tdf#148198: merge identical hyperlinks of adjacent text ranges on ODF export

The true hyperlink boundaries are available as SwpHints starts/ends, which are used
in DOC(X) export (see SwWW8AttrIter::OutAttrWithRange).

However, I don't see a reasonable way to expose this information to xmloff, so I
decided instead to just merge the identical hyperlink properties of adjacent ranges
into a single hyperlink. This will allow to fix already split hyperlinks saved in
previous versions. The downside is that this disallows to have separate adjacent
identical hyperlinks - I hope that this would not be a real issue.

Change-Id: I901e6035a5e89bc515b5742c6a5f564c77faf05b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/137013
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

3 files changed, 169 insertions, 184 deletions

diff --git a/xmloff/source/text/txtflde.cxx b/xmloff/source/text/txtflde.cxx
index 242b641369c8..54341aab0b01 100644
--- a/xmloff/source/text/txtflde.cxx
+++ b/xmloff/source/text/txtflde.cxx
@@ -991,44 +991,15 @@ void XMLTextFieldExport::ExportField(
                     ? aStates
                     : nullptr;
 
-    // find out whether we need to set the style or hyperlink
-    bool bHasHyperlink;
+    // find out whether we need to set the style
     bool bIsUICharStyle;
     bool bHasAutoStyle;
     OUString sStyle = GetExport().GetTextParagraphExport()->
-        FindTextStyleAndHyperlink( xRangePropSet, bHasHyperlink, bIsUICharStyle,
-                                   bHasAutoStyle, pStates );
+        FindTextStyle( xRangePropSet, bIsUICharStyle, bHasAutoStyle, pStates );
     bool bHasStyle = !sStyle.isEmpty();
 
-    // export hyperlink (if we have one)
-    Reference < XPropertySetInfo > xRangePropSetInfo;
-    if( bHasHyperlink )
-    {
-        Reference<XPropertyState> xRangePropState( xRangePropSet, UNO_QUERY );
-        xRangePropSetInfo = xRangePropSet->getPropertySetInfo();
-        bHasHyperlink =
-            GetExport().GetTextParagraphExport()->addHyperlinkAttributes(
-                xRangePropSet, xRangePropState,
-                xRangePropSetInfo );
-    }
-    SvXMLElementExport aHyperlink( GetExport(), bHasHyperlink,
-                                   XML_NAMESPACE_TEXT, XML_A,
-                                   false, false );
-
-    if( bHasHyperlink )
-    {
-        // export events (if supported)
-        OUString sHyperLinkEvents("HyperLinkEvents");
-        if (xRangePropSetInfo->hasPropertyByName(sHyperLinkEvents))
-        {
-            Any aAny = xRangePropSet->getPropertyValue(sHyperLinkEvents);
-            Reference<XNameReplace> xName;
-            aAny >>= xName;
-            GetExport().GetEventExport().Export(xName, false);
-        }
-    }
-
     {
+        Reference<XPropertySetInfo> xRangePropSetInfo;
         XMLTextCharStyleNamesElementExport aCharStylesExport(
             GetExport(), bIsUICharStyle &&
                          GetExport().GetTextParagraphExport()
diff --git a/xmloff/source/text/txtftne.cxx b/xmloff/source/text/txtftne.cxx
index ff37cd990b32..333ec7c6f783 100644
--- a/xmloff/source/text/txtftne.cxx
+++ b/xmloff/source/text/txtftne.cxx
@@ -89,38 +89,10 @@ void XMLTextParagraphExport::exportTextFootnote(
     {
         // create span (for citation mark) if necessary; footnote content
         // will be handled via exportTextFootnoteHelper, exportText
-        bool bHasHyperlink;
         bool bIsUICharStyle = false;
         bool bHasAutoStyle = false;
 
-        OUString sStyle = FindTextStyleAndHyperlink( rPropSet, bHasHyperlink,
-                                                     bIsUICharStyle, bHasAutoStyle );
-
-        // export hyperlink (if we have one)
-        Reference < XPropertySetInfo > xPropSetInfo;
-        if( bHasHyperlink )
-        {
-            Reference<XPropertyState> xPropState( rPropSet, UNO_QUERY );
-            xPropSetInfo = rPropSet->getPropertySetInfo();
-            bHasHyperlink =
-                addHyperlinkAttributes( rPropSet, xPropState, xPropSetInfo );
-        }
-        SvXMLElementExport aHyperlink( GetExport(), bHasHyperlink,
-                                       XML_NAMESPACE_TEXT, XML_A,
-                                       false, false );
-
-        if( bHasHyperlink )
-        {
-            // export events (if supported)
-            OUString sHyperLinkEvents("HyperLinkEvents");
-            if (xPropSetInfo->hasPropertyByName(sHyperLinkEvents))
-            {
-                Any a = rPropSet->getPropertyValue(sHyperLinkEvents);
-                Reference<XNameReplace> xName;
-                a >>= xName;
-                GetExport().GetEventExport().Export(xName, false);
-            }
-        }
+        OUString sStyle = FindTextStyle( rPropSet, bIsUICharStyle, bHasAutoStyle );
 
