Convert "page" <-> "view" coordinates for PdfSelection and LinkModel

Use FPDF_PageToDevice() rather than subtracting from page height,
in all the remaining places where we were doing that.
Also use FPDF_DeviceToPage() to convert coordinates coming from
the view, for hit-testing links and for selecting ranges of text.
Mark the private utility functions that we're modifying const
while we're at it.

Task-number: QTBUG-100630
Fixes: QTBUG-106565
Change-Id: Ide4f73b80888a0e08381c6e4995f69ebeaa2d12f
Reviewed-by: Axel Spoerl <axel.spoerl@qt.io>
(cherry picked from commit b78c78349d786b294ee0845717ad5db38ae2565a)
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
(cherry picked from commit 5642f33e17aa82fd361b972cd8ac88ba83604273)
Reviewed-by: Qt Cherry-pick Bot <cherrypick_bot@qt-project.org>
(cherry picked from commit 61a65ca969aca620da334de4af14616433ee66a9)

8 files changed, 363 insertions, 40 deletions

diff --git a/src/pdf/qpdfdocument.cpp b/src/pdf/qpdfdocument.cpp
index 11d886dc4..17fdb29b9 100644
--- a/src/pdf/qpdfdocument.cpp
+++ b/src/pdf/qpdfdocument.cpp
@@ -446,7 +446,7 @@ void QPdfDocumentPrivate::fpdf_AddSegment(_FX_DOWNLOADHINTS *pThis, size_t offse
     Q_UNUSED(size);
 }
 
-QString QPdfDocumentPrivate::getText(FPDF_TEXTPAGE textPage, int startIndex, int count)
+QString QPdfDocumentPrivate::getText(FPDF_TEXTPAGE textPage, int startIndex, int count) const
 {
     QList<ushort> buf(count + 1);
     // TODO is that enough space in case one unicode character is more than one in utf-16?
@@ -455,23 +455,38 @@ QString QPdfDocumentPrivate::getText(FPDF_TEXTPAGE textPage, int startIndex, int
     return QString::fromUtf16(reinterpret_cast<const char16_t *>(buf.constData()), len - 1);
 }
 
-QPointF QPdfDocumentPrivate::getCharPosition(FPDF_TEXTPAGE textPage, double pageHeight, int charIndex)
+QPointF QPdfDocumentPrivate::getCharPosition(FPDF_PAGE pdfPage, FPDF_TEXTPAGE textPage, int charIndex) const
 {
     double x, y;
-    int count = FPDFText_CountChars(textPage);
-    bool ok = FPDFText_GetCharOrigin(textPage, qMin(count - 1, charIndex), &x, &y);
-    if (!ok)
-        return QPointF();
-    return QPointF(x, pageHeight - y);
+    const int count = FPDFText_CountChars(textPage);
+    if (FPDFText_GetCharOrigin(textPage, qMin(count - 1, charIndex), &x, &y))
+        return mapPageToView(pdfPage, x, y);
+    return {};
 }
 
-QRectF QPdfDocumentPrivate::getCharBox(FPDF_TEXTPAGE textPage, double pageHeight, int charIndex)
+QRectF QPdfDocumentPrivate::getCharBox(FPDF_PAGE pdfPage, FPDF_TEXTPAGE textPage, int charIndex) const
 {
     double l, t, r, b;
-    bool ok = FPDFText_GetCharBox(textPage, charIndex, &l, &r, &b, &t);
-    if (!ok)
-        return QRectF();
-    return QRectF(l, pageHeight - t, r - l, t - b);
+    if (FPDFText_GetCharBox(textPage, charIndex, &l, &r, &b, &t))
+        return mapPageToView(pdfPage, l, t, r, b);
+    return {};
+}
+
+/*! \internal
+    Convert the point \a x , \a y to the usual 1x (pixels = points)
+    4th-quadrant "view" coordinate system relative to the top-left corner of
+    the rendered page. Some PDF files have internal transforms that make this
+    coordinate system different from "page coordinates", so we cannot just
+    subtract from page height to invert the y coordinates, in general.
+ */
+QPointF QPdfDocumentPrivate::mapPageToView(FPDF_PAGE pdfPage, double x, double y) const
+{
+    const auto pageHeight = FPDF_GetPageHeight(pdfPage);
+    const auto pageWidth = FPDF_GetPageWidth(pdfPage);
+    int rx, ry;
+    if (FPDF_PageToDevice(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, x, y, &rx, &ry))
+        return QPointF(rx, ry);
+    return {};
 }
 
