocale.Country = "KH";
+
+    const sal_Unicode KHMER[] = { 0x17B2, 0x17D2, 0x1799, 0x1782, 0x17C1 };
+
+    OUString aTest(KHMER, SAL_N_ELEMENTS(KHMER));
+    i18n::Boundary aBounds = m_xBreak->getWordBoundary(aTest, 0, aLocale,
+        i18n::WordType::DICTIONARY_WORD, true);
+
+    CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 3);
+
+    aBounds = m_xBreak->getWordBoundary(aTest, aBounds.endPos, aLocale,
+        i18n::WordType::DICTIONARY_WORD, true);
+
+    CPPUNIT_ASSERT(aBounds.startPos == 3 && aBounds.endPos == 5);
+}
+#endif
+
+void TestBreakIterator::doTestJapanese(uno::Reference< i18n::XBreakIterator > const &xBreak)
+{
+    lang::Locale aLocale;
+    aLocale.Language = "ja";
+    aLocale.Country = "JP";
+    i18n::Boundary aBounds;
+
+    {
+        const sal_Unicode JAPANESE[] = { 0x30B7, 0x30E3, 0x30C3, 0x30C8, 0x30C0, 0x30A6, 0x30F3 };
+
+        OUString aTest(JAPANESE, SAL_N_ELEMENTS(JAPANESE));
+        aBounds = xBreak->getWordBoundary(aTest, 5, aLocale,
+            i18n::WordType::DICTIONARY_WORD, true);
+
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 7);
+    }
+
+    {
+        const sal_Unicode JAPANESE[] = { 0x9EBB, 0x306E, 0x8449, 0x9EBB, 0x306E, 0x8449 };
+
+        OUString aTest(JAPANESE, SAL_N_ELEMENTS(JAPANESE));
+        aBounds = xBreak->getWordBoundary(aTest, 1, aLocale,
+            i18n::WordType::DICTIONARY_WORD, true);
+
+        CPPUNIT_ASSERT(aBounds.startPos == 0 && aBounds.endPos == 3);
+
+        aBounds = xBreak->getWordBoundary(aTest, 5, aLocale,
+            i18n::WordType::DICTIONARY_WORD, true);
+
+        CPPUNIT_ASSERT(aBounds.startPos == 3 && aBounds.endPos == 6);
+    }
+}
+
+void TestBreakIterator::testJapanese()
+{
+    doTestJapanese(m_xBreak);
+
+    // fdo#78479 - test second / cached instantiation of xdictionary
+    uno::Reference< i18n::XBreakIterator > xTmpBreak(m_xSFactory->createInstance(
+        "com.sun.star.i18n.BreakIterator"), uno::UNO_QUERY_THROW);
+
+    doTestJapanese(xTmpBreak);
+}
+
+void TestBreakIterator::testChinese()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "zh";
+    aLocale.Country = "CN";
+    i18n::Boundary aBounds;
+
+    {
+        const sal_Unicode CHINESE[] = { 0x6A35, 0x6A30, 0x69FE, 0x8919, 0xD867, 0xDEDB  };
+
+        OUString aTest(CHINESE, SAL_N_ELEMENTS(CHINESE));
+        aBounds = m_xBreak->getWordBoundary(aTest, 4, aLocale,
+            i18n::WordType::DICTIONARY_WORD, true);
+        CPPUNIT_ASSERT(aBounds.startPos == 4 && aBounds.endPos == 6);
+    }
+}
+void TestBreakIterator::setUp()
+{
+    BootstrapFixtureBase::setUp();
+    m_xBreak.set(m_xSFactory->createInstance("com.sun.star.i18n.BreakIterator"), uno::UNO_QUERY_THROW);
+}
+
+void TestBreakIterator::tearDown()
+{
+    m_xBreak.clear();
+    BootstrapFixtureBase::tearDown();
+}
+
+CPPUNIT_TEST_SUITE_REGISTRATION(TestBreakIterator);
+
+CPPUNIT_PLUGIN_IMPLEMENT();
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/i18npool/qa/cppunit/test_characterclassification.cxx b/i18npool/qa/cppunit/test_characterclassification.cxx
new file mode 100644
index 000000000..4af398440
--- /dev/null
+++ b/i18npool/qa/cppunit/test_characterclassification.cxx
@@ -0,0 +1,105 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ */
+
+#include <com/sun/star/i18n/XCharacterClassification.hpp>
+#include <unotest/bootstrapfixturebase.hxx>
+
+using namespace ::com::sun::star;
+
+class TestCharacterClassification : public test::BootstrapFixtureBase
+{
+public:
+    virtual void setUp() override;
+    virtual void tearDown() override;
+
+    void testTitleCase();
+    void testStringType();
+
+    CPPUNIT_TEST_SUITE(TestCharacterClassification);
+    CPPUNIT_TEST(testTitleCase);