         {
             XMLTextCharStyleNamesElementExport aCharStylesExport(
diff --git a/xmloff/source/text/txtparae.cxx b/xmloff/source/text/txtparae.cxx
index 5b2cb233dd38..13378e3902e2 100644
--- a/xmloff/source/text/txtparae.cxx
+++ b/xmloff/source/text/txtparae.cxx
@@ -280,6 +280,148 @@ namespace
 
             void ExportParameter(const OUString& sKey, const OUString& sValue);
     };
+
+    struct HyperlinkData
+    {
+        OUString href, name, targetFrame, ustyleName, vstyleName;
+        bool serverMap = false;
+        css::uno::Reference<css::container::XNameReplace> events;
+
+        HyperlinkData() = default;
+        HyperlinkData(const css::uno::Reference<css::beans::XPropertySet>& rPropSet);
+
+        bool operator==(const HyperlinkData&);
+        bool operator!=(const HyperlinkData& rOther) { return !operator==(rOther); }
+
+        bool addHyperlinkAttributes(SvXMLExport& rExport);
+        void exportEvents(SvXMLExport& rExport);
+    };
+
+    HyperlinkData::HyperlinkData(const css::uno::Reference<css::beans::XPropertySet>& rPropSet)
+    {
+        const css::uno::Reference<css::beans::XPropertyState> xPropState(rPropSet, UNO_QUERY);
+        const auto xPropSetInfo(rPropSet->getPropertySetInfo());
+        if (xPropSetInfo->hasPropertyByName(gsHyperLinkURL)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE == xPropState->getPropertyState(gsHyperLinkURL)))
+        {
+            rPropSet->getPropertyValue(gsHyperLinkURL) >>= href;
+        }
+
+        if (href.isEmpty())
+            return;
+
+        if (xPropSetInfo->hasPropertyByName(gsHyperLinkName)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE == xPropState->getPropertyState(gsHyperLinkName)))
+        {
+            rPropSet->getPropertyValue(gsHyperLinkName) >>= name;
+        }
+
+        if (xPropSetInfo->hasPropertyByName(gsHyperLinkTarget)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE == xPropState->getPropertyState(gsHyperLinkTarget)))
+        {
+            rPropSet->getPropertyValue(gsHyperLinkTarget) >>= targetFrame;
+        }
+
+        if (xPropSetInfo->hasPropertyByName(gsServerMap)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE == xPropState->getPropertyState(gsServerMap)))
+        {
+            serverMap = *o3tl::doAccess<bool>(rPropSet->getPropertyValue(gsServerMap));
+        }
+
+        if (xPropSetInfo->hasPropertyByName(gsUnvisitedCharStyleName)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE
+                       == xPropState->getPropertyState(gsUnvisitedCharStyleName)))
+        {
+            rPropSet->getPropertyValue(gsUnvisitedCharStyleName) >>= ustyleName;
+        }
+
+        if (xPropSetInfo->hasPropertyByName(gsVisitedCharStyleName)
+            && (!xPropState.is()
+                || PropertyState_DIRECT_VALUE
+                       == xPropState->getPropertyState(gsVisitedCharStyleName)))
+        {
+            rPropSet->getPropertyValue(gsVisitedCharStyleName) >>= vstyleName;
+        }
+
+        static constexpr OUStringLiteral sHyperLinkEvents(u"HyperLinkEvents");
+        if (xPropSetInfo->hasPropertyByName(sHyperLinkEvents))
+        {
+            events.set(rPropSet->getPropertyValue(sHyperLinkEvents), uno::UNO_QUERY);
+        }
+    }
+
+    bool HyperlinkData::operator==(const HyperlinkData& rOther)
+    {
+        if (href != rOther.href || name != rOther.name || targetFrame != rOther.targetFrame
+            || ustyleName != rOther.ustyleName || vstyleName != rOther.vstyleName
+            || serverMap != rOther.serverMap)
+            return false;
+
+        if (events == rOther.events)
+            return true;
+        if (!events || !rOther.events)
+            return false;
+
+        const css::uno::Sequence<OUString> aNames = events->getElementNames();
+        if (aNames != rOther.events->getElementNames())
+            return false;
+        for (const auto& rName : aNames)
+        {
+            const css::uno::Any aAny = events->getByName(rName);
+            const css::uno::Any aOtherAny = rOther.events->getByName(rName);
+            if (aAny != aOtherAny)
+                return false;
+        }
+        return true;
+    }
+
+    bool HyperlinkData::addHyperlinkAttributes(SvXMLExport& rExport)
+    {
+        if (href.isEmpty())
+        {
+            // hyperlink without a URL does not make sense
+            OSL_ENSURE(false, "hyperlink without a URL --> no export to ODF");
+            return false;
+        }
+
+        rExport.AddAttribute(XML_NAMESPACE_XLINK, XML_TYPE, XML_SIMPLE);
+        rExport.AddAttribute(XML_NAMESPACE_XLINK, XML_HREF, rExport.GetRelativeReference(href));
+
+        if (!name.isEmpty())
+            rExport.AddAttribute(XML_NAMESPACE_OFFICE, XML_NAME, name);
+
+        if (!targetFrame.isEmpty())
+        {
+            rExport.AddAttribute(XML_NAMESPACE_OFFICE, XML_TARGET_FRAME_NAME, targetFrame);
+            enum XMLTokenEnum eTok = targetFrame == "_blank" ? XML_NEW : XML_REPLACE;
+            rExport.AddAttribute(XML_NAMESPACE_XLINK, XML_SHOW, eTok);
+        }
+
+        if (serverMap)
+            rExport.AddAttribute(XML_NAMESPACE_OFFICE, XML_SERVER_MAP, XML_TRUE);
+
+        if (!ustyleName.isEmpty())
+            rExport.AddAttribute(XML_NAMESPACE_TEXT, XML_STYLE_NAME,
+                                 rExport.EncodeStyleName(ustyleName));
+
+        if (!vstyleName.isEmpty())
+            rExport.AddAttribute(XML_NAMESPACE_TEXT, XML_VISITED_STYLE_NAME,
+                                 rExport.EncodeStyleName(vstyleName));
+
+        return true;
+    }
+
+    void HyperlinkData::exportEvents(SvXMLExport& rExport)
+    {
+        // export events (if supported)
+        if (events)
+            rExport.GetEventExport().Export(events, false);
+    }
 }
 
