norohind/vanilla
1
0
Fork 0
Code Issues 1 Pull Requests Projects Releases Wiki Activity

use MetadataRetriever instead of BASTP

Browse Source
This commit is contained in:
Adrian Ulrich 2014-10-13 09:02:27 +02:00
parent e3a423e6c4
commit 9993ff64c6
1 changed files with 30 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

52
src/ch/blinkenlights/android/vanilla/MediaUtils.java
View File

@ -39,8 +39,8 @@ import android.net.Uri;
import android.os.Environment;
import android.provider.MediaStore;
import android.database.MatrixCursor;
import ch.blinkenlights.bastp.Bastp;
import android.media.MediaMetadataRetriever;
import android.util.Log;
/**
@ -563,32 +563,40 @@ public class MediaUtils {
* */
public static Cursor getCursorForFileQuery(String path) {
MatrixCursor matrixCursor = new MatrixCursor(Song.FILLED_PROJECTION);
String[] keys = { "", "", "TITLE", "ALBUM", "ARTIST" };
HashMap tags = (new Bastp()).getTags(path);
MediaMetadataRetriever data = new MediaMetadataRetriever();
if(tags.containsKey("type")) { // File was parseable
/* This file is not in the media database but we still need
* to give it an unique id. We are going to use a negative
* ID calculated using the file path.
* This is not perfect as the same file may be accessed using
* multiple path variations - but getting the inode is not possible
* and hashing the file is out of question */
try {
data.setDataSource(path);
} catch (Exception e) {
Log.w("VanillaMusic", "Failed to extract metadata from " + path);
}
String title = data.extractMetadata(MediaMetadataRetriever.METADATA_KEY_TITLE);
String album = data.extractMetadata(MediaMetadataRetriever.METADATA_KEY_ALBUM);
String artist = data.extractMetadata(MediaMetadataRetriever.METADATA_KEY_ARTIST);
String duration = data.extractMetadata(MediaMetadataRetriever.METADATA_KEY_DURATION);
if (duration != null) { // looks like we will be able to play this file
// Vanilla requires each file to be identified by its unique id in the media database.
// However: This file is not in the database, so we are going to roll our own
// using the negative crc32 sum of the path value. While this is not perfect
// (the same file may be accessed using various paths) it's the fastest method
// and far good enough.
CRC32 crc = new CRC32();
crc.update(path.getBytes());
Long songId = (Long)(2+crc.getValue())*-1; // must at least be -2 (-1 defines Song-Object to be empty)
Object[] objData = new Object[] { songId, path, "", "", "", 0, 0, 600000, 0 };
for (int i=0; i<keys.length; i++) { // fill in values from BASTP
if (tags.containsKey(keys[i])) {
objData[i] = (String)((Vector)tags.get(keys[i])).get(0);
}
}
// Build minimal fake-database entry for this file
Object[] objData = new Object[] { songId, path, "", "", "", 0, 0, 0, 0 };
// Overwrite hardcoded failback duration with actual data from tagparser
if(tags.containsKey("duration")) {
long songDuration = 1000L * (Integer)tags.get("duration");
objData[7] = songDuration;
}
if (title != null)
objData[2] = title;
if (album != null)
objData[3] = album;
if (artist != null)
objData[4] = artist;
if (duration != null)
objData[7] = Long.parseLong(duration, 10);
matrixCursor.addRow(objData);
}