+    CPPUNIT_TEST(testStringType);
+    CPPUNIT_TEST_SUITE_END();
+private:
+    uno::Reference<i18n::XCharacterClassification> m_xCC;
+};
+
+//A test to ensure that our Title Case functionality is working
+//http://lists.freedesktop.org/archives/libreoffice/2012-June/032767.html
+//https://bz.apache.org/ooo/show_bug.cgi?id=30863
+void TestCharacterClassification::testTitleCase()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "en";
+    aLocale.Country = "US";
+
+    {
+        //basic example
+        OUString sTest("Some text");
+        OUString sTitleCase = m_xCC->toTitle(sTest, 0, sTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should be title", OUString("Some Text"), sTitleCase);
+        OUString sUpperCase = m_xCC->toUpper(sTest, 0, sTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should be upper", OUString("SOME TEXT"), sUpperCase);
+        OUString sLowerCase = m_xCC->toLower(sTest, 0, sTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_EQUAL_MESSAGE("Should be lower ", OUString("some text"), sLowerCase);
+    }
+
+    {
+        //tricky one
+        const sal_Unicode LATINSMALLLETTERDZ[] = { 0x01F3 };
+        OUString aTest(LATINSMALLLETTERDZ, SAL_N_ELEMENTS(LATINSMALLLETTERDZ));
+        OUString sTitleCase = m_xCC->toTitle(aTest, 0, aTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_MESSAGE("Should be title", sTitleCase.getLength() == 1 && sTitleCase[0] == 0x01F2);
+        OUString sUpperCase = m_xCC->toUpper(aTest, 0, aTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_MESSAGE("Should be upper", sUpperCase.getLength() == 1 && sUpperCase[0] == 0x01F1);
+        OUString sLowerCase = m_xCC->toLower(aTest, 0, aTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_MESSAGE("Should be lower ", sLowerCase.getLength() == 1 && sLowerCase[0] == 0x01F3);
+    }
+}
+
+//https://bugs.libreoffice.org/show_bug.cgi?id=69641
+void TestCharacterClassification::testStringType()
+{
+    lang::Locale aLocale;
+    aLocale.Language = "en";
+    aLocale.Country = "US";
+
+    {
+        //simple case
+        OUString sTest("Some text");
+        sal_Int32 nResult = m_xCC->getStringType(sTest, 0, sTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_EQUAL(sal_Int32(230), nResult);
+    }
+
+    {
+        //tricky case
+        const sal_Unicode MATHEMATICAL_ITALIC_SMALL_THETA[] = { 0xD835, 0xDF03 };
+        OUString sTest(MATHEMATICAL_ITALIC_SMALL_THETA, SAL_N_ELEMENTS(MATHEMATICAL_ITALIC_SMALL_THETA));
+        sal_Int32 nResult = m_xCC->getStringType(sTest, 0, sTest.getLength(), aLocale);
+        CPPUNIT_ASSERT_EQUAL(sal_Int32(228), nResult);
+    }
+
+}
+
+void TestCharacterClassification::setUp()
+{
+    BootstrapFixtureBase::setUp();
+    m_xCC.set(m_xSFactory->createInstance("com.sun.star.i18n.CharacterClassification"), uno::UNO_QUERY_THROW);
+}
+
+void TestCharacterClassification::tearDown()
+{
+    BootstrapFixtureBase::tearDown();
+    m_xCC.clear();
+}
+
+CPPUNIT_TEST_SUITE_REGISTRATION(TestCharacterClassification);
+
+CPPUNIT_PLUGIN_IMPLEMENT();
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/i18npool/qa/cppunit/test_defaultnumberingprovider.cxx b/i18npool/qa/cppunit/test_defaultnumberingprovider.cxx
new file mode 100644
index 000000000..dbe55050a
--- /dev/null
+++ b/i18npool/qa/cppunit/test_defaultnumberingprovider.cxx
@@ -0,0 +1,131 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ */
+
+#include <test/bootstrapfixture.hxx>
+
+#include <com/sun/star/style/NumberingType.hpp>
+#include <com/sun/star/text/DefaultNumberingProvider.hpp>
+#include <com/sun/star/text/XNumberingFormatter.hpp>
+
+#include <comphelper/propertyvalue.hxx>
+
+using namespace ::com::sun::star;
+
+/// i18npool defaultnumberingprovider tests.