 namespace xmloff
@@ -769,9 +911,8 @@ OUString XMLTextParagraphExport::Find(
     return sName;
 }
 
-OUString XMLTextParagraphExport::FindTextStyleAndHyperlink(
+OUString XMLTextParagraphExport::FindTextStyle(
            const Reference < XPropertySet > & rPropSet,
-        bool& rbHyperlink,
         bool& rbHasCharStyle,
         bool& rbHasAutoStyle,
         const XMLPropertyState** ppAddStates ) const
@@ -781,7 +922,7 @@ OUString XMLTextParagraphExport::FindTextStyleAndHyperlink(
 
     // Get parent and remove hyperlinks (they aren't of interest)
     OUString sName;
-    rbHyperlink = rbHasCharStyle = rbHasAutoStyle = false;
+    rbHasCharStyle = rbHasAutoStyle = false;
     sal_uInt16 nIgnoreProps = 0;
     rtl::Reference< XMLPropertySetMapper > xPM(xPropMapper->getPropertySetMapper());
     ::std::vector< XMLPropertyState >::iterator aFirstDel = aPropStates.end();
@@ -808,7 +949,6 @@ OUString XMLTextParagraphExport::FindTextStyleAndHyperlink(
             nIgnoreProps++;
             break;
         case CTF_HYPERLINK_URL:
-            rbHyperlink = true;
             i->mnIndex = -1;
             if( nIgnoreProps )
                 aSecondDel = i;
@@ -2123,12 +2263,29 @@ void XMLTextParagraphExport::exportTextRangeEnumeration(
      * bookmarks are used instead of fieldmarks. */
     FieldmarkType openFieldMark = NONE;
 
+    std::optional<SvXMLElementExport> oTextA;
+    HyperlinkData aHyperlinkData;
+
     while( rTextEnum->hasMoreElements() )
     {
         Reference<XPropertySet> xPropSet(rTextEnum->nextElement(), UNO_QUERY);
         Reference < XTextRange > xTxtRange(xPropSet, uno::UNO_QUERY);
         Reference<XPropertySetInfo> xPropInfo(xPropSet->getPropertySetInfo());
 
