Use checked string iteration in case conversions

The Unicode table code can only be safely called on valid code-points.
So code that calls it must only pass it valid Unicode data. The string
iterator's Unchecked Unchecked methods only provide this guarantee
when the string being iterated is guaranteed to be valid UTF-16; while
client code should only use QString, QStringView and friends on valid
UTF-16 data, we have no way to be sure they have respected that.

So take the few extra cycles to actually check validity in the course
of iterating strings, when the resulting code-points are to be passed
to the Unicode table look-ups. Add tests that case mapping doesn't
access Unicode tables out of range (it'll trigger the new assertion).
Added some comments to qchar.h that helped me understand surrogates.

Change-Id: Iec2c3106bf1a875bdaa1d622f6cf94d7007e281e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This commit is contained in:
Edward Welbourne 2020-07-31 14:55:57 +02:00
parent ab5e444c8f
commit 78cf89c07d
6 changed files with 40 additions and 7 deletions

View File

@ -1553,6 +1553,7 @@ static auto fullConvertCase(char32_t uc, QUnicodeTables::Case which) noexcept
auto data() const { return chars; }
auto size() const { return sz; }
} result;
Q_ASSERT(uc <= QChar::LastValidCodePoint);
auto pp = result.chars;

View File

@ -504,11 +504,11 @@ public:
}
static constexpr inline bool isHighSurrogate(char32_t ucs4) noexcept
{
return ((ucs4 & 0xfffffc00) == 0xd800);
return (ucs4 & 0xfffffc00) == 0xd800; // 0xd800 + up to 1023 (0x3ff)
}
static constexpr inline bool isLowSurrogate(char32_t ucs4) noexcept
{
return ((ucs4 & 0xfffffc00) == 0xdc00);
return (ucs4 & 0xfffffc00) == 0xdc00; // 0xdc00 + up to 1023 (0x3ff)
}
static constexpr inline bool isSurrogate(char32_t ucs4) noexcept
{
@ -520,6 +520,8 @@ public:
}
static constexpr inline char32_t surrogateToUcs4(char16_t high, char16_t low) noexcept
{
// 0x010000 through 0x10ffff, provided params are actual high, low surrogates.
// 0x010000 + ((high - 0xd800) << 10) + (low - 0xdc00), optimized:
return (char32_t(high)<<10) + low - 0x35fdc00;
}
static constexpr inline char32_t surrogateToUcs4(QChar high, QChar low) noexcept

View File

@ -4755,7 +4755,7 @@ bool QString::isUpper() const
QStringIterator it(*this);
while (it.hasNext()) {
const char32_t uc = it.nextUnchecked();
const char32_t uc = it.next();
if (qGetProp(uc)->cases[QUnicodeTables::UpperCase].diff)
return false;
}
@ -4781,7 +4781,7 @@ bool QString::isLower() const
QStringIterator it(*this);
while (it.hasNext()) {
const char32_t uc = it.nextUnchecked();
const char32_t uc = it.next();
if (qGetProp(uc)->cases[QUnicodeTables::LowerCase].diff)
return false;
}
@ -6143,7 +6143,7 @@ static QString detachAndConvertCase(T &str, QStringIterator it, QUnicodeTables::
QChar *pp = s.begin() + it.index(); // will detach if necessary
do {
const auto folded = fullConvertCase(it.nextUnchecked(), which);
const auto folded = fullConvertCase(it.next(), which);
if (Q_UNLIKELY(folded.size() > 1)) {
if (folded.chars[0] == *pp && folded.size() == 2) {
// special case: only second actually changed (e.g. surrogate pairs),
@ -6183,9 +6183,9 @@ static QString convertCase(T &str, QUnicodeTables::Case which)
QStringIterator it(p, e);
while (it.hasNext()) {
const char32_t uc = it.nextUnchecked();
const char32_t uc = it.next();
if (qGetProp(uc)->cases[which].diff) {
it.recedeUnchecked();
it.recede();
return detachAndConvertCase(str, it, which);
}
}

View File

@ -9592,6 +9592,7 @@ static const Properties uc_properties[] = {
Q_DECL_CONST_FUNCTION static inline const Properties *qGetProp(char32_t ucs4) noexcept
{
Q_ASSERT(ucs4 <= QChar::LastValidCodePoint);
if (ucs4 < 0x11000)
return uc_properties + uc_property_trie[uc_property_trie[ucs4 >> 5] + (ucs4 & 0x1f)];

View File

@ -458,6 +458,8 @@ private slots:
void simplified_data();
void simplified();
void trimmed();
void unicodeTableAccess_data();
void unicodeTableAccess();
void toUpper();
void toLower();
void isLower_isUpper_data();
@ -1949,6 +1951,32 @@ void tst_QString::rightJustified()
QCOMPARE(a, QLatin1String("ABC"));
}
void tst_QString::unicodeTableAccess_data()
{
QTest::addColumn<QString>("invalid");
const auto join = [](char16_t high, char16_t low) {
const QChar pair[2] = { high, low };
return QString(pair, 2);
};
// Least high surrogate for which an invalid successor produces an error:
QTest::newRow("least-high") << join(0xdbf8, 0xfc00);
// Least successor that, after a high surrogate, produces invalid:
QTest::newRow("least-follow") << join(0xdbff, 0xe000);
}
void tst_QString::unicodeTableAccess()
{
// QString processing must not access unicode tables out of bounds.
QFETCH(QString, invalid);
// Exercise methods, to see if any assertions trigger:
const auto upper = invalid.toUpper();
const auto lower = invalid.toLower();
const auto folded = invalid.toCaseFolded();
// Fatuous test, just to use those.
QVERIFY(upper == invalid || lower == invalid || folded == invalid || lower != upper);
}
void tst_QString::toUpper()
{
QCOMPARE( QString().toUpper(), QString() );

View File

@ -2501,6 +2501,7 @@ static QByteArray createPropertyInfo()
out += "Q_DECL_CONST_FUNCTION static inline const Properties *qGetProp(char32_t ucs4) noexcept\n"
"{\n"
" Q_ASSERT(ucs4 <= QChar::LastValidCodePoint);\n"
" if (ucs4 < 0x" + QByteArray::number(BMP_END, 16) + ")\n"
" return uc_properties + uc_property_trie[uc_property_trie[ucs4 >> "
+ QByteArray::number(BMP_SHIFT) + "] + (ucs4 & 0x"