+class I18npoolDefaultnumberingproviderTest : public test::BootstrapFixture
+{
+};
+
+CPPUNIT_TEST_FIXTURE(I18npoolDefaultnumberingproviderTest, testArabicZero)
+{
+    // 1 -> "01"
+    uno::Reference<text::XNumberingFormatter> xFormatter(
+        text::DefaultNumberingProvider::create(mxComponentContext), uno::UNO_QUERY);
+    uno::Sequence<beans::PropertyValue> aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(1)),
+    };
+    lang::Locale aLocale;
+    OUString aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    // Without the accompanying fix in place, this test would have failed with a
+    // lang.IllegalArgumentException, support for ARABIC_ZERO was missing.
+    CPPUNIT_ASSERT_EQUAL(OUString("01"), aActual);
+
+    // 10 -> "10"
+    aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(10)),
+    };
+    aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    CPPUNIT_ASSERT_EQUAL(OUString("10"), aActual);
+}
+
+CPPUNIT_TEST_FIXTURE(I18npoolDefaultnumberingproviderTest, testArabicZero3)
+{
+    // 10 -> "010"
+    uno::Reference<text::XNumberingFormatter> xFormatter(
+        text::DefaultNumberingProvider::create(mxComponentContext), uno::UNO_QUERY);
+    uno::Sequence<beans::PropertyValue> aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO3)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(10)),
+    };
+    lang::Locale aLocale;
+    OUString aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    // Without the accompanying fix in place, this test would have failed with a
+    // lang.IllegalArgumentException, support for ARABIC_ZERO3 was missing.
+    CPPUNIT_ASSERT_EQUAL(OUString("010"), aActual);
+
+    // 100 -> "100"
+    aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO3)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(100)),
+    };
+    aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    CPPUNIT_ASSERT_EQUAL(OUString("100"), aActual);
+}
+
+CPPUNIT_TEST_FIXTURE(I18npoolDefaultnumberingproviderTest, testArabicZero4)
+{
+    // 100 -> "0100"
+    uno::Reference<text::XNumberingFormatter> xFormatter(
+        text::DefaultNumberingProvider::create(mxComponentContext), uno::UNO_QUERY);
+    uno::Sequence<beans::PropertyValue> aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO4)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(100)),
+    };
+    lang::Locale aLocale;
+    OUString aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    // Without the accompanying fix in place, this test would have failed with a
+    // lang.IllegalArgumentException, support for ARABIC_ZERO4 was missing.
+    CPPUNIT_ASSERT_EQUAL(OUString("0100"), aActual);
+
+    // 1000 -> "1000"
+    aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO4)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(1000)),
+    };
+    aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    CPPUNIT_ASSERT_EQUAL(OUString("1000"), aActual);
+}
+
+CPPUNIT_TEST_FIXTURE(I18npoolDefaultnumberingproviderTest, testArabicZero5)
+{
+    // 1000 -> "01000"
+    uno::Reference<text::XNumberingFormatter> xFormatter(
+        text::DefaultNumberingProvider::create(mxComponentContext), uno::UNO_QUERY);
+    uno::Sequence<beans::PropertyValue> aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO5)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(1000)),
+    };
+    lang::Locale aLocale;
+    OUString aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    // Without the accompanying fix in place, this test would have failed with a
+    // lang.IllegalArgumentException, support for ARABIC_ZERO5 was missing.
