| 1 |
#include <QtCore/QScopedPointer> |
| 2 |
#include <QtTest/QTest> |
| 3 |
|
| 4 |
#include <poppler-private.h> |
| 5 |
|
| 6 |
#include <cstring> |
| 7 |
#include <cstdint> // for uint16_t |
| 8 |
|
| 9 |
#include "GlobalParams.h" |
| 10 |
#include "UnicodeTypeTable.h" |
| 11 |
#include "UTF.h" |
| 12 |
|
| 13 |
class TestUTFConversion : public QObject |
| 14 |
{ |
| 15 |
Q_OBJECT |
| 16 |
public: |
| 17 |
explicit TestUTFConversion(QObject *parent = nullptr) : QObject(parent) { } |
| 18 |
private Q_SLOTS: |
| 19 |
void testUTF_data(); |
| 20 |
void testUTF(); |
| 21 |
void testUnicodeToAscii7(); |
| 22 |
void testUnicodeLittleEndian(); |
| 23 |
}; |
| 24 |
|
| 25 |
static bool compare(const Unicode *a, const char *b, int len) |
| 26 |
{ |
| 27 |
for (int i = 0; i < len; i++) { |
| 28 |
if (a[i] != (Unicode)b[i]) { |
| 29 |
return false; |
| 30 |
} |
| 31 |
} |
| 32 |
|
| 33 |
return true; |
| 34 |
} |
| 35 |
|
| 36 |
static bool compare(const Unicode *a, const uint16_t *b, int len) |
| 37 |
{ |
| 38 |
for (int i = 0; i < len; i++) { |
| 39 |
if (a[i] != b[i]) { |
| 40 |
return false; |
| 41 |
} |
| 42 |
} |
| 43 |
|
| 44 |
return true; |
| 45 |
} |
| 46 |
|
| 47 |
void TestUTFConversion::testUTF_data() |
| 48 |
{ |
| 49 |
QTest::addColumn<QString>("s"); |
| 50 |
|
| 51 |
QTest::newRow("<empty>") << QString(QLatin1String("")); |
| 52 |
QTest::newRow("a") << QStringLiteral("a"); |
| 53 |
QTest::newRow("abc") << QStringLiteral("abc"); |
| 54 |
QTest::newRow("Latin") << QStringLiteral("Vitrum edere possum; mihi non nocet"); |
| 55 |
QTest::newRow("Greek") << QStringLiteral("Μπορώ να φάω σπασμένα γυαλιά χωρίς να πάθω τίποτα"); |
| 56 |
QTest::newRow("Icelandic") << QStringLiteral("Ég get etið gler án þess að meiða mig"); |
| 57 |
QTest::newRow("Russian") << QStringLiteral("Я могу есть стекло, оно мне не вредит."); |
| 58 |
QTest::newRow("Sanskrit") << QStringLiteral("काचं शक्नोम्यत्तुम् । नोपहिनस्ति माम् ॥"); |
| 59 |
QTest::newRow("Arabic") << QStringLiteral("أنا قادر على أكل الزجاج و هذا لا يؤلمني"); |
| 60 |
QTest::newRow("Chinese") << QStringLiteral("我能吞下玻璃而不伤身体。"); |
| 61 |
QTest::newRow("Thai") << QStringLiteral("ฉันกินกระจกได้ แต่มันไม่ทำให้ฉันเจ็บ"); |
| 62 |
QTest::newRow("non BMP") << QStringLiteral("𝓹𝓸𝓹𝓹𝓵𝓮𝓻"); |
| 63 |
} |
| 64 |
|
| 65 |
void TestUTFConversion::testUTF() |
| 66 |
{ |
| 67 |
std::string utf8String; |
| 68 |
int len; |
| 69 |
|
| 70 |
QFETCH(QString, s); |
| 71 |
const std::string str = s.toStdString(); |
| 72 |
|
| 73 |
// UTF-8 to UTF-16 |
| 74 |
|
| 75 |
len = utf8CountUtf16CodeUnits(str); |
| 76 |
QCOMPARE(len, s.size()); // QString size() returns number of code units, not code points |
| 77 |
|
| 78 |
std::u16string utf16String = utf8ToUtf16(str); |
| 79 |
QCOMPARE(utf16String, s.toStdU16String()); |
| 80 |
QCOMPARE(len, s.size()); |
| 81 |
|
| 82 |
std::string gsUtf16_a(utf8ToUtf16WithBom(str)); |
| 83 |
std::unique_ptr<GooString> gsUtf16_b(Poppler::QStringToUnicodeGooString(s)); |
| 84 |
QCOMPARE(gsUtf16_b->cmp(gsUtf16_a), 0); |
| 85 |
|
| 86 |
// UTF-16 to UTF-8 |
| 87 |
|
| 88 |
len = utf16CountUtf8Bytes(s.utf16()); |
| 89 |
QCOMPARE(len, str.size()); |
| 90 |
|
| 91 |
utf8String = utf16ToUtf8(s.utf16(), INT_MAX); |
| 92 |
QCOMPARE(utf8String, str); |
| 93 |