+        if (!bAutoStyles)
+        {
+            if (HyperlinkData aNewHyperlinkData(xPropSet); aNewHyperlinkData != aHyperlinkData)
+            {
+                aHyperlinkData = aNewHyperlinkData;
+                oTextA.reset();
+                if (aHyperlinkData.addHyperlinkAttributes(GetExport()))
+                {
+                    oTextA.emplace(GetExport(), true, XML_NAMESPACE_TEXT, XML_A, false, false);
+                    aHyperlinkData.exportEvents(GetExport());
+                }
+            }
+        }
+
         if (xPropInfo->hasPropertyByName(gsTextPortionType))
         {
             OUString sType;
@@ -2890,7 +3047,6 @@ void XMLTextParagraphExport::exportAnyTextFrame(
     else
     {
         Reference< XPropertySetInfo > xPropSetInfo(xPropSet->getPropertySetInfo());
-        Reference< XPropertyState > xPropState( xPropSet, UNO_QUERY );
         {
             bool bAddCharStyles = pRangePropSet &&
                 lcl_txtpara_isBoundAsChar( xPropSet, xPropSetInfo );
@@ -2901,10 +3057,7 @@ void XMLTextParagraphExport::exportAnyTextFrame(
             OUString sStyle;
 
             if( bAddCharStyles )
-            {
-                bool bDummy;
-                sStyle = FindTextStyleAndHyperlink( *pRangePropSet, bDummy, bIsUICharStyle, bHasAutoStyle );
-            }
+                sStyle = FindTextStyle( *pRangePropSet, bIsUICharStyle, bHasAutoStyle );
             else
                 bIsUICharStyle = false;
 
@@ -2924,8 +3077,7 @@ void XMLTextParagraphExport::exportAnyTextFrame(
                 {
                     SvXMLElementExport aElement( GetExport(),
                         FrameType::Shape != eType &&
-                        addHyperlinkAttributes( xPropSet,
-                                                xPropState,xPropSetInfo ),
+                            HyperlinkData(xPropSet).addHyperlinkAttributes(GetExport()),
                         XML_NAMESPACE_DRAW, XML_A, false, false );
                     switch( eType )
                     {
@@ -3364,89 +3516,6 @@ void XMLTextParagraphExport::exportTitleAndDescription(
     }
 }
 
