QUrl: Implement UTS #46

UTS #46 (https://unicode.org/reports/tr46/) is a successor to
IDNA 2003/2008 standards from Unicode.

The current implementation uses nontransitional processing by default.
An optional argument is added to QUrl::toAce() and QUrl::fromAce() to
allow using transitional processing and to ignore the IDN whitelist.

[ChangeLog][QtCore][QUrl] ACE processing is now performed according
to the UTS #46 standard based on IDNA 2008 instead of IDNA 2003.

Task-number: QTBUG-85371
Change-Id: I46b2e86792bc9699cb6961bae8e283fbff72f874
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>

src/corelib/io/qurl.cpp

@@ -389,6 +389,25 @@
     \sa fromUserInput()
 */
 
+/*!
+    \enum  QUrl::AceProcessingOption
+    \since 6.3
+
+    The ACE processing options control the way URLs are transformed to and from
+    ASCII-Compatible Encoding.
+
+    \value IgnoreIDNWhitelist         Ignore the IDN whitelist when converting URLs
+                                      to Unicode.
+    \value AceTransitionalProcessing  Use transitional processing described in UTS #46.
+                                      This allows better compatibility with IDNA 2003
+                                      specification.
+
+    The default is to use nontransitional processing and to allow non-ASCII
+    characters only inside URLs whose top-level domains are listed in the IDN whitelist.
+
+    \sa toAce(), fromAce(), idnWhitelist()
+*/
+
 /*!
     \fn QUrl::QUrl(QUrl &&other)
 
@@ -1177,7 +1196,7 @@ inline void QUrlPrivate::appendHost(QString &appendTo, QUrl::FormattingOptions o
         // this is either an IPv4Address or a reg-name
         // if it is a reg-name, it is already stored in Unicode form
         if (options & QUrl::EncodeUnicode && !(options & 0x4000000))
-            appendTo += qt_ACE_do(host, ToAceOnly, AllowLeadingDot);
+            appendTo += qt_ACE_do(host, ToAceOnly, AllowLeadingDot, {});
         else
             appendTo += host;
     }
@@ -1339,7 +1358,7 @@ inline bool QUrlPrivate::setHost(const QString &value, int from, int iend, QUrl:
     //  Unicode encoding (some non-ASCII characters case-fold to digits
     //                    when nameprepping is done)
     //
-    // The qt_ACE_do function below applies nameprepping and the STD3 check.
+    // The qt_ACE_do function below does IDNA normalization and the STD3 check.
     // That means a Unicode string may become an IPv4 address, but it cannot
     // produce a '[' or a '%'.
 
@@ -1358,7 +1377,7 @@ inline bool QUrlPrivate::setHost(const QString &value, int from, int iend, QUrl:
         return setHost(s, 0, s.length(), QUrl::StrictMode);
     }
 
-    s = qt_ACE_do(QStringView(begin, len), NormalizeAce, ForbidLeadingDot);
+    s = qt_ACE_do(value.mid(from, iend - from), NormalizeAce, ForbidLeadingDot, {});
     if (s.isEmpty()) {
         setError(InvalidRegNameError, value);
         return false;
@@ -3013,50 +3032,72 @@ QByteArray QUrl::toPercentEncoding(const QString &input, const QByteArray &exclu
 }
 
 /*!
-    \since 4.2
+    \since 6.3
 
     Returns the Unicode form of the given domain name
     \a domain, which is encoded in the ASCII Compatible Encoding (ACE).
+    The output can be customized by passing flags with \a options.
     The result of this function is considered equivalent to \a domain.
 
     If the value in \a domain cannot be encoded, it will be converted
     to QString and returned.
 
-    The ASCII Compatible Encoding (ACE) is defined by RFC 3490, RFC 3491
+    The ASCII-Compatible Encoding (ACE) is defined by RFC 3490, RFC 3491
-    and RFC 3492. It is part of the Internationalizing Domain Names in
+    and RFC 3492 and updated by the Unicode Technical Standard #46. It is part
-    Applications (IDNA) specification, which allows for domain names
+    of the Internationalizing Domain Names in Applications (IDNA) specification,
-    (like \c "example.com") to be written using international
+    which allows for domain names (like \c "example.com") to be written using
-    characters.
+    non-US-ASCII characters.
+*/
+QString QUrl::fromAce(const QByteArray &domain, QUrl::AceProcessingOptions options)
+{
+    return qt_ACE_do(QString::fromLatin1(domain), NormalizeAce,
+                     ForbidLeadingDot /*FIXME: make configurable*/, options);
+}
+
+#if QT_VERSION < QT_VERSION_CHECK(7, 0, 0)
+/*!
+    \since 4.2
+    \overload
 */
 QString QUrl::fromAce(const QByteArray &domain)
 {
-    QVarLengthArray<char16_t> buffer;
+    return fromAce(domain, {});
-    buffer.resize(domain.size());
-    qt_from_latin1(buffer.data(), domain.data(), domain.size());
-    return qt_ACE_do(QStringView{buffer.data(), buffer.size()},
-                     NormalizeAce, ForbidLeadingDot /*FIXME: make configurable*/);
 }
+#endif
 
 /*!
-    \since 4.2
+    \since 6.3
 
     Returns the ASCII Compatible Encoding of the given domain name \a domain.
+    The output can be customized by passing flags with \a options.
     The result of this function is considered equivalent to \a domain.
 