 /*! \internal
@@ -482,14 +497,31 @@ QRectF QPdfDocumentPrivate::getCharBox(FPDF_TEXTPAGE textPage, double pageHeight
     system different from "page coordinates", so we cannot just
     subtract from page height to invert the y coordinates, in general.
  */
-QRectF QPdfDocumentPrivate::mapPageToView(FPDF_PAGE pdfPage, double left, double top, double right, double bottom)
+QRectF QPdfDocumentPrivate::mapPageToView(FPDF_PAGE pdfPage, double left, double top, double right, double bottom) const
 {
     const auto pageHeight = FPDF_GetPageHeight(pdfPage);
     const auto pageWidth = FPDF_GetPageWidth(pdfPage);
     int xfmLeft, xfmTop, xfmRight, xfmBottom;
-    FPDF_PageToDevice(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, left, top, &xfmLeft, &xfmTop);
-    FPDF_PageToDevice(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, right, bottom, &xfmRight, &xfmBottom);
-    return QRectF(xfmLeft, xfmTop, xfmRight - xfmLeft, xfmBottom - xfmTop);
+    if ( FPDF_PageToDevice(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, left, top, &xfmLeft, &xfmTop) &&
+         FPDF_PageToDevice(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, right, bottom, &xfmRight, &xfmBottom) )
+        return QRectF(xfmLeft, xfmTop, xfmRight - xfmLeft, xfmBottom - xfmTop);
+    return {};
+}
+
+/*! \internal
+    Convert the point \a x , \a y \a from the usual 1x (pixels = points)
+    4th-quadrant "view" coordinate system relative to the top-left corner of
+    the rendered page, to "page coordinates" suited to the given \a pdfPage,
+    which may have arbitrary internal transforms.
+ */
+QPointF QPdfDocumentPrivate::mapViewToPage(FPDF_PAGE pdfPage, QPointF position) const
+{
+    const auto pageHeight = FPDF_GetPageHeight(pdfPage);
+    const auto pageWidth = FPDF_GetPageWidth(pdfPage);
+    double rx, ry;
+    if (FPDF_DeviceToPage(pdfPage, 0, 0, qRound(pageWidth), qRound(pageHeight), 0, position.x(), position.y(), &rx, &ry))
+        return QPointF(rx, ry);
+    return {};
 }
 
 QPdfDocumentPrivate::TextPosition QPdfDocumentPrivate::hitTest(int page, QPointF position)
@@ -498,14 +530,14 @@ QPdfDocumentPrivate::TextPosition QPdfDocumentPrivate::hitTest(int page, QPointF
 
     TextPosition result;
     FPDF_PAGE pdfPage = FPDF_LoadPage(doc, page);
-    double pageHeight = FPDF_GetPageHeight(pdfPage);
     FPDF_TEXTPAGE textPage = FPDFText_LoadPage(pdfPage);
-    int hitIndex = FPDFText_GetCharIndexAtPos(textPage, position.x(), pageHeight - position.y(),
+    const QPointF pagePos = mapViewToPage(pdfPage, position);
+    int hitIndex = FPDFText_GetCharIndexAtPos(textPage, pagePos.x(), pagePos.y(),
                                               CharacterHitTolerance, CharacterHitTolerance);
     if (hitIndex >= 0) {
-        QPointF charPos = getCharPosition(textPage, pageHeight, hitIndex);
+        QPointF charPos = getCharPosition(pdfPage, textPage, hitIndex);
         if (!charPos.isNull()) {
-            QRectF charBox = getCharBox(textPage, pageHeight, hitIndex);
+            QRectF charBox = getCharBox(pdfPage, textPage, hitIndex);
             // If the given position is past the end of the line, i.e. if the right edge of the found character's
             // bounding box is closer to it than the left edge is, we say that we "hit" the next character index after
             if (qAbs(charBox.right() - position.x()) < qAbs(charPos.x() - position.x())) {
@@ -949,11 +981,12 @@ QPdfSelection QPdfDocument::getSelection(int page, QPointF start, QPointF end)
 {
     const QPdfMutexLocker lock;
     FPDF_PAGE pdfPage = FPDF_LoadPage(d->doc, page);
-    double pageHeight = FPDF_GetPageHeight(pdfPage);
+    const QPointF pageStart = d->mapViewToPage(pdfPage, start);
+    const QPointF pageEnd = d->mapViewToPage(pdfPage, end);
     FPDF_TEXTPAGE textPage = FPDFText_LoadPage(pdfPage);
-    int startIndex = FPDFText_GetCharIndexAtPos(textPage, start.x(), pageHeight - start.y(),
+    int startIndex = FPDFText_GetCharIndexAtPos(textPage, pageStart.x(), pageStart.y(),
                                                 CharacterHitTolerance, CharacterHitTolerance);
-    int endIndex = FPDFText_GetCharIndexAtPos(textPage, end.x(), pageHeight - end.y(),
+    int endIndex = FPDFText_GetCharIndexAtPos(textPage, pageEnd.x(), pageEnd.y(),
                                               CharacterHitTolerance, CharacterHitTolerance);
 