+    CPPUNIT_ASSERT_EQUAL(OUString("01000"), aActual);
+
+    // 10000 -> "10000"
+    aProperties = {
+        comphelper::makePropertyValue("NumberingType",
+                                      static_cast<sal_uInt16>(style::NumberingType::ARABIC_ZERO5)),
+        comphelper::makePropertyValue("Value", static_cast<sal_Int32>(10000)),
+    };
+    aActual = xFormatter->makeNumberingString(aProperties, aLocale);
+    CPPUNIT_ASSERT_EQUAL(OUString("10000"), aActual);
+}
+
+CPPUNIT_PLUGIN_IMPLEMENT();
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/i18npool/qa/cppunit/test_ordinalsuffix.cxx b/i18npool/qa/cppunit/test_ordinalsuffix.cxx
new file mode 100644
index 000000000..fb06a41fa
--- /dev/null
+++ b/i18npool/qa/cppunit/test_ordinalsuffix.cxx
@@ -0,0 +1,100 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ */
+#include <algorithm>
+#include <com/sun/star/i18n/XOrdinalSuffix.hpp>
+#include <com/sun/star/lang/Locale.hpp>
+#include <unotest/bootstrapfixturebase.hxx>
+
+using namespace com::sun::star;
+
+class TestOrdinalSuffix : public test::BootstrapFixtureBase
+{
+private:
+    uno::Reference<i18n::XOrdinalSuffix> m_xOrdinal;
+
+public:
+    virtual void setUp() override;
+    virtual void tearDown() override;
+
+    void testFrench();
+    void testEnglish();
+
+    CPPUNIT_TEST_SUITE(TestOrdinalSuffix);
+    CPPUNIT_TEST(testFrench);
+    CPPUNIT_TEST(testEnglish);
+    CPPUNIT_TEST_SUITE_END();
+};
+
+void TestOrdinalSuffix::setUp()
+{
+    BootstrapFixtureBase::setUp();
+    m_xOrdinal.set(m_xSFactory->createInstance("com.sun.star.i18n.OrdinalSuffix"), uno::UNO_QUERY_THROW);
+}
+
+void TestOrdinalSuffix::tearDown()
+{
+    m_xOrdinal.clear();
+    BootstrapFixtureBase::tearDown();
+}
+
+void TestOrdinalSuffix::testFrench()
+{
+    lang::Locale aLocale("fr", "LU", "");
+    uno::Sequence< OUString > aSuffixes;
+    OUString *pStart, *pEnd, *pFind;
+
+    //1er
+    aSuffixes = m_xOrdinal->getOrdinalSuffix(1, aLocale);
+    pStart = aSuffixes.begin();
+    pEnd = aSuffixes.end();
+    pFind = std::find(pStart, pEnd, OUString("er"));
+    CPPUNIT_ASSERT(pFind != pEnd);
+
+    //2e, 3e, etc.
+    aSuffixes = m_xOrdinal->getOrdinalSuffix(2, aLocale);
+    pStart = aSuffixes.begin();
+    pEnd = aSuffixes.end();
+    pFind = std::find(pStart, pEnd, OUString("e"));
+    CPPUNIT_ASSERT(pFind != pEnd);
+}
+
+void TestOrdinalSuffix::testEnglish()
+{
+    lang::Locale aLocale("en", "US", "");
+    uno::Sequence< OUString > aSuffixes;
+    OUString *pStart, *pEnd, *pFind;
+
+    //1st
+    aSuffixes = m_xOrdinal->getOrdinalSuffix(1, aLocale);
+    pStart = aSuffixes.begin();
+    pEnd = aSuffixes.end();
+    pFind = std::find(pStart, pEnd, OUString("st"));
+    CPPUNIT_ASSERT(pFind != pEnd);
+
+    //2nd
+    aSuffixes = m_xOrdinal->getOrdinalSuffix(2, aLocale);
+    pStart = aSuffixes.begin();
+    pEnd = aSuffixes.end();
+    pFind = std::find(pStart, pEnd, OUString("nd"));
+    CPPUNIT_ASSERT(pFind != pEnd);
+
+    //3rd
+    aSuffixes = m_xOrdinal->getOrdinalSuffix(3, aLocale);
+    pStart = aSuffixes.begin();
+    pEnd = aSuffixes.end();
+    pFind = std::find(pStart, pEnd, OUString("rd"));
+    CPPUNIT_ASSERT(pFind != pEnd);
+}
+
+
+CPPUNIT_TEST_SUITE_REGISTRATION( TestOrdinalSuffix );
+
+CPPUNIT_PLUGIN_IMPLEMENT();
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/i18npool/qa/cppunit/test_textsearch.cxx b/i18npool/qa/cppunit/test_textsearch.cxx
new file mode 100644
index 000000000..22ded53cd
--- /dev/null
+++ b/i18npool/qa/cppunit/test_textsearch.cxx
@@ -0,0 +1,402 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ *
+ * This file incorporates work covered by the following license notice:
+ *
+ *   Licensed to the Apache Software Foundation (ASF) under one or more
+ *   contributor license agreements. See the NOTICE file distributed
+ *   with this work for additional information regarding copyright
+ *   ownership. The ASF licenses this file to you under the Apache
+ *   License, Version 2.0 (the "License"); you may not use this file
