Fix IPvFuture use in QUrl

We have no idea what it might contain, but test it anyway to make sure
it works. Turns out there were a few bugs the unit tests have now
caught.

Change-Id: I0a6c868365feec31c2360b3c341c8ca6944f4352
Reviewed-by: David Faure (KDE) <faure@kde.org>

2 files changed, 70 insertions, 5 deletions

diff --git a/src/corelib/io/qurl.cpp b/src/corelib/io/qurl.cpp
index 17447d9490..be4f177515 100644
--- a/src/corelib/io/qurl.cpp
+++ b/src/corelib/io/qurl.cpp
@@ -1087,7 +1087,7 @@ inline void QUrlPrivate::appendHost(QString &appendTo, QUrl::FormattingOptions o
 
 // the whole IPvFuture is passed and parsed here, including brackets;
 // returns null if the parsing was successful, or the QChar of the first failure
-static const QChar *parseIpFuture(QString &host, const QChar *begin, const QChar *end)
+static const QChar *parseIpFuture(QString &host, const QChar *begin, const QChar *end, QUrl::ParsingMode mode)
 {
     //    IPvFuture     = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
     static const char acceptable[] =
@@ -1098,17 +1098,18 @@ static const QChar *parseIpFuture(QString &host, const QChar *begin, const QChar
     // the brackets and the "v" have been checked
     if (begin[3].unicode() != '.')
         return &begin[3];
-    if ((begin[2].unicode() >= 'A' && begin[2].unicode() >= 'F') ||
+    if ((begin[2].unicode() >= 'A' && begin[2].unicode() <= 'F') ||
             (begin[2].unicode() >= 'a' && begin[2].unicode() <= 'f') ||
             (begin[2].unicode() >= '0' && begin[2].unicode() <= '9')) {
         // this is so unlikely that we'll just go down the slow path
         // decode the whole string, skipping the "[vH." and "]" which we already know to be there
         host += QString::fromRawData(begin, 4);
+
         begin += 4;
         --end;
 
         QString decoded;
-        if (qt_urlRecode(decoded, begin, end, QUrl::FullyEncoded, 0)) {
+        if (mode == QUrl::TolerantMode && qt_urlRecode(decoded, begin, end, QUrl::FullyDecoded, 0)) {
             begin = decoded.constBegin();
             end = decoded.constEnd();
         }
@@ -1180,7 +1181,7 @@ inline bool QUrlPrivate::setHost(const QString &value, int from, int iend, QUrl:
         }
 
         if (len > 5 && begin[1].unicode() == 'v') {
-            const QChar *c = parseIpFuture(host, begin, end);
+            const QChar *c = parseIpFuture(host, begin, end, mode);
             if (c)
                 setError(InvalidIPvFutureError, value, c - value.constData());
             return !c;
@@ -3653,7 +3654,7 @@ static QString errorMessage(QUrlPrivate::ErrorCode errorCode, const QString &err
     case QUrlPrivate::InvalidIPv6AddressError:
         return QStringLiteral("Invalid IPv6 address");
     case QUrlPrivate::InvalidIPvFutureError:
-        return QStringLiteral("Invalid IPvFuture address");
+        return QStringLiteral("Invalid IPvFuture address (character '%1' not permitted)").arg(c);
     case QUrlPrivate::HostMissingEndBracket:
         return QStringLiteral("Expected ']' to match '[' in hostname");
 
diff --git a/tests/auto/corelib/io/qurl/tst_qurl.cpp b/tests/auto/corelib/io/qurl/tst_qurl.cpp
index cda187b037..ad4c0c1224 100644
--- a/tests/auto/corelib/io/qurl/tst_qurl.cpp
+++ b/tests/auto/corelib/io/qurl/tst_qurl.cpp
@@ -111,6 +111,8 @@ private slots:
     void percentEncoding();
     void swap();
     void symmetry();
+    void ipvfuture_data();
+    void ipvfuture();
     void ipv6_data();
     void ipv6();
     void ipv6_2_data();
@@ -1616,6 +1618,66 @@ void tst_QUrl::symmetry()
     }
 }
 