     QPdfSelection result;
@@ -964,7 +997,7 @@ QPdfSelection QPdfDocument::getSelection(int page, QPointF start, QPointF end)
 
         // If the given end position is past the end of the line, i.e. if the right edge of the last character's
         // bounding box is closer to it than the left edge is, then extend the char range by one
-        QRectF endCharBox = d->getCharBox(textPage, pageHeight, endIndex);
+        QRectF endCharBox = d->getCharBox(pdfPage, textPage, endIndex);
         if (qAbs(endCharBox.right() - end.x()) < qAbs(endCharBox.x() - end.x()))
             ++endIndex;
 
@@ -976,7 +1009,7 @@ QPdfSelection QPdfDocument::getSelection(int page, QPointF start, QPointF end)
         for (int i = 0; i < rectCount; ++i) {
             double l, r, b, t;
             FPDFText_GetRect(textPage, i, &l, &t, &r, &b);
-            QRectF rect(l, pageHeight - t, r - l, t - b);
+            const QRectF rect = d->mapPageToView(pdfPage, l, t, r, b);
             if (hull.isNull())
                 hull = rect;
             else
@@ -1006,7 +1039,6 @@ QPdfSelection QPdfDocument::getSelectionAtIndex(int page, int startIndex, int ma
         return {};
     const QPdfMutexLocker lock;
     FPDF_PAGE pdfPage = FPDF_LoadPage(d->doc, page);
-    double pageHeight = FPDF_GetPageHeight(pdfPage);
     FPDF_TEXTPAGE textPage = FPDFText_LoadPage(pdfPage);
     int pageCount = FPDFText_CountChars(textPage);
     if (startIndex >= pageCount)
@@ -1021,7 +1053,7 @@ QPdfSelection QPdfDocument::getSelectionAtIndex(int page, int startIndex, int ma
         for (int i = 0; i < rectCount; ++i) {
             double l, r, b, t;
             FPDFText_GetRect(textPage, i, &l, &t, &r, &b);
-            QRectF rect(l, pageHeight - t, r - l, t - b);
+            const QRectF rect = d->mapPageToView(pdfPage, l, t, r, b);
             if (hull.isNull())
                 hull = rect;
             else
@@ -1030,7 +1062,7 @@ QPdfSelection QPdfDocument::getSelectionAtIndex(int page, int startIndex, int ma
         }
     }
     if (bounds.isEmpty())
-        hull = QRectF(d->getCharPosition(textPage, pageHeight, startIndex), QSizeF());
+        hull = QRectF(d->getCharPosition(pdfPage, textPage, startIndex), QSizeF());
     qCDebug(qLcDoc) << "on page" << page << "at index" << startIndex << "maxLength" << maxLength
                     << "got" << text.size() << "chars," << rectCount << "rects within" << hull;
 
