From 306fa0f40c89fa64c7bb9821f86f96a572c6f4f4 Mon Sep 17 00:00:00 2001
From: Anna Schumaker <Anna@NoWheyCreamery.com>
Date: Mon, 28 Jan 2019 16:05:11 -0500
Subject: [PATCH] curds: Add function for album lookups

We want to share album objects whenever possible, so add a lookup
function for checking this.  Python built-in objects (like dictionaries)
are supposedly threadsafe, so we don't need a lock to access them but I
added the threadpool test anyway just in case.

Signed-off-by: Anna Schumaker <Anna@NoWheyCreamery.com>
---
 curds/album.py      | 15 +++++++++++++++
 curds/test_album.py | 23 +++++++++++++++++++++++
 2 files changed, 38 insertions(+)

diff --git a/curds/album.py b/curds/album.py
index 0f248bd..3b25749 100644
--- a/curds/album.py
+++ b/curds/album.py
@@ -1,4 +1,7 @@
 # Copyright 2019 (c) Anna Schumaker.
+import hashlib
+
+album_map = dict()
 
 class Album:
     def __init__(self, fileinfo):
@@ -9,3 +12,15 @@ class Album:
         self.albumartist = fileinfo.get("albumartist",
                                 fileinfo.get("album artist",
                                     fileinfo.get("artist", [ "Unknown Artist" ])))[0]
+
+    def hash(self):
+        md5 = hashlib.md5()
+        md5.update(self.album.encode('utf-8'))
+        md5.update(self.albumartist.encode('utf-8'))
+        md5.update(str(self.date).encode('utf-8'))
+        return md5.hexdigest()
+
+
+def lookup(fileinfo):
+    album = Album(fileinfo)
+    return album_map.setdefault(album.hash(), album)
diff --git a/curds/test_album.py b/curds/test_album.py
index d699083..155e4f8 100644
--- a/curds/test_album.py
+++ b/curds/test_album.py
@@ -1,4 +1,6 @@
 # Copyright 2019 (c) Anna Schumaker
+import hashlib
+from multiprocessing.pool import ThreadPool
 import unittest
 import album
 
@@ -13,6 +15,7 @@ class TestAlbumClass(unittest.TestCase):
         self.assertEqual(a.date,   2019)
         self.assertEqual(a.albumartist, "Test Artist")
         self.assertEqual(a.tracktotal,   1)
+        self.assertEqual(a.hash(), hashlib.md5("Test AlbumTest Artist2019".encode('utf-8')).hexdigest())
 
     def test_init_empty(self):
         a = album.Album({})
@@ -21,6 +24,7 @@ class TestAlbumClass(unittest.TestCase):
         self.assertEqual(a.date,   0)
         self.assertEqual(a.albumartist, "Unknown Artist")
         self.assertEqual(a.tracktotal,   0)
+        self.assertEqual(a.hash(), hashlib.md5("Unknown AlbumUnknown Artist0".encode('utf-8')).hexdigest())
 
     def test_init_artist_fallback(self):
         test_info = {"albumartist" : [ "1" ], "album artist" : [ "2" ], "artist" : [ "3" ]}
@@ -32,5 +36,24 @@ class TestAlbumClass(unittest.TestCase):
         test_info.pop("artist")
         self.assertEqual(album.Album(test_info).albumartist, "Unknown Artist")
 
+    def test_album_lookup(self):
+        album.album_map.clear()
+        a = album.lookup(album_info)
+        self.assertIsNotNone(a)
+
+        for i in range(10):
+            b = album.lookup(album_info)
+            self.assertEqual(a, b)
+        self.assertEqual(len(album.album_map), 1)
+
+    def test_parallel_lookup(self):
+        album.album_map.clear()
+        with ThreadPool(processes=5) as pool:
+            res = pool.map(album.lookup, [ album_info ] * 20)
+            self.assertIsNotNone(res[0])
+            self.assertEqual(res.count(res[0]), 20)
+        self.assertEqual(len(album.album_map), 1)
+
+
 if __name__ == '__main__':
     unittest.main()