QCOMPARE(len, str.size()); |
| 94 |
|
| 95 |
utf8String = utf16ToUtf8(s.utf16()); |
| 96 |
QCOMPARE(utf8String, str); |
| 97 |
} |
| 98 |
|
| 99 |
void TestUTFConversion::testUnicodeToAscii7() |
| 100 |
{ |
| 101 |
globalParams = std::make_unique<GlobalParams>(); |
| 102 |
|
| 103 |
// Test string is one 'Registered' and twenty 'Copyright' chars |
| 104 |
// so it's long enough to reproduce the bug given that glibc |
| 105 |
// malloc() always returns 8-byte aligned memory addresses. |
| 106 |
std::unique_ptr<GooString> goo = Poppler::QStringToUnicodeGooString(QString::fromUtf8("®©©©©©©©©©©©©©©©©©©©©")); // clazy:exclude=qstring-allocations |
| 107 |
|
| 108 |
const std::vector<Unicode> in = TextStringToUCS4(goo->toStr()); |
| 109 |
|
| 110 |
int in_norm_len; |
| 111 |
int *in_norm_idx; |
| 112 |
Unicode *in_norm = unicodeNormalizeNFKC(in.data(), in.size(), &in_norm_len, &in_norm_idx, true); |
| 113 |
|
| 114 |
Unicode *out; |
| 115 |
int out_len; |
| 116 |
int *out_ascii_idx; |
| 117 |
|
| 118 |
unicodeToAscii7(std::span(in_norm, in_norm_len), &out, &out_len, in_norm_idx, &out_ascii_idx); |
| 119 |
|
| 120 |
free(in_norm); |
| 121 |
free(in_norm_idx); |
| 122 |
|
| 123 |
// ascii7 conversion: ® -> (R) © -> (c) |
| 124 |
const char *expected_ascii = (char *)"(R)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)(c)"; |
| 125 |
|
| 126 |
QCOMPARE(out_len, (int)strlen(expected_ascii)); |
| 127 |
QVERIFY(compare(out, expected_ascii, out_len)); |
| 128 |
|
| 129 |
free(out); |
| 130 |
free(out_ascii_idx); |
| 131 |
} |
| 132 |
|
| 133 |
void TestUTFConversion::testUnicodeLittleEndian() |
| 134 |
{ |
| 135 |
uint16_t UTF16LE_hi[5] { 0xFFFE, 0x4800, 0x4900, 0x2100, 0x1126 }; // UTF16-LE "HI!☑" |
| 136 |
std::string GooUTF16LE(reinterpret_cast<const char *>(UTF16LE_hi), sizeof(UTF16LE_hi)); |
| 137 |
|
| 138 |
uint16_t UTF16BE_hi[5] { 0xFEFF, 0x0048, 0x0049, 0x0021, 0x2611 }; // UTF16-BE "HI!☑" |
| 139 |
std::string GooUTF16BE(reinterpret_cast<const char *>(UTF16BE_hi), sizeof(UTF16BE_hi)); |
| 140 |
|
| 141 |
// Let's assert both GooString's are different |
| 142 |
QVERIFY(GooUTF16LE != GooUTF16BE); |
| 143 |
|
| 144 |
const std::vector<Unicode> UCS4fromLE = TextStringToUCS4(GooUTF16LE); |
| 145 |
const std::vector<Unicode> UCS4fromBE = TextStringToUCS4(GooUTF16BE); |
| 146 |
|
| 147 |
// len is 4 because TextStringToUCS4() removes the two leading Byte Order Mark (BOM) code points |
| 148 |
QCOMPARE(UCS4fromLE.size(), UCS4fromBE.size()); |
| 149 |
QCOMPARE(UCS4fromLE.size(), 4); |
| 150 |
|
| 151 |
// Check that now after conversion, UCS4fromLE and UCS4fromBE are now the same |
| 152 |
for (size_t i = 0; i < UCS4fromLE.size(); i++) { |
| 153 |
QCOMPARE(UCS4fromLE[i], UCS4fromBE[i]); |
| 154 |
} |
| 155 |
|
| 156 |
const QString expected = QString::fromUtf8("HI!☑"); // clazy:exclude=qstring-allocations |
| 157 |
|
| 158 |
// Do some final verifications, checking the strings to be "HI!" |
| 159 |
QVERIFY(UCS4fromLE == UCS4fromBE); |
| 160 |
QVERIFY(compare(UCS4fromLE.data(), expected.utf16(), UCS4fromLE.size())); |
| 161 |
QVERIFY(compare(UCS4fromBE.data(), expected.utf16(), UCS4fromLE.size())); |
| 162 |
} |
| 163 |
|
| 164 |
QTEST_GUILESS_MAIN(TestUTFConversion) |
| 165 |
#include "check_utf_conversion.moc" |