@@ -1047,7 +1079,6 @@ QPdfSelection QPdfDocument::getAllText(int page)
 {
     const QPdfMutexLocker lock;
     FPDF_PAGE pdfPage = FPDF_LoadPage(d->doc, page);
-    double pageHeight = FPDF_GetPageHeight(pdfPage);
     FPDF_TEXTPAGE textPage = FPDFText_LoadPage(pdfPage);
     int count = FPDFText_CountChars(textPage);
     if (count < 1)
@@ -1059,7 +1090,7 @@ QPdfSelection QPdfDocument::getAllText(int page)
     for (int i = 0; i < rectCount; ++i) {
         double l, r, b, t;
         FPDFText_GetRect(textPage, i, &l, &t, &r, &b);
-        QRectF rect(l, pageHeight - t, r - l, t - b);
+        const QRectF rect = d->mapPageToView(pdfPage, l, t, r, b);
         if (hull.isNull())
             hull = rect;
         else
diff --git a/src/pdf/qpdfdocument_p.h b/src/pdf/qpdfdocument_p.h
index 5ec485012..12092c4ad 100644
--- a/src/pdf/qpdfdocument_p.h
+++ b/src/pdf/qpdfdocument_p.h
@@ -78,10 +78,12 @@ public:
     static int fpdf_GetBlock(void* param, unsigned long position, unsigned char* pBuf, unsigned long size);
     static void fpdf_AddSegment(struct _FX_DOWNLOADHINTS* pThis, size_t offset, size_t size);
     void updateLastError();
-    QString getText(FPDF_TEXTPAGE textPage, int startIndex, int count);
-    QPointF getCharPosition(FPDF_TEXTPAGE textPage, double pageHeight, int charIndex);
-    QRectF getCharBox(FPDF_TEXTPAGE textPage, double pageHeight, int charIndex);
-    QRectF mapPageToView(FPDF_PAGE pdfPage, double left, double top, double right, double bottom);
+    QString getText(FPDF_TEXTPAGE textPage, int startIndex, int count) const;
+    QPointF getCharPosition(FPDF_PAGE pdfPage, FPDF_TEXTPAGE textPage, int charIndex) const;
+    QRectF getCharBox(FPDF_PAGE pdfPage, FPDF_TEXTPAGE textPage, int charIndex) const;
+    QPointF mapPageToView(FPDF_PAGE pdfPage, double x, double y) const;
+    QRectF mapPageToView(FPDF_PAGE pdfPage, double left, double top, double right, double bottom) const;
+    QPointF mapViewToPage(FPDF_PAGE pdfPage, QPointF position) const;
 
     // FPDF takes the rotation parameter as an int.
     // This enum is mapping the int values defined in fpdfview.h:956.
diff --git a/src/pdf/qpdflinkmodel.cpp b/src/pdf/qpdflinkmodel.cpp
index 366b77344..645dff21a 100644
--- a/src/pdf/qpdflinkmodel.cpp
+++ b/src/pdf/qpdflinkmodel.cpp
@@ -179,7 +179,6 @@ void QPdfLinkModelPrivate::update()
         qCWarning(qLcLink) << "failed to load page" << page;
         return;
     }
-    double pageHeight = FPDF_GetPageHeight(pdfPage);
     q->beginResetModel();
     links.clear();
 
@@ -204,8 +203,7 @@ void QPdfLinkModelPrivate::update()
             std::swap(rect.bottom, rect.top);
 