+ *   except in compliance with the License. You may obtain a copy of
+ *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
+ */
+
+#include <com/sun/star/util/SearchFlags.hpp>
+#include <com/sun/star/util/SearchOptions.hpp>
+#include <com/sun/star/util/SearchAlgorithms2.hpp>
+#include <com/sun/star/util/XTextSearch2.hpp>
+#include <unotest/bootstrapfixturebase.hxx>
+#include <i18nutil/transliteration.hxx>
+
+#include <unicode/regex.h>
+
+using namespace ::com::sun::star;
+typedef U_ICU_NAMESPACE::UnicodeString IcuUniString;
+
+class TestTextSearch : public test::BootstrapFixtureBase
+{
+public:
+    virtual void setUp() override;
+    virtual void tearDown() override;
+
+    void testICU();
+    void testSearches();
+    void testWildcardSearch();
+    void testApostropheSearch();
+
+    CPPUNIT_TEST_SUITE(TestTextSearch);
+    CPPUNIT_TEST(testICU);
+    CPPUNIT_TEST(testSearches);
+    CPPUNIT_TEST(testWildcardSearch);
+    CPPUNIT_TEST(testApostropheSearch);
+    CPPUNIT_TEST_SUITE_END();
+private:
+    uno::Reference<util::XTextSearch> m_xSearch;
+    uno::Reference<util::XTextSearch2> m_xSearch2;
+};
+
+// Sanity check our ICU first ...
+void TestTextSearch::testICU()
+{
+    UErrorCode nErr = U_ZERO_ERROR;
+    sal_uInt32 nSearchFlags = UREGEX_UWORD | UREGEX_CASE_INSENSITIVE;
+
+    OUString aString( "abcdefgh" );
+    OUString aPattern( "e" );
+    IcuUniString aSearchPat( reinterpret_cast<const UChar*>(aPattern.getStr()), aPattern.getLength() );
+
+    std::unique_ptr<icu::RegexMatcher> pRegexMatcher(new icu::RegexMatcher( aSearchPat, nSearchFlags, nErr ));
+
+    IcuUniString aSource( reinterpret_cast<const UChar*>(aString.getStr()), aString.getLength() );
+    pRegexMatcher->reset( aSource );
+
+    CPPUNIT_ASSERT( pRegexMatcher->find( 0, nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+    CPPUNIT_ASSERT_EQUAL( static_cast<int32_t>(4), pRegexMatcher->start( nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+    CPPUNIT_ASSERT_EQUAL( static_cast<int32_t>(5), pRegexMatcher->end( nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+
+    OUString aString2( "acababaabcababadcdaa" );
+    OUString aPattern2( "a" );
+
+    IcuUniString aSearchPat2( reinterpret_cast<const UChar*>(aPattern2.getStr()), aPattern2.getLength() );
+    pRegexMatcher.reset(new icu::RegexMatcher( aSearchPat2, nSearchFlags, nErr ));
+
+    IcuUniString aSource2( reinterpret_cast<const UChar*>(aString2.getStr()), aString2.getLength() );
+    pRegexMatcher->reset( aSource2 );
+
+    CPPUNIT_ASSERT( pRegexMatcher->find( 0, nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+    CPPUNIT_ASSERT_EQUAL( static_cast<int32_t>(0), pRegexMatcher->start( nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+    CPPUNIT_ASSERT_EQUAL( static_cast<int32_t>(1), pRegexMatcher->end( nErr ) );
+    CPPUNIT_ASSERT_EQUAL( U_ZERO_ERROR, nErr );
+}
+
+void TestTextSearch::testSearches()
+{
+    OUString str( "acababaabcababadcdaa" );
+    sal_Int32 startPos = 2, endPos = 20 ;
+    OUString const searchStr( "(ab)*a(c|d)+" );
+    sal_Int32 const fStartRes = 10, fEndRes = 18 ;
+    sal_Int32 const bStartRes = 18, bEndRes = 10 ;
+
+    // set options
+    util::SearchOptions aOptions;
+    aOptions.algorithmType = util::SearchAlgorithms_REGEXP ;
+    aOptions.searchFlag = util::SearchFlags::ALL_IGNORE_CASE;
+    aOptions.searchString = searchStr;
+    m_xSearch->setOptions( aOptions );
+
+    util::SearchResult aRes;
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( fStartRes, aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( fEndRes, aRes.endOffset[0] );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( bStartRes, aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( bEndRes, aRes.endOffset[0] );
+
+    aOptions.transliterateFlags = static_cast<int>(TransliterationFlags::IGNORE_CASE
+                                | TransliterationFlags::IGNORE_WIDTH);
+    aOptions.searchString = "([^ ]*)[ ]*([^ ]*)";
