From d2ed22d22972dc68a1d9962ff53e42d818ba4302 Mon Sep 17 00:00:00 2001
From: Jonathan Bernard <jdbernard@gmail.com>
Date: Wed, 16 Mar 2016 05:49:29 -0500
Subject: [PATCH] Rescanning now finds moved files by their hash.

More precisely, the MediaLibrary matches up files based on their file hash. If
it finds a files that matches an already known entry it doesn't create a new
entry, only updates the file path to the found file in the existing entry.

This is useful if files get re-organized or if an existing database is loaded
at an entirely new library root that contains mostly the same files (relocating
the library to another computer for example). This method preserves information
added to the database that was not originally present in the ID3 tags in the
media files (custom tags, name corrections, etc).
---
 .../com/jdbernard/wdiwtlt/MediaLibrary.groovy | 28 +++++++++++++------
 .../com/jdbernard/wdiwtlt/db/DbApi.groovy     |  4 +++
 2 files changed, 23 insertions(+), 9 deletions(-)

diff --git a/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy b/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy
index 828664f..599d4b5 100644
--- a/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy
+++ b/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy
@@ -70,17 +70,31 @@ public class MediaLibrary {
             return null }
 
         def relPath = getRelativePath(libraryRoot, f)
-        MediaFile mf = dbapi.getMediaFileByFilePath(relPath)
+        MediaFile found = dbapi.getMediaFileByFilePath(relPath)
 
-        if (mf) {
+        if (found) {
             logger.info(
                 "Ignoring a media file I already know about: {}", relPath)
-            return mf }
+            return found }
+
+        MediaFile mf = new MediaFile()
+        mf.filePath = relPath
+
+        // Hash the file
+        mf.fileHash = f.withInputStream { DigestUtils.md5Hex(it) }
+
+        // Look for an entry for that hash
+        found = dbapi.getMediaFileByFileHash(mf.fileHash)
+
+        if (found) {
+            logger.info('Found a media file by hash in a new location. ' +
+                "I'm updating my relative path to this file:\n\t{}\n\t\t--> {}.",
+                found.filePath, mf.filePath)
+            found.filePath = mf.filePath
+            return dbapi.update(found) }
 
         // Read in the media's tags
-        mf = new MediaFile()
         def af
-
         try { af = AudioFileIO.read(f) }
         catch (Exception e) {
             logger.info("Ignoring a file because I can't " +
@@ -91,7 +105,6 @@ public class MediaLibrary {
         def fileTag = af.tag
 
         mf.name = fileTag?.getFirst(TITLE)?.trim() ?: f.name
-        mf.filePath = relPath
         mf.comment = fileTag?.getAll(COMMENT)?.collect { it.trim() }?.join('\n\n')
         mf.discNumber = safeToInteger(fileTag?.getFirst(DISC_NO)) ?: 1
         mf.trackNumber = safeToInteger(fileTag?.getFirst(TRACK))
@@ -110,9 +123,6 @@ public class MediaLibrary {
             mf.metaInfoSource = MediaFile.FILE_LOCATION
             albumNames = [folderParts.peekLast()] }
 
-        // Hash the file
-        mf.fileHash = f.withInputStream { DigestUtils.md5Hex(it) }
-
         dbapi.withTransaction {
             dbapi.create(mf)
             associateWithArtistsAndAlbums(mf, artistNames, albumNames,
diff --git a/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy b/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy
index 6b8dd8f..02a0126 100644
--- a/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy
+++ b/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy
@@ -411,6 +411,10 @@ public class DbApi {
         def files = getBy(MediaFile, ['file_path'], [filePath])
         return files ? files[0] : null }
 
+    public MediaFile getMediaFileByFileHash(String fileHash) {
+      def files = getBy(MediaFile, ['file_hash'], [fileHash])
+      return files ? files[0] : null }
+
     public List<MediaFile> getMediaFilesWhere(Map params) {
         def query = new StringBuilder()
         def sqlParams = []