         QPdfLink linkData;
-        linkData.d->rects << QRectF(rect.left, pageHeight - rect.top,
-                               rect.right - rect.left, rect.top - rect.bottom);
+        linkData.d->rects << document->d->mapPageToView(pdfPage, rect.left, rect.top, rect.right, rect.bottom);
         FPDF_DEST dest = FPDFLink_GetDest(doc, linkAnnot);
         FPDF_ACTION action = FPDFLink_GetAction(linkAnnot);
         switch (FPDFAction_GetType(action)) {
@@ -224,7 +222,7 @@ void QPdfLinkModelPrivate::update()
                 break; // at least we got a page number, so the link will jump there
             }
             if (hasX && hasY)
-                linkData.d->location = QPointF(x, pageHeight - y);
+                linkData.d->location = document->d->mapPageToView(pdfPage, x, y);
             if (hasZoom)
                 linkData.d->zoom = zoom;
             break;
@@ -287,7 +285,7 @@ void QPdfLinkModelPrivate::update()
                     double left, top, right, bottom;
                     bool success = FPDFLink_GetRect(webLinks, i, r, &left, &top, &right, &bottom);
                     if (success) {
-                        linkData.d->rects << QRectF(left, pageHeight - top, right - left, top - bottom);
+                        linkData.d->rects << document->d->mapPageToView(pdfPage, left, top, right, bottom);
                         links << linkData;
                     }
                 }
diff --git a/tests/auto/pdf/qpdfdocument/CMakeLists.txt b/tests/auto/pdf/qpdfdocument/CMakeLists.txt
index aadc192b3..b8300ef27 100644
--- a/tests/auto/pdf/qpdfdocument/CMakeLists.txt
+++ b/tests/auto/pdf/qpdfdocument/CMakeLists.txt
@@ -12,5 +12,7 @@ qt_internal_add_test(tst_qpdfdocument
     TESTDATA
         pdf-sample.protected.pdf
         pdf-sample.metadata.pdf
+        rotated_text.pdf
+        tagged_mcr_multipage.pdf
         test.pdf
 )
diff --git a/tests/auto/pdf/qpdfdocument/rotated_text.pdf b/tests/auto/pdf/qpdfdocument/rotated_text.pdf
new file mode 100644
index 000000000..d6d8db84e
--- /dev/null
+++ b/tests/auto/pdf/qpdfdocument/rotated_text.pdf
@@ -0,0 +1,70 @@
+%PDF-1.7
+% ò¤ô
+1 0 obj <<
+  /Type /Catalog
+  /Pages 2 0 R
+>>
+endobj
+2 0 obj <<
+  /Type /Pages
+  /MediaBox [ 0 0 200 200 ]
+  /Count 1
+  /Kids [ 3 0 R ]
+>>
+endobj
+3 0 obj <<
+  /Type /Page
+  /Parent 2 0 R
+  /Resources <<
+    /Font <<
+      /F1 4 0 R
+    >>
+  >>
+  /Contents 5 0 R
+>>
+endobj
+4 0 obj <<
+  /Type /Font
+  /Subtype /Type1
+  /BaseFont /Times-Roman
+>>
+endobj
+5 0 obj <<
+  /Length 406
+>>
+stream
+BT
+0 0 Td
+/F1 12 Tf
+0.70710678118 -0.70710678118 0.70710678118 0.70710678118 100 100 Tm
+(Hello,) Tj
+0 0 Td
+/F1 12 Tf
+-0.70710678118 -0.70710678118 0.70710678118 -0.70710678118 100 100 Tm
+( world!\r\n) Tj
+0 0 Td
+/F1 12 Tf
+-0.70710678118 0.70710678118 -0.70710678118 -0.70710678118 100 100 Tm
+(Goodbye,) Tj
