QUrl::fromUserInput: improvements, corrections and make the demo

browser use it.

- Handle windows files names by looking for paths first (and don't check
  that it exists first)
- Handle host names without dots (it was not handled because of
  difficulties with the case host:port)
- Return the parsed url only if the host or the path is not empty
  instead of returning a url that looks like "http:"

Reviewed-by: Thiago Macieira

4 files changed, 29 insertions, 66 deletions

diff --git a/demos/browser/browsermainwindow.cpp b/demos/browser/browsermainwindow.cpp
index fba3ac5110..8e3986ba67 100644
--- a/demos/browser/browsermainwindow.cpp
+++ b/demos/browser/browsermainwindow.cpp
@@ -552,47 +552,6 @@ void BrowserMainWindow::slotViewStatusbar()
     m_autoSaver->changeOccurred();
 }
 
-QUrl BrowserMainWindow::guessUrlFromString(const QString &string)
-{
-    QString urlStr = string.trimmed();
-    QRegExp test(QLatin1String("^[a-zA-Z]+\\:.*"));
-
-    // Check if it looks like a qualified URL. Try parsing it and see.
-    bool hasSchema = test.exactMatch(urlStr);
-    if (hasSchema) {
-        QUrl url = QUrl::fromEncoded(urlStr.toUtf8(), QUrl::TolerantMode);
-        if (url.isValid())
-            return url;
-    }
-
-    // Might be a file.
-    if (QFile::exists(urlStr)) {
-        QFileInfo info(urlStr);
-        return QUrl::fromLocalFile(info.absoluteFilePath());
-    }
-
-    // Might be a shorturl - try to detect the schema.
-    if (!hasSchema) {
-        int dotIndex = urlStr.indexOf(QLatin1Char('.'));
-        if (dotIndex != -1) {
-            QString prefix = urlStr.left(dotIndex).toLower();
-            QByteArray schema = (prefix == QLatin1String("ftp")) ? prefix.toLatin1() : QByteArray("http");
-            QUrl url =
-                QUrl::fromEncoded(schema + "://" + urlStr.toUtf8(), QUrl::TolerantMode);
-            if (url.isValid())
-                return url;
-        }
-    }
-
-    // Fall back to QUrl's own tolerant parser.
-    QUrl url = QUrl::fromEncoded(string.toUtf8(), QUrl::TolerantMode);
-
-    // finally for cases where the user just types in a hostname add http
-    if (url.scheme().isEmpty())
-        url = QUrl::fromEncoded("http://" + string.toUtf8(), QUrl::TolerantMode);
-    return url;
-}
-
 void BrowserMainWindow::loadUrl(const QUrl &url)
 {
     if (!currentTab() || !url.isValid())
@@ -873,7 +832,7 @@ void BrowserMainWindow::slotSwapFocus()
 
 void BrowserMainWindow::loadPage(const QString &page)
 {
-    QUrl url = guessUrlFromString(page);
+    QUrl url = QUrl::fromUserInput(page);
     loadUrl(url);
 }
 
diff --git a/demos/browser/browsermainwindow.h b/demos/browser/browsermainwindow.h
index f5c14ea80f..f0f95b1999 100644
--- a/demos/browser/browsermainwindow.h
+++ b/demos/browser/browsermainwindow.h
@@ -68,7 +68,6 @@ public:
     QSize sizeHint() const;
 
 public:
-    static QUrl guessUrlFromString(const QString &url);
     TabWidget *tabWidget() const;
     WebView *currentTab() const;
     QByteArray saveState(bool withTabs = true) const;
diff --git a/src/corelib/io/qurl.cpp b/src/corelib/io/qurl.cpp
index 86680a5ddd..fd51bcf50b 100644
--- a/src/corelib/io/qurl.cpp
+++ b/src/corelib/io/qurl.cpp
@@ -6210,8 +6210,8 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
     \list
     \o qt.nokia.com becomes http://qt.nokia.com
     \o ftp.qt.nokia.com becomes ftp://ftp.qt.nokia.com
-    \o localhost becomes http://localhost
-    \o /home/user/test.html becomes file:///home/user/test.html (if exists)
+    \o hostname becomes http://hostname
+    \o /home/user/test.html becomes file:///home/user/test.html
     \endlist
 
     \section2 Tips to avoid erroneous character conversion when dealing with
@@ -6228,31 +6228,32 @@ QUrl QUrl::fromUserInput(const QString &userInput)
 {
     QString trimmedString = userInput.trimmed();
 