+    m_xSearch->setOptions(aOptions);
+    aRes = m_xSearch->searchForward("11 22 33", 2, 7);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(3), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 2) && (aRes.endOffset[0] == 5));
+    CPPUNIT_ASSERT((aRes.startOffset[1] == 2) && (aRes.endOffset[1] == 2));
+    CPPUNIT_ASSERT((aRes.startOffset[2] == 3) && (aRes.endOffset[2] == 5));
+}
+
+void TestTextSearch::testWildcardSearch()
+{
+    util::SearchOptions2 aOptions;
+    OUString aText;
+    util::SearchResult aRes;
+
+    aOptions.AlgorithmType2 = util::SearchAlgorithms2::WILDCARD ;
+    aOptions.WildcardEscapeCharacter = '~';
+    // aOptions.searchFlag = ::css::util::SearchFlags::WILD_MATCH_SELECTION;
+    // is not set, so substring match is allowed.
+    aOptions.transliterateFlags = sal_Int32(::css::i18n::TransliterationModules::TransliterationModules_IGNORE_CASE);
+    aText = "abAca";
+
+    aOptions.searchString = "a";
+    m_xSearch2->setOptions2( aOptions );
+    // match first "a", [0,1)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 1));
+    // match last "a", (5,4]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 5) && (aRes.endOffset[0] == 4));
+
+    aOptions.searchString = "a?";
+    m_xSearch2->setOptions2( aOptions );
+    // match "ab", [0,2)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 2));
+    // match "ac", (4,2]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 4) && (aRes.endOffset[0] == 2));
+
+    aOptions.searchString = "a*c";
+    m_xSearch2->setOptions2( aOptions );
+    // match "abac", [0,4) XXX NOTE: first match forward
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 4));
+    // match "ac", (4,2] XXX NOTE: first match backward, not greedy
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 4) && (aRes.endOffset[0] == 2));
+
+    aOptions.searchString = "b*a";
+    m_xSearch2->setOptions2( aOptions );
+    // match "ba", [1,3) XXX NOTE: first match forward, not greedy
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 1) && (aRes.endOffset[0] == 3));
+    // match "baca", (5,1] XXX NOTE: first match backward
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 5) && (aRes.endOffset[0] == 1));
+
+    aText = "ab?ca";
+
+    aOptions.searchString = "?~??";
+    m_xSearch2->setOptions2( aOptions );
+    // match "b?c", [1,4)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 1) && (aRes.endOffset[0] == 4));
+    // match "b?c", (4,1]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 4) && (aRes.endOffset[0] == 1));
+
+    aText = "ab*ca";
+
+    aOptions.searchString = "?~*?";
+    m_xSearch2->setOptions2( aOptions );
+    // match "b?c", [1,4)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 1) && (aRes.endOffset[0] == 4));
+    // match "b?c", (4,1]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 4) && (aRes.endOffset[0] == 1));
+
+    aOptions.searchString = "ca?";
+    m_xSearch2->setOptions2( aOptions );
+    // no match
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(0), aRes.subRegExpressions);
+    // no match
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(0), aRes.subRegExpressions);
+
+    aOptions.searchString = "ca*";
+    m_xSearch2->setOptions2( aOptions );
+    // match "ca", [3,5)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 3) && (aRes.endOffset[0] == 5));
+    // match "ca", (5,3]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 5) && (aRes.endOffset[0] == 3));
+
+    aOptions.searchString = "*ca*";
+    m_xSearch2->setOptions2( aOptions );
+    // match "abaca", [0,5)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 5));
+    // match "abaca", (5,0]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 5) && (aRes.endOffset[0] == 0));
+
+    aText = "123123";
+    aOptions.searchString = "*2?";
+    m_xSearch2->setOptions2( aOptions );
+    // match first "123", [0,3)
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 3));
+    // match "123123", (6,0]    Yes this looks odd, but it is as searching "?2*" forward.