+0 0 Td
+/F1 12 Tf
+0.70710678118 0.70710678118 -0.70710678118 0.70710678118 100 100 Tm
+( world!) Tj
+ET
+endstream
+endobj
+xref
+0 6
+0000000000 65535 f 
+0000000015 00000 n 
+0000000068 00000 n 
+0000000161 00000 n 
+0000000287 00000 n 
+0000000365 00000 n 
+trailer <<
+  /Root 1 0 R
+  /Size 6
+>>
+startxref
+823
+%%EOF
diff --git a/tests/auto/pdf/qpdfdocument/tagged_mcr_multipage.pdf b/tests/auto/pdf/qpdfdocument/tagged_mcr_multipage.pdf
new file mode 100644
index 000000000..fcc5fafda
--- /dev/null
+++ b/tests/auto/pdf/qpdfdocument/tagged_mcr_multipage.pdf
@@ -0,0 +1,136 @@
+%PDF-1.7
+% ò¤ô
+1 0 obj <<
+  /Type /Catalog
+  /MarkInfo <<
+    /Type /MarkInfo
+    /Marked true
+  >>
+  /Pages 2 0 R
+  /StructTreeRoot 8 0 R
+>>
+endobj
+2 0 obj <<
+  /Type /Pages
+  /CropBox [ 10.8197 8.459 605.705 801.639 ]
+  /MediaBox [ 0.0 0.0 616.721 809.902 ]
+  /Count 2
+  /Kids [
+    4 0 R
+    6 0 R
+  ]
+>>
+endobj
+3 0 obj <<
+  /Type /Font
+  /Subtype /Type1
+  /BaseFont /Times-Roman
+>>
+endobj
+4 0 obj <<
+  /Type /Page
+  /Tabs /S
+  /Parent 2 0 R
+  /StructParents 0
+  /Contents 5 0 R
+  /Resources <<
+    /ProcSet [/PDF /Text]
+    /Font <<
+      /F1 3 0 R
+    >>
+  >>
+>>
+endobj
+5 0 obj <<
+  /Length 83
+>>
+stream
+BT
+/Document <</MCID 0 >>BDC
+0 i
+/F1 1 Tf
+12 0 0 12 43.073 771.625 Tm
+(1)Tj
+EMC
+ET
+endstream
+endobj
+6 0 obj <<
+  /Type /Page
+  /Tabs /S
+  /Parent 2 0 R
+  /StructParents 1
+  /Contents 7 0 R
+  /Resources <<
+    /ProcSet [/PDF /Text]
+    /Font <<
+      /F1 3 0 R
+    >>
+  >>
+>>
+endobj
+7 0 obj <<
+  /Length 83
+>>
+stream
+BT
+/Document <</MCID 0 >>BDC
+0 i
+/F1 1 Tf
+12 0 0 12 43.073 771.625 Tm
+(2)Tj
+EMC
+ET
+endstream
+endobj
+8 0 obj <<
+  /Type /StructTreeRoot
+  /K 10 0 R
+  /ParentTree 9 0 R
+  /ParentTreeNextKey 2
+>>
+endobj
+9 0 obj <<
+  /Nums [
+    0
+    [10 0 R]
+    1
+    [10 0 R]
+  ]
+>>
+endobj
+10 0 obj <<
+  /T ()
+  /S /Document
+  /P 8 0 R
+  /Pg 4 0 R
+  /K [
+    0
+    <<
+      /MCID 0
+      /Pg 6 0 R
+      /Type /MCR
+    >>
+  ]
+>>
+%endobj
+xref
+0 11
+0000000000 65535 f 
+0000000015 00000 n 
+0000000149 00000 n 
+0000000315 00000 n 
+0000000393 00000 n 
+0000000575 00000 n 
+0000000709 00000 n 
+0000000891 00000 n 
+0000001025 00000 n 
+0000001125 00000 n 
+0000001198 00000 n 
+trailer <<
+  /Root 1 0 R
+  /Size 11
+>>
+startxref
+1345
+%%EOF
diff --git a/tests/auto/pdf/qpdfdocument/tst_qpdfdocument.cpp b/tests/auto/pdf/qpdfdocument/tst_qpdfdocument.cpp
index 6bb007e7f..d222bff0c 100644
--- a/tests/auto/pdf/qpdfdocument/tst_qpdfdocument.cpp
+++ b/tests/auto/pdf/qpdfdocument/tst_qpdfdocument.cpp
@@ -36,6 +36,10 @@ private slots:
     void passwordClearedOnClose();
     void metaData();
     void pageLabels();
+    void getSelection_data();
+    void getSelection();
+    void getSelectionAtIndex_data();
+    void getSelectionAtIndex();
 
