QTextBrowser: assume Markdown is UTF-8

That's how CommonMark specifies it.  The HTML codec-guessing algorithm
was making it fall back to Latin1 in practice, which was screwing up
any Unicode characters found in the markdown source.

Change-Id: I4021adc4a68591ecfd56ef24971af53ce3e9c96d
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>

3 files changed, 37 insertions, 4 deletions

diff --git a/src/widgets/widgets/qtextbrowser.cpp b/src/widgets/widgets/qtextbrowser.cpp
index 2f992b1cff..7a77f86de2 100644
--- a/src/widgets/widgets/qtextbrowser.cpp
+++ b/src/widgets/widgets/qtextbrowser.cpp
@@ -312,13 +312,17 @@ void QTextBrowserPrivate::setSource(const QUrl &url, QTextDocument::ResourceType
         if (data.type() == QVariant::String) {
             txt = data.toString();
         } else if (data.type() == QVariant::ByteArray) {
+            if (type == QTextDocument::HtmlResource) {
 #if QT_CONFIG(textcodec)
-            QByteArray ba = data.toByteArray();
-            QTextCodec *codec = Qt::codecForHtml(ba);
-            txt = codec->toUnicode(ba);
+                QByteArray ba = data.toByteArray();
+                QTextCodec *codec = Qt::codecForHtml(ba);
+                txt = codec->toUnicode(ba);
 #else
-            txt = data.toString();
+                txt = data.toString();
 #endif
+            } else {
+                txt = QString::fromUtf8(data.toByteArray());
+            }
         }
         if (Q_UNLIKELY(txt.isEmpty()))
             qWarning("QTextBrowser: No document for %s", url.toString().toLatin1().constData());
diff --git a/tests/auto/widgets/widgets/qtextbrowser/quotesAndFractions.md b/tests/auto/widgets/widgets/qtextbrowser/quotesAndFractions.md
new file mode 100644
index 0000000000..6bad9cb95b
--- /dev/null
+++ b/tests/auto/widgets/widgets/qtextbrowser/quotesAndFractions.md
@@ -0,0 +1 @@
+you’ll hope to see ❝quotes❞ ﹠1½ ⅔ ¼ ⅗ ⅚ ⅝ some “vulgar” fractions (pardon my «French»)
diff --git a/tests/auto/widgets/widgets/qtextbrowser/tst_qtextbrowser.cpp b/tests/auto/widgets/widgets/qtextbrowser/tst_qtextbrowser.cpp
index 083e297941..27bf0ce7be 100644
--- a/tests/auto/widgets/widgets/qtextbrowser/tst_qtextbrowser.cpp
+++ b/tests/auto/widgets/widgets/qtextbrowser/tst_qtextbrowser.cpp
@@ -94,6 +94,8 @@ private slots:
     void urlEncoding();
     void sourceType_data();
     void sourceType();
+    void unicode_data();
+    void unicode();
 
 private:
     TestBrowser *browser;
@@ -721,5 +723,31 @@ void tst_QTextBrowser::sourceType()
     QCOMPARE(maxHeadingLevel, expectedMaxHeadingLevel);
 }
 
+void tst_QTextBrowser::unicode_data()
+{
+    QTest::addColumn<QString>("sourceFile");
+    QTest::addColumn<QTextDocument::ResourceType>("sourceType");
+    QTest::addColumn<QString>("expectedText");
+
+#if QT_CONFIG(textmarkdownreader)
+    QTest::newRow("markdown with quotes and fractions") << "quotesAndFractions.md" << QTextDocument::MarkdownResource <<
+        "you\u2019ll hope to see \u275Dquotes\u275E \uFE601\u00BD \u2154 \u00BC \u2157 \u215A \u215D some \u201Cvulgar\u201D fractions (pardon my \u00ABFrench\u00BB)";
+#endif
+}
+
+void tst_QTextBrowser::unicode()
+{
+    QFETCH(QString, sourceFile);
+    QFETCH(QTextDocument::ResourceType, sourceType);
+    QFETCH(QString, expectedText);
+    browser->setSource(QUrl::fromLocalFile(QFINDTESTDATA(sourceFile)), sourceType);
+    QTextFrame::iterator iterator = browser->document()->rootFrame()->begin();
+    while (!iterator.atEnd()) {
+        QString blockText = iterator++.currentBlock().text();
+        if (!blockText.isEmpty())
+            QCOMPARE(blockText, expectedText);
+    }
+}
+
 QTEST_MAIN(tst_QTextBrowser)
 #include "tst_qtextbrowser.moc"