Make the context API in ExtractorEngine type-independent as well

autotests/extractortest.cpp

@@ -58,7 +58,7 @@ private Q_SLOTS:
         QTest::addColumn<QString>("contextFile");
         QTest::addColumn<QString>("inputFile");
 
-        for (const QDir &baseDir :  {QStringLiteral(SOURCE_DIR "/extractordata"), QStringLiteral(SOURCE_DIR "/../../kitinerary-tests")}) {
+        for (const QDir baseDir :  {QStringLiteral(SOURCE_DIR "/extractordata"), QStringLiteral(SOURCE_DIR "/../../kitinerary-tests")}) {
             if (!baseDir.exists()) {
                 continue;
             }
@@ -96,14 +96,13 @@ private Q_SLOTS:
         if (cf.open(QFile::ReadOnly)) {
             contextMsg.setContent(cf.readAll());
             contextMsg.parse();
-            m_engine.setContext(&contextMsg);
+            m_engine.setContext(QVariant::fromValue(&contextMsg), u"message/rfc822");
         } else if (inputFile.endsWith(QLatin1String(".eml"))) {
             contextMsg.setContent(inFile.readAll());
             inFile.seek(0);
             contextMsg.parse();
-            m_engine.setContext(&contextMsg);
+            m_engine.setContext(QVariant::fromValue(&contextMsg), u"message/rfc822");
         } else {
-            m_engine.setContext(nullptr);
             m_engine.setContextDate(QDateTime({2018, 1, 1}, {0, 0}));
         }
 

src/lib/engine/extractorengine.cpp

@@ -109,6 +109,8 @@ public:
     std::vector<GenericExtractor::Result> m_genericResults;
     QJsonArray m_result;
     QJSEngine m_engine;
+    ExtractorDocumentNode m_rootNode;
+    ExtractorDocumentNode m_contextNode;
     ExtractorDocumentNodeFactory m_nodeFactory;
     ExtractorRepository m_repo;
     BarcodeDecoder m_barcodeDecoder;
@@ -248,6 +250,9 @@ void ExtractorEngine::clear()
     d->m_ownedMimeContent.reset();
     d->m_barcodeDecoder.clearCache();
     d->m_usedExtractor.clear();
+
+    d->m_rootNode = {};
+    d->m_contextNode = {};
 }
 
 void ExtractorEnginePrivate::resetContent()
@@ -309,6 +314,8 @@ void ExtractorEngine::setData(const QByteArray &data, QStringView fileName, QStr
     const auto nameType = ExtractorInput::typeFromFileName(fileName.toString());
     const auto contentType = ExtractorInput::typeFromContent(data);
     setData(data, nameType == ExtractorInput::Unknown ? contentType : nameType);
+
+    d->m_rootNode = d->m_nodeFactory.createNode(data, fileName, mimeType);
 }
 
 void ExtractorEngine::setData(const QByteArray &data, ExtractorInput::Type type)
@@ -324,6 +331,7 @@ void ExtractorEngine::setData(const QByteArray &data, ExtractorInput::Type type)
 
 void ExtractorEngine::setContent(const QVariant &data, QStringView mimeType)
 {
+    d->m_rootNode = d->m_nodeFactory.createNode(data, mimeType);
     // ### temporary scaffolding until we have the new extractor engine
     switch (ExtractorInput::typeFromMimeType(mimeType.toString())) {
         case ExtractorInput::Text:
@@ -467,6 +475,20 @@ void ExtractorEngine::setContext(KMime::Content *context)
     setContextDate({});
 }
 
+void ExtractorEngine::setContext(const QVariant &data, QStringView mimeType)
+{
+    // ### temporary migration scaffolding
+    if (mimeType == QLatin1String("message/rfc822")) {
+        if (data.value<KMime::Content*>()) {
+            setContext(data.value<KMime::Content*>());
+        } else {
+            setContext(data.value<KMime::Message*>());
+        }
+    }
+
+    d->m_contextNode = d->m_nodeFactory.createNode(data, mimeType);
+}
+
 void ExtractorEnginePrivate::setContext(PdfDocument *pdf)
 {
     auto dt = pdf->modificationTime();
@@ -480,6 +502,8 @@ void ExtractorEnginePrivate::setContext(PdfDocument *pdf)
 
 void ExtractorEngine::setContextDate(const QDateTime &dt)
 {
+    d->m_contextNode.setContextDateTime(dt);
+
     d->m_context->m_senderDate = dt;
     d->m_jsonLdApi->setContextDate(dt);
     d->m_barcodeApi->setContextDate(dt);

src/lib/engine/extractorengine.h

@@ -179,7 +179,15 @@ public:
      *  Calling this method is not necessary when using setContent,
      *  only when using any of the other content setter methods directly.
      */
-    void setContext(KMime::Content *context);
+    [[deprecated("set setContext()")]] void setContext(KMime::Content *context);
+
+    /** Provide a document part that is only used to determine which extractor to use,
+     *  but not for extraction itself.
+     *  This can for example be the MIME message part wrapping a document to extract.
+     *  Using this is not necessary when this document part is already included in
+     *  what is passed to setContent() already anyway.
+     */
+    void setContext(const QVariant &data, QStringView mimeType);
 
     /** Set the date the extracted document has been issued at.
      *  This does not need to be perfectly accurate and is used to