 private:
     void consistencyCheck(QPdfDocument &doc) const;
@@ -394,6 +398,86 @@ void tst_QPdfDocument::pageLabels()
     QCOMPARE(doc.pageLabel(2), "i"); // i of the tiger!
 }
 
+void tst_QPdfDocument::getSelection_data()
+{
+    QTest::addColumn<QString>("pdfPath");
+    QTest::addColumn<int>("page");
+    QTest::addColumn<QPointF>("start");
+    QTest::addColumn<QPointF>("end");
+    QTest::addColumn<QString>("expectedText");
+    QTest::addColumn<int>("expectedStartIndex");
+    QTest::addColumn<int>("expectedEndIndex");
+    QTest::addColumn<QRect>("expectedBounds");
+    QTest::addColumn<int>("expectedPolygonCount");
+
+    QTest::newRow("raid") << QFINDTESTDATA("test.pdf")
+            << 1 << QPointF(316.4, 206) << QPointF(339, 201)
+            << "raid" << 80 << 84 << QRect(316, 201, 21, 12) << 1;
+    QTest::newRow("rotated text") << QFINDTESTDATA("rotated_text.pdf")
+            << 0 << QPointF(102, 94) << QPointF(125, 73)
+            << "world!" << 25 << 31 << QRect(98, 70, 26, 28) << 1;
+}
+
+void tst_QPdfDocument::getSelection()
+{
+    QFETCH(QString, pdfPath);
+    QFETCH(int, page);
+    QFETCH(QPointF, start);
+    QFETCH(QPointF, end);
+    QFETCH(QString, expectedText);
+    QFETCH(int, expectedStartIndex);
+    QFETCH(int, expectedEndIndex);
+    QFETCH(QRect, expectedBounds);
+    QFETCH(int, expectedPolygonCount);
+
+    QPdfDocument doc;
+    QCOMPARE(doc.load(pdfPath), QPdfDocument::Error::None);
+
+    QPdfSelection sel = doc.getSelection(page, start, end);
+    QCOMPARE(sel.text(), expectedText);
+    QCOMPARE(sel.startIndex(), expectedStartIndex);
+    QCOMPARE(sel.endIndex(), expectedEndIndex);
+    QCOMPARE(sel.boundingRectangle().toRect(), expectedBounds);
+    QCOMPARE(sel.bounds().size(), expectedPolygonCount);
+}
+
+void tst_QPdfDocument::getSelectionAtIndex_data()
+{
+    QTest::addColumn<QString>("pdfPath");
+    QTest::addColumn<int>("page");
+    QTest::addColumn<int>("start");
+    QTest::addColumn<int>("maxLen");
+    QTest::addColumn<QString>("expectedText");
+    QTest::addColumn<QRect>("expectedBounds");
+    QTest::addColumn<int>("expectedPolygonCount");
+
+    QTest::newRow("raid") << QFINDTESTDATA("test.pdf")
+            << 1 << 80 << 4 << "raid" << QRect(316, 201, 21, 12) << 1;
+    QTest::newRow("rotated text") << QFINDTESTDATA("rotated_text.pdf")
+            << 0 << 7 << 6 << "world!" << QRect(76, 102, 26, 28) << 1;
+    QTest::newRow("displaced text") << QFINDTESTDATA("tagged_mcr_multipage.pdf")
+            << 0 << 0 << 10 << "1" << QRect(34, 22, 3, 8) << 1;
+}
+
+void tst_QPdfDocument::getSelectionAtIndex()
+{
+    QFETCH(QString, pdfPath);
+    QFETCH(int, page);
+    QFETCH(int, start);
+    QFETCH(int, maxLen);
+    QFETCH(QString, expectedText);
+    QFETCH(QRect, expectedBounds);
+    QFETCH(int, expectedPolygonCount);
+
+    QPdfDocument doc;
+    QCOMPARE(doc.load(pdfPath), QPdfDocument::Error::None);
+
+    QPdfSelection sel = doc.getSelectionAtIndex(page, start, maxLen);
+    QCOMPARE(sel.text(), expectedText);
+    QCOMPARE(sel.boundingRectangle().toRect(), expectedBounds);
+    QCOMPARE(sel.bounds().size(), expectedPolygonCount);
+}
+
 QTEST_MAIN(tst_QPdfDocument)
 
 #include "tst_qpdfdocument.moc"
diff --git a/tests/auto/pdfquick/multipageview/tst_multipageview.cpp b/tests/auto/pdfquick/multipageview/tst_multipageview.cpp
index b64cd47b1..c5e0b30db 100644
--- a/tests/auto/pdfquick/multipageview/tst_multipageview.cpp
+++ b/tests/auto/pdfquick/multipageview/tst_multipageview.cpp
@@ -62,7 +62,7 @@ void tst_MultiPageView::internalLink_data()
     QTest::addColumn<qreal>("expectedZoom");
     QTest::addColumn<QPoint>("expectedScroll");
 
-    QTest::newRow("first link") << 0 << 1 << qreal(1) << QPoint(134, 1276);
+    QTest::newRow("first link") << 0 << 1 << qreal(1) << QPoint(134, 1286);
     // TODO fails because it zooms out, and the view leaves gaps between pages currently
 //    QTest::newRow("second link") << 1 << 2 << qreal(0.5) << QPoint(0, 717);
 }