-    // Check the most common case of a valid url with scheme and host first
+    // Check first for files, since on Windows drive letters can be interpretted as schemes
+    if (QDir::isAbsolutePath(trimmedString))
+        return QUrl::fromLocalFile(trimmedString);
+
     QUrl url = QUrl::fromEncoded(trimmedString.toUtf8(), QUrl::TolerantMode);
-    if (url.isValid() && !url.scheme().isEmpty() && !url.host().isEmpty())
+    QUrl urlPrepended = QUrl::fromEncoded((QLatin1String("http://") + trimmedString).toUtf8(), QUrl::TolerantMode);
+
+    // Check the most common case of a valid url with scheme and host
+    // We check if the port would be valid by adding the scheme to handle the case host:port
+    // where the host would be interpretted as the scheme
+    if (url.isValid()
+        && !url.scheme().isEmpty()
+        && (!url.host().isEmpty() || !url.path().isEmpty())
+        && urlPrepended.port() == -1)
         return url;
 
-    // Absolute files that exists
-    if (QDir::isAbsolutePath(trimmedString) && QFile::exists(trimmedString))
-        return QUrl::fromLocalFile(trimmedString);
-
-    // If the string is missing the scheme or the scheme is not valid prepend a scheme
-    QString scheme = url.scheme();
-    if (scheme.isEmpty() || scheme.contains(QLatin1Char('.')) || scheme == QLatin1String("localhost")) {
-        // Do not do anything for strings such as "foo", only "foo.com"
+    // Else, try the prepended one and adjust the scheme from the host name
+    if (urlPrepended.isValid() && (!urlPrepended.host().isEmpty() || !urlPrepended.path().isEmpty()))
+    {
         int dotIndex = trimmedString.indexOf(QLatin1Char('.'));
-        if (dotIndex != -1 || trimmedString.startsWith(QLatin1String("localhost"))) {
-            const QString hostscheme = trimmedString.left(dotIndex).toLower();
-            QByteArray scheme = (hostscheme == QLatin1String("ftp")) ? "ftp" : "http";
-            trimmedString = QLatin1String(scheme) + QLatin1String("://") + trimmedString;
-        }
-        url = QUrl::fromEncoded(trimmedString.toUtf8(), QUrl::TolerantMode);
+        const QString hostscheme = trimmedString.left(dotIndex).toLower();
+        if (hostscheme == QLatin1String("ftp"))
+            urlPrepended.setScheme(QLatin1String("ftp"));
+        return urlPrepended;
     }
 
-    if (url.isValid())
-        return url;
-
     return QUrl();
 }
 // end of BSD code
diff --git a/tests/auto/qurl/tst_qurl.cpp b/tests/auto/qurl/tst_qurl.cpp
index c8fe4e57b0..03e77aa94f 100644
--- a/tests/auto/qurl/tst_qurl.cpp
+++ b/tests/auto/qurl/tst_qurl.cpp
@@ -3713,7 +3713,7 @@ void tst_QUrl::fromUserInput_data()
     QTest::newRow("add scheme-0") << "example.org" << QUrl("http://example.org");
     QTest::newRow("add scheme-1") << "www.example.org" << QUrl("http://www.example.org");
     QTest::newRow("add scheme-2") << "ftp.example.org" << QUrl("ftp://ftp.example.org");
-    QTest::newRow("add scheme-3") << "webkit" << QUrl("webkit");
+    QTest::newRow("add scheme-3") << "hostname" << QUrl("http://hostname");
 
     // QUrl's tolerant parser should already handle this
     QTest::newRow("not-encoded-0") << "http://example.org/test page.html" << QUrl::fromEncoded("http://example.org/test%20page.html");
@@ -3723,6 +3723,9 @@ void tst_QUrl::fromUserInput_data()
     portUrl.setPort(80);
     QTest::newRow("port-0") << "example.org:80" << portUrl;
     QTest::newRow("port-1") << "http://example.org:80" << portUrl;
+    portUrl.setPath("path");
+    QTest::newRow("port-1") << "example.org:80/path" << portUrl;
+    QTest::newRow("port-1") << "http://example.org:80/path" << portUrl;
 
     // mailto doesn't have a ://, but is valid
     QUrl mailto("ben@example.net");
@@ -3730,10 +3733,11 @@ void tst_QUrl::fromUserInput_data()
     QTest::newRow("mailto") << "mailto:ben@example.net" << mailto;
 
     // misc
-    QTest::newRow("localhost-0") << "localhost" << QUrl("http://localhost");
     QTest::newRow("localhost-1") << "localhost:80" << QUrl("http://localhost:80");
     QTest::newRow("spaces-0") << "  http://example.org/test page.html " << QUrl("http://example.org/test%20page.html");
     QTest::newRow("trash-0") << "example.org/test?someData=42%&someOtherData=abcde#anchor" << QUrl::fromEncoded("http://example.org/test?someData=42%25&someOtherData=abcde#anchor");
+    QTest::newRow("other-scheme-0") << "spotify:track:0hO542doVbfGDAGQULMORT" << QUrl("spotify:track:0hO542doVbfGDAGQULMORT");
+    QTest::newRow("other-scheme-1") << "weirdscheme:80:otherstuff" << QUrl("weirdscheme:80:otherstuff");
 
     // FYI: The scheme in the resulting url user
     QUrl authUrl("user:pass@domain.com");