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 6) && (aRes.endOffset[0] == 0));
+
+    aOptions.searchFlag |= util::SearchFlags::WILD_MATCH_SELECTION;
+    m_xSearch2->setOptions2( aOptions );
+    // match "123123", [0,6) with greedy '*'
+    aRes = m_xSearch2->searchForward( aText, 0, aText.getLength());
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 0) && (aRes.endOffset[0] == 6));
+    // match "123123", (6,0]
+    aRes = m_xSearch2->searchBackward( aText, aText.getLength(), 0);
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(1), aRes.subRegExpressions);
+    CPPUNIT_ASSERT((aRes.startOffset[0] == 6) && (aRes.endOffset[0] == 0));
+}
+
+void TestTextSearch::testApostropheSearch()
+{
+    // A) find typographic apostrophes also by using ASCII apostrophe in searchString
+    OUString str( u"It\u2019s an apostrophe." );
+    sal_Int32 startPos = 0, endPos = str.getLength();
+
+    // set options
+    util::SearchOptions aOptions;
+    aOptions.algorithmType = util::SearchAlgorithms_ABSOLUTE;
+    aOptions.searchFlag = util::SearchFlags::ALL_IGNORE_CASE;
+    aOptions.searchString = "'";
+    m_xSearch->setOptions( aOptions );
+
+    util::SearchResult aRes;
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    // This was 0.
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.endOffset[0] );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    // This was 0.
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.endOffset[0] );
+
+    // check with transliteration
+    aOptions.transliterateFlags = static_cast<int>(TransliterationFlags::IGNORE_CASE
+                                | TransliterationFlags::IGNORE_WIDTH);
+    m_xSearch->setOptions(aOptions);
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    // This was 0.
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.endOffset[0] );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    // This was 0.
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.endOffset[0] );
+
+    // B) search ASCII apostrophe in a text with ASCII apostrophes
+    str = str.replace(u'\u2019', '\'');
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.endOffset[0] );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.endOffset[0] );
+
+    // C) search typographic apostrophe in a text with ASCII apostrophes (no result)
+    aOptions.searchString = OUString(u"\u2019");
+    m_xSearch->setOptions( aOptions );
+
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(0), aRes.subRegExpressions);
+
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int32>(0), aRes.subRegExpressions);
+
+    // D) search typographic apostrophe in a text with typographic apostrophes
+    str = str.replace('\'', u'\u2019');
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, endPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.endOffset[0] );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, endPos, startPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(3), aRes.startOffset[0] );
+    CPPUNIT_ASSERT_EQUAL( static_cast<sal_Int32>(2), aRes.endOffset[0] );
+
+    // E) search mixed apostrophes in a text with mixed apostrophes:
+    aOptions.searchString = OUString(u"'\u2019");
+    m_xSearch->setOptions( aOptions );
+    str = u"test: \u2019'";
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, str.getLength());
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, str.getLength(), startPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+
+    // F) search mixed apostrophes in a text with ASCII apostrophes:
+    str = u"test: ''";
+
+    // search forward
+    aRes = m_xSearch->searchForward( str, startPos, str.getLength());
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+
+    // search backwards
+    aRes = m_xSearch->searchBackward( str, str.getLength(), startPos );
+    CPPUNIT_ASSERT( aRes.subRegExpressions > 0 );
+}
+
+void TestTextSearch::setUp()
+{
+    BootstrapFixtureBase::setUp();
+    m_xSearch.set(m_xSFactory->createInstance("com.sun.star.util.TextSearch"), uno::UNO_QUERY_THROW);
+    m_xSearch2.set(m_xSFactory->createInstance("com.sun.star.util.TextSearch2"), uno::UNO_QUERY_THROW);
+}
+
+void TestTextSearch::tearDown()
+{
+    m_xSearch.clear();
+    m_xSearch2.clear();
+    BootstrapFixtureBase::tearDown();
+}
+
+CPPUNIT_TEST_SUITE_REGISTRATION(TestTextSearch);
+
+CPPUNIT_PLUGIN_IMPLEMENT();
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */