Dive pictures: find moved pictures based on filename

Users might have edited their pictures. Therefore, instead of identifying
pictures by the hash of the file-content, use the file path. The match
between original and new filename is graded by a score. Currently, this
is the number of path components that match, starting from the filename.
Camparison is case-insensitive.

After having identified the matching images, write the caches so that they
are saved even if the user doesn't cleanly quit the application.

Since the new code uses significantly less resources, it can be run in a
single background thread. Thus, the multi-threading can be simplified.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>

core/qthelper.cpp

@@ -1277,37 +1277,95 @@ QStringList imageExtensionFilters() {
 	return filters;
 }
 
-// This works on a copy of the string, because it runs in asynchronous context
+// Compare two full paths and return the number of matching levels, starting from the filename.
-static void learnImage(QString filename)
+// String comparison is case-insensitive.
+static int matchFilename(const QString &path1, const QString &path2)
 {
+	QFileInfo f1(path1);
+	QFileInfo f2(path2);
+
+	int score = 0;
+	for (;;) {
+		QString fn1 = f1.fileName();
+		QString fn2 = f2.fileName();
+		if (fn1.isEmpty() || fn2.isEmpty())
+			break;
+		if (fn1 == ".") {
+			f1 = QFileInfo(f1.path());
+			continue;
+		}
+		if (fn2 == ".") {
+			f2 = QFileInfo(f2.path());
+			continue;
+		}
+		if (QString::compare(fn1, fn2, Qt::CaseInsensitive) != 0)
+			break;
+		f1 = QFileInfo(f1.path());
+		f2 = QFileInfo(f2.path());
+		++score;
+	}
+	return score;
+}
+
+struct ImageMatch {
+	QString localFilename;
+	int score;
+};
+
+static void learnImage(const QString &filename, QMap<QString, ImageMatch> &matches)
+{
+	// Find the original filenames with the highest match-score
+	QStringList newMatches;
 	QByteArray hash = hashFile(filename);
-	// TODO: This is inefficient: we search the hash map by value. But firstly,
+	int bestScore = 1;
-	// this is running in asynchronously, so it doesn't block the UI. Secondly,
-	// we might not want to learn pictures by hash anyway (the user might have
-	// edited the picture, which changes the hash.
 	for (auto it = hashOf.cbegin(); it != hashOf.cend(); ++it) {
-		if (it.value() == hash)
+		int score = matchFilename(filename, it.key());
-			learnPictureFilename(it.key(), filename);
+		if (score < bestScore)
+			continue;
+		if (score > bestScore)
+			newMatches.clear();
+		newMatches.append(it.key());
+		bestScore = score;
+	}
+
+	// Add the new original filenames to the list of matches, if the score is higher than previously
+	for (const QString &originalFilename: newMatches) {
+		auto it = matches.find(originalFilename);
+		if (it == matches.end())
+			matches.insert(originalFilename, { filename, bestScore });
+		else if (it->score < bestScore)
+			*it = { filename, bestScore };
 	}
 }
 
-void learnImages(const QDir dir, int max_recursions)
+void learnImages(const QStringList &dirNames, int max_recursions)
 {
-	QStringList files;
 	QStringList filters = imageExtensionFilters();
+	QMap<QString, ImageMatch> matches;
 
-	if (max_recursions) {
+	QVector<QStringList> stack; // Use a stack to recurse into directories
-		foreach (QString dirname, dir.entryList(QStringList(), QDir::NoDotAndDotDot | QDir::Dirs)) {
+	stack.reserve(max_recursions + 1);
-			learnImages(QDir(dir.filePath(dirname)), max_recursions - 1);
+	stack.append(dirNames);
+	while (!stack.isEmpty()) {
+		if (stack.last().isEmpty()) {
+			stack.removeLast();
+			continue;
+		}
+		QDir dir(stack.last().takeLast());
+
+		for (const QString &file: dir.entryList(filters, QDir::Files))
+			learnImage(dir.absoluteFilePath(file), matches);
+		if (stack.size() <= max_recursions) {
+			stack.append(QStringList());
+			for (const QString &dirname: dir.entryList(QStringList(), QDir::NoDotAndDotDot | QDir::Dirs))
+				stack.last().append(dir.filePath(dirname));
 		}
 	}
 
+	for (auto it = matches.begin(); it != matches.end(); ++it)
+		learnPictureFilename(it.key(), it->localFilename);
 
-	foreach (QString file, dir.entryList(filters, QDir::Files)) {
+	write_hashes();
-		files.append(dir.absoluteFilePath(file));
-	}
-
-	QtConcurrent::blockingMap(files, learnImage);
 }
 
 extern "C" const char *local_file_path(struct picture *picture)

core/qthelper.h

@@ -31,7 +31,7 @@ void updateHash(struct picture *picture);
 QByteArray hashFile(const QString &filename);
 QString hashString(const char *filename);
 QString thumbnailFileName(const QString &filename);
-void learnImages(const QDir dir, int max_recursions);
+void learnImages(const QStringList &dirNames, int max_recursions);
 void learnPictureFilename(const QString &originalName, const QString &localName);
 void hashPicture(QString filename);
 extern "C" char *hashstring(const char *filename);

desktop-widgets/mainwindow.cpp

@@ -701,13 +701,10 @@ void MainWindow::on_actionCloudOnline_triggered()
 	updateCloudOnlineStatus();
 }
 
-void learnImageDirs(QStringList dirnames)
+static void learnImageDirs(QStringList dirnames)
 {
-	QList<QFuture<void> > futures;
+	learnImages(dirnames, 10);
-	foreach (QString dir, dirnames) {
+	DivePictureModel::instance()->updateDivePictures();
-		futures << QtConcurrent::run(learnImages, QDir(dir), 10);
-	}
-	DivePictureModel::instance()->updateDivePicturesWhenDone(futures);
 }
 
 void MainWindow::on_actionHash_images_triggered()

qt-models/divepicturemodel.cpp

@@ -23,14 +23,6 @@ DivePictureModel::DivePictureModel() : rowDDStart(0),
 		this, &DivePictureModel::updateThumbnail, Qt::QueuedConnection);
 }
 
-void DivePictureModel::updateDivePicturesWhenDone(QList<QFuture<void>> futures)
-{
-	Q_FOREACH (QFuture<void> f, futures) {
-		f.waitForFinished();
-	}
-	updateDivePictures();
-}
-
 void DivePictureModel::setZoomLevel(int level)
 {
 	zoomLevel = level / 10.0;

qt-models/divepicturemodel.h

@@ -21,7 +21,6 @@ public:
 	virtual QVariant data(const QModelIndex &index, int role = Qt::DisplayRole) const;
 	virtual int rowCount(const QModelIndex &parent = QModelIndex()) const;
 	virtual void updateDivePictures();
-	void updateDivePicturesWhenDone(QList<QFuture<void>>);
 	void removePictures(const QVector<QString> &fileUrls);
 	int rowDDStart, rowDDEnd;
 	void updateDivePictureOffset(const QString &filename, int offsetSeconds);