+void tst_QUrl::ipvfuture_data()
+{
+    QTest::addColumn<QString>("input");
+    QTest::addColumn<bool>("isValid");
+    QTest::addColumn<QString>("output");
+
+    // No one uses IPvFuture yet, so we have no clue what it might contain
+    // We're just testing that it can hold what the RFC says it should hold:
+    //    IPvFuture     = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
+    QTest::newRow("missing-version-dot") << "x://[v]" << false;
+    QTest::newRow("missing-version") << "x://[v.]" << false;
+    QTest::newRow("missing-version-2") << "x://[v.1234]" << false;
+    QTest::newRow("missing-dot") << "x://[v7]" << false;
+    QTest::newRow("missing-dot-2") << "x://[v71234]" << false;
+    QTest::newRow("missing-data") << "x://[v7.]" << false;
+    QTest::newRow("non-hex-version") << "x://[vz.1234]" << false;
+
+    QTest::newRow("digit-ver") << "x://[v7.1]" << true << "x://[v7.1]";
+    QTest::newRow("lowercase-hex-ver") << "x://[va.1]" << true << "x://[va.1]";
+    QTest::newRow("lowercase-hex-ver") << "x://[vA.1]" << true << "x://[vA.1]";
+
+    QTest::newRow("data-digits") << "x://[v7.1234]" << true << "x://[v7.1234]";
+    QTest::newRow("data-unreserved") << "x://[v7.hello~-WORLD_.com]" << true << "x://[v7.hello~-WORLD_.com]";
+    QTest::newRow("data-sub-delims-colon") << "x://[v7.!$&'()*+,;=:]" << true << "x://[v7.!$&'()*+,;=:]";
+
+    // we're using the tolerant parser
+    QTest::newRow("data-encoded-digits") << "x://[v7.%31%32%33%34]" << true << "x://[v7.1234]";
+    QTest::newRow("data-encoded-unreserved") << "x://[v7.%7E%2D%54%5f%2E]" << true << "x://[v7.~-T_.]";
+    QTest::newRow("data-encoded-sub-delims-colon") << "x://[v7.%21%24%26%27%28%29%2A%2B%2C%3B%3D%3A]" << true << "x://[v7.!$&'()*+,;=:]";
+
+    // should we test "[%76%37%2ex]" -> "[v7.x]" ?
+
+    QTest::newRow("data-invalid-space") << "x://[v7.%20]" << false;
+    QTest::newRow("data-invalid-control") << "x://[v7.\x7f]" << false;
+    QTest::newRow("data-invalid-other-1") << "x://[v7.{1234}]" << false;
+    QTest::newRow("data-invalid-other-2") << "x://[v7.<hello>]" << false;
+    QTest::newRow("data-invalid-unicode") << "x://[v7.æøå]" << false;
+    QTest::newRow("data-invalid-percent") << "x://[v7.%]" << false;
+    QTest::newRow("data-invalid-percent-percent") << "x://[v7.%25]" << false;
+}
+
+void tst_QUrl::ipvfuture()
+{
+    QFETCH(QString, input);
+    QFETCH(bool, isValid);
+
+    QUrl url(input);
+    if (isValid) {
+        QVERIFY2(url.isValid(), qPrintable(url.errorString()));
+
+        QFETCH(QString, output);
+        QCOMPARE(url.toString(), output);
+
+        QUrl url2(output);
+        QCOMPARE(url2, url);
+    } else {
+        QVERIFY(!url.isValid());
+    }
+}
+
 
 void tst_QUrl::ipv6_data()
 {
@@ -1953,7 +2015,9 @@ void tst_QUrl::strictParser_data()
     QTest::newRow("invalid-ipvfuture-1") << "http://[v7]" << "Invalid IPvFuture address";
     QTest::newRow("invalid-ipvfuture-2") << "http://[v7.]" << "Invalid IPvFuture address";
     QTest::newRow("invalid-ipvfuture-3") << "http://[v789]" << "Invalid IPvFuture address";
+    QTest::newRow("invalid-ipvfuture-char1") << "http://[v7.^]" << "Invalid IPvFuture address";
     QTest::newRow("invalid-encoded-ipv6") << "x://[%3a%3a%31]" << "Invalid IPv6 address";
+    QTest::newRow("invalid-encoded-ipvfuture") << "x://[v7.%7E%2D%54%5f%2E]" << "Invalid IPvFuture address";
     QTest::newRow("unbalanced-brackets") << "http://[ff02::1" << "Expected ']' to match '[' in hostname";
 
     // invalid hostnames happen in TolerantMode too