Remove FilesPreprocessor and FilesPreprocessor to introduce SequenceExtractor

Author: chesnokoff

modules/core/src/main/java/jsymbolic2/processing/FeatureExtractionJobProcessor.java

@@ -4,6 +4,7 @@
 import jsymbolic2.configuration.ConfigurationFileData;
 import jsymbolic2.featureutils.FeatureExtractorAccess;
 import org.apache.commons.lang3.tuple.ImmutablePair;
+import org.apache.commons.lang3.tuple.Pair;
 import org.ddmal.jmei2midi.MeiSequence;
 import org.ddmal.jmei2midi.meielements.meispecific.MeiSpecificStorage;
 
@@ -19,8 +20,7 @@
  *
  * @author Cory McKay and Tristano Tenaglia
  */
-public enum FeatureExtractionJobProcessor {
-    ;
+public class FeatureExtractionJobProcessor {
     /* PUBLIC STATIC METHODS ********************************************************************************/
 
     /**
@@ -66,11 +66,11 @@ public static List<String> extractAndSaveSpecificFeatures(List<File> paths_of_fi
                     FeatureExtractorAccess.getAllImplementedFeatureExtractors(),
                     features_to_extract,
                     saveInfo.save_overall_recording_features());
-            FilesPreprocessor filesPreprocessor = new FilesPreprocessor(paths_of_files_or_folders_to_parse,
-                    printStreams.error_print_stream(), error_log);
-            FilesReader filesReader = new FilesReader(List.of(new SequencePreprocessor()));
-            List<ImmutablePair<String, Sequence>> midiPairs = filesReader.extractMidi(filesPreprocessor.getMidiFilesList());
-            List<ImmutablePair<String, MeiSequence>> meiPairs = filesReader.extractMei(filesPreprocessor.getMeiFilesList());
+
+            SequenceExtractor sequenceExtractor = new SequenceExtractor(paths_of_files_or_folders_to_parse);
+
+            List<Pair<String, Sequence>> midiPairs = sequenceExtractor.getMIDISequences();
+            List<Pair<String, MeiSequence>> meiPairs = sequenceExtractor.getMEISequences();
             // Extract features and save the feature values in DataBoard
             DataBoard dataBoard = FeatureExtractionJobProcessor.extractFeatures(midiPairs, meiPairs,
                     processor,
@@ -206,8 +206,8 @@ public static List<String> extractAndSaveFeaturesConfigFileSettings(List<File> p
      *                                 displayed and a direct printing	of the associated error message to
      *                                 standard error.
      */
-    private static DataBoard extractFeatures(List<ImmutablePair<String, Sequence>> midiSequences,
-                                             List<ImmutablePair<String, MeiSequence>> meiSequences,
+    private static DataBoard extractFeatures(List<Pair<String, Sequence>> midiSequences,
+                                             List<Pair<String, MeiSequence>> meiSequences,
                                              MIDIFeatureProcessor processor,
                                              String feature_values_save_path,
                                              PrintStreams printStreams, List<String> error_log,

modules/core/src/main/java/jsymbolic2/processing/FilesPreprocessor.java

@@ -0,0 +1,62 @@
+package jsymbolic2.processing;
+
+import java.io.File;
+import java.io.IOException;
+import java.util.List;
+import javax.sound.midi.InvalidMidiDataException;
+import javax.sound.midi.MidiSystem;
+import javax.sound.midi.Sequence;
+import org.apache.commons.lang3.tuple.Pair;
+import org.ddmal.jmei2midi.MeiSequence;
+
+public class SequenceExtractor {
+
+  private final List<File> filesList;
+  SequencePreprocessor sequencePreprocessor = new SequencePreprocessor();
+
+  SequenceExtractor(List<File> filesList) {
+    this.filesList = filesList;
+  }
+
+  List<Pair<String, Sequence>> getMIDISequences() {
+    return filesList.stream()
+        .filter(file -> file.getName().endsWith(".mid") || file.getName().endsWith(".midi"))
+        .map(file -> {
+          try {
+            return Pair.of(file.getName(),
+                sequencePreprocessor.apply(MidiSystem.getSequence(file)));
+          } catch (InvalidMidiDataException | IOException e) {
+            throw new RuntimeException("Could not parse " + file.getAbsolutePath() + " " + e);
+          }
+        })
+        .toList();
+  }
+
+  List<Pair<String, MeiSequence>> getMEISequences() {
+    return filesList.stream()
+        .filter(file -> file.getName().endsWith(".mei"))
+        .map(file -> {
+          try {
+            return Pair.of(file.getName(), new MeiSequence(file));
+          } catch (InvalidMidiDataException e) {
+            throw new RuntimeException("Could not parse " + file.getAbsolutePath() + " " + e);
+          }
+        })
+        .toList();
+  }
+
+  public static Sequence getMidiSequenceFromMidiFile(File file, List<String> error_log)
+      throws IOException, InvalidMidiDataException {
+    Sequence sequence = null;
+    try {
+      sequence = MidiSystem.getSequence(file);
+    } catch (IOException e) {
+      error_log.add("The specified path, " + file + ", does not refer to a valid file.");
+      throw e;
+    } catch (InvalidMidiDataException e) {
+      error_log.add("The specified file, " + file + ", is not a valid MIDI or MEI file.");
+      throw e;
+    }
+    return sequence;
+  }
+}