-bool XMLTextParagraphExport::addHyperlinkAttributes(
-    const Reference< XPropertySet > & rPropSet,
-    const Reference< XPropertyState > & rPropState,
-    const Reference< XPropertySetInfo > & rPropSetInfo )
-{
-    OUString sHRef;
-
-    if( rPropSetInfo->hasPropertyByName( gsHyperLinkURL ) &&
-        ( !rPropState.is() || PropertyState_DIRECT_VALUE ==
-                    rPropState->getPropertyState( gsHyperLinkURL ) ) )
-    {
-        rPropSet->getPropertyValue( gsHyperLinkURL ) >>= sHRef;
-    }
-
-    if ( sHRef.isEmpty() )
-    {
-        // hyperlink without a URL does not make sense
-        OSL_ENSURE( false, "hyperlink without a URL --> no export to ODF" );
-        return false;
-    }
-
-    GetExport().AddAttribute(XML_NAMESPACE_XLINK, XML_TYPE, XML_SIMPLE);
-    GetExport().AddAttribute(XML_NAMESPACE_XLINK, XML_HREF,
-                             GetExport().GetRelativeReference(sHRef));
-
-    if ( rPropSetInfo->hasPropertyByName( gsHyperLinkName )
-         && ( !rPropState.is()
-              || PropertyState_DIRECT_VALUE == rPropState->getPropertyState( gsHyperLinkName ) ) )
-    {
-        OUString sName;
-        rPropSet->getPropertyValue( gsHyperLinkName ) >>= sName;
-        if( !sName.isEmpty() )
-            GetExport().AddAttribute(XML_NAMESPACE_OFFICE, XML_NAME, sName);
-    }
-
-    if ( rPropSetInfo->hasPropertyByName( gsHyperLinkTarget )
-         && ( !rPropState.is()
-              || PropertyState_DIRECT_VALUE == rPropState->getPropertyState( gsHyperLinkTarget ) ) )
-    {
-        OUString sTargetFrame;
-        rPropSet->getPropertyValue( gsHyperLinkTarget ) >>= sTargetFrame;
-        if( !sTargetFrame.isEmpty() )
-        {
-            GetExport().AddAttribute(XML_NAMESPACE_OFFICE, XML_TARGET_FRAME_NAME, sTargetFrame);
-            enum XMLTokenEnum eTok = sTargetFrame == "_blank" ? XML_NEW : XML_REPLACE;
-            GetExport().AddAttribute(XML_NAMESPACE_XLINK, XML_SHOW, eTok);
-        }
-    }
-
-    if ( rPropSetInfo->hasPropertyByName( gsServerMap )
-         && ( !rPropState.is()
-              || PropertyState_DIRECT_VALUE == rPropState->getPropertyState( gsServerMap ) ) )
-    {
-        bool bServerMap = *o3tl::doAccess<bool>(rPropSet->getPropertyValue( gsServerMap ));
-        if ( bServerMap )
-            GetExport().AddAttribute(XML_NAMESPACE_OFFICE, XML_SERVER_MAP, XML_TRUE);
-    }
-
-    if ( rPropSetInfo->hasPropertyByName( gsUnvisitedCharStyleName )
-         && ( !rPropState.is()
-              || PropertyState_DIRECT_VALUE == rPropState->getPropertyState( gsUnvisitedCharStyleName ) ) )
-    {
-        OUString sUStyleName;
-        rPropSet->getPropertyValue( gsUnvisitedCharStyleName ) >>= sUStyleName;
-        if( !sUStyleName.isEmpty() )
-            GetExport().AddAttribute(XML_NAMESPACE_TEXT, XML_STYLE_NAME,
-                                     GetExport().EncodeStyleName(sUStyleName));
-    }
-
-    if ( rPropSetInfo->hasPropertyByName( gsVisitedCharStyleName )
-         && ( !rPropState.is()
-              || PropertyState_DIRECT_VALUE == rPropState->getPropertyState( gsVisitedCharStyleName ) ) )
-    {
-        OUString sVStyleName;
-        rPropSet->getPropertyValue( gsVisitedCharStyleName ) >>= sVStyleName;
-        if( !sVStyleName.isEmpty() )
-            GetExport().AddAttribute(XML_NAMESPACE_TEXT, XML_VISITED_STYLE_NAME,
-                                     GetExport().EncodeStyleName(sVStyleName));
-    }
-
-    return true;
-}
-
 void XMLTextParagraphExport::exportTextRangeSpan(
     const css::uno::Reference< css::text::XTextRange > & rTextRange,
     Reference< XPropertySet > const & xPropSet,
@@ -3492,40 +3561,13 @@ void XMLTextParagraphExport::exportTextRange(
     }
     else
     {
-        bool bHyperlink = false;
         bool bIsUICharStyle = false;
         bool bHasAutoStyle = false;
         const OUString sStyle(
-            FindTextStyleAndHyperlink( xPropSet, bHyperlink, bIsUICharStyle, bHasAutoStyle ) );
+            FindTextStyle( xPropSet, bIsUICharStyle, bHasAutoStyle ) );
 
         Reference < XPropertySetInfo > xPropSetInfo;
-        bool bHyperlinkAttrsAdded = false;
-        if ( bHyperlink )
-        {
-            Reference< XPropertyState > xPropState( xPropSet, UNO_QUERY );
-            xPropSetInfo.set( xPropSet->getPropertySetInfo() );
-            bHyperlinkAttrsAdded = addHyperlinkAttributes( xPropSet, xPropState, xPropSetInfo );
-        }
-
-        if ( bHyperlink && bHyperlinkAttrsAdded )
-        {
-            SvXMLElementExport aElem( GetExport(), true, XML_NAMESPACE_TEXT, XML_A, false, false );
-
-            // export events (if supported)
-            OUString sHyperLinkEvents(
-                "HyperLinkEvents");
-            if (xPropSetInfo->hasPropertyByName(sHyperLinkEvents))
-            {
-                Reference< XNameReplace > xName( xPropSet->getPropertyValue( sHyperLinkEvents ), uno::UNO_QUERY );
-                GetExport().GetEventExport().Export( xName, false );
-            }
-
-            exportTextRangeSpan( rTextRange, xPropSet, xPropSetInfo, bIsUICharStyle, bHasAutoStyle, sStyle, rPrevCharIsSpace, openFieldMark );
-        }
-        else
-        {
-            exportTextRangeSpan( rTextRange, xPropSet, xPropSetInfo, bIsUICharStyle, bHasAutoStyle, sStyle, rPrevCharIsSpace, openFieldMark );
-        }
+        exportTextRangeSpan( rTextRange, xPropSet, xPropSetInfo, bIsUICharStyle, bHasAutoStyle, sStyle, rPrevCharIsSpace, openFieldMark );
     }
 }