     The ASCII-Compatible Encoding (ACE) is defined by RFC 3490, RFC 3491
-    and RFC 3492. It is part of the Internationalizing Domain Names in
+    and RFC 3492 and updated by the Unicode Technical Standard #46. It is part
-    Applications (IDNA) specification, which allows for domain names
+    of the Internationalizing Domain Names in Applications (IDNA) specification,
-    (like \c "example.com") to be written using international
+    which allows for domain names (like \c "example.com") to be written using
-    characters.
+    non-US-ASCII characters.
 
     This function returns an empty QByteArray if \a domain is not a valid
     hostname. Note, in particular, that IPv6 literals are not valid domain
     names.
 */
+QByteArray QUrl::toAce(const QString &domain, AceProcessingOptions options)
+{
+    return qt_ACE_do(domain, ToAceOnly, ForbidLeadingDot /*FIXME: make configurable*/, options)
+            .toLatin1();
+}
+
+#if QT_VERSION < QT_VERSION_CHECK(7, 0, 0)
+/*!
+    \since 4.2
+    \overload
+*/
 QByteArray QUrl::toAce(const QString &domain)
 {
-    return qt_ACE_do(domain, ToAceOnly, ForbidLeadingDot /*FIXME: make configurable*/).toLatin1();
+    return toAce(domain, {});
 }
+#endif
 
 /*!
     \internal

src/corelib/io/qurl.h

@@ -281,8 +281,22 @@ public:
     NSURL *toNSURL() const Q_DECL_NS_RETURNS_AUTORELEASED;
 #endif
 
+    enum AceProcessingOption : unsigned int {
+        IgnoreIDNWhitelist = 0x1,
+        AceTransitionalProcessing = 0x2,
+    };
+    Q_DECLARE_FLAGS(AceProcessingOptions, AceProcessingOption)
+
+#if QT_VERSION >= QT_VERSION_CHECK(7, 0, 0)
+    static QString fromAce(const QByteArray &, AceProcessingOptions options = {});
+    static QByteArray toAce(const QString &, AceProcessingOptions options = {});
+#else
     static QString fromAce(const QByteArray &);
     static QByteArray toAce(const QString &);
+    static QString fromAce(const QByteArray &, AceProcessingOptions options);
+    static QByteArray toAce(const QString &, AceProcessingOptions options);
+#endif
+
     static QStringList idnWhitelist();
     static QStringList toStringList(const QList<QUrl> &uris, FormattingOptions options = FormattingOptions(PrettyDecoded));
     static QList<QUrl> fromStringList(const QStringList &uris, ParsingMode mode = TolerantMode);
@@ -302,6 +316,7 @@ public:
 Q_DECLARE_SHARED(QUrl)
 Q_DECLARE_OPERATORS_FOR_FLAGS(QUrl::ComponentFormattingOptions)
 //Q_DECLARE_OPERATORS_FOR_FLAGS(QUrl::FormattingOptions)
+Q_DECLARE_OPERATORS_FOR_FLAGS(QUrl::AceProcessingOptions)
 
 #ifndef Q_QDOC
 constexpr inline QUrl::FormattingOptions operator|(QUrl::UrlFormattingOption f1, QUrl::UrlFormattingOption f2)

src/corelib/io/qurl_p.h

@@ -65,7 +65,8 @@ extern Q_AUTOTEST_EXPORT qsizetype qt_urlRecode(QString &appendTo, QStringView u
 // in qurlidna.cpp
 enum AceLeadingDot { AllowLeadingDot, ForbidLeadingDot };
 enum AceOperation { ToAceOnly, NormalizeAce };
-extern QString qt_ACE_do(QStringView domain, AceOperation op, AceLeadingDot dot);
+extern QString qt_ACE_do(const QString &domain, AceOperation op, AceLeadingDot dot,
+                         QUrl::AceProcessingOptions options);
 extern Q_AUTOTEST_EXPORT void qt_punycodeEncoder(QStringView in, QString *output);
 extern Q_AUTOTEST_EXPORT QString qt_punycodeDecoder(const QString &pc);