curds: Add a test for parallel track scanning

And add code for generating a large test library to make stress testing
possible.

Signed-off-by: Anna Schumaker <Anna@NoWheyCreamery.com>

Makefile

@@ -3,6 +3,7 @@
 clean:
 	find . -type d -name __pycache__ -exec rm -r {} \+
 	find trier -type d -name "Test Album" -exec rm -r {} \+
+	find trier -type d -name "Test Library" -exec rm -r {} \+
 
 .PHONY: trier
 trier:

curds/test_track.py

@@ -1,10 +1,13 @@
 # Copyright 2019 (c) Anna Schumaker.
+import asyncio
+import concurrent.futures
 import mutagen
 import unittest
 import os
 import track
 
-test_tracks = os.path.abspath("./trier/Test Album")
+test_tracks  = os.path.abspath("./trier/Test Album")
+test_library = os.path.abspath("./trier/Test Library")
 
 class TestTrackClass(unittest.TestCase):
     def test_init_basic(self):
@@ -50,6 +53,7 @@ class TestTrackClass(unittest.TestCase):
         self.assertEqual(track.Track(join(test_tracks, "10 - Test {Disc 20}.ogg")).discnumber,   20)
 
     def test_track_add(self):
+        track.tracklist.clear()
         t = track.add(os.path.join(test_tracks, "01 - Test Track.ogg"))
         self.assertIsNotNone(t)
         self.assertEqual(len(track.tracklist), 1)
@@ -65,5 +69,19 @@ class TestTrackClass(unittest.TestCase):
         self.assertEqual(len(track.tracklist), 2)
         self.assertIn(u, track.tracklist)
 
+    def test_parallel_add(self):
+        track.tracklist.clear()
+        count  = 0
+        tracks = [ ]
+        with concurrent.futures.ThreadPoolExecutor(max_workers=5) as pool:
+            for dirname, subdirs, files in os.walk(test_library):
+                for f in files:
+                    tracks.append(pool.submit(track.add, os.path.join(dirname, f)))
+                    count += 1
+        self.assertEqual(len(track.tracklist), count)
+        self.assertEqual(len(tracks), count)
+        for t in tracks:
+            self.assertIsNotNone(t)
+
 if __name__ == '__main--':
         unittest.main()

trier/generate_tracks.py

@@ -8,6 +8,8 @@ ffmpeg = "ffmpeg -hide_banner -nostdin -f s16le -i /dev/zero -codec libvorbis -l
 
 def generate_track(length, filename, tags={}):
     path = os.path.join(trier, filename)
+    if os.path.exists(path):
+        return
     os.makedirs(os.path.dirname(path), exist_ok=True)
     subprocess.run(ffmpeg + [ "-t", str(length), path ])
 
@@ -16,6 +18,7 @@ def generate_track(length, filename, tags={}):
         fileinfo[key] = value
     fileinfo.save()
 
+# Create a bunch of tracks in the Test Album directory
 generate_track( 0, "Test Album/00 - Empty Track.ogg")
 generate_track(10, "Test Album/01 - Test Track.ogg", { "title"  : "Test Track",
                                                        "artist" : "Test Artist",
@@ -43,3 +46,18 @@ generate_track(50, "Test Album/09 - Test {Disc 02}.ogg", { "Title" : "Test {Disc
                                                            "album" : "Test Album {Disc 02}" })
 generate_track(55, "Test Album/10 - Test {Disc 20}.ogg", { "Title" : "Test {Disc 20}",
                                                            "album" : "Test Album {Disc 20}" })
+
+# Create a giant library for testing
+for artistno in range(1, 26):
+    artist = f"Test Artist {artistno:02}"
+    for albumno in range(1, 6):
+        album = f"Test Album {albumno}"
+        for trackno in range(1, 11):
+            title = f"Test Track {trackno:02}"
+            generate_track(trackno, f"Test Library/{artist}/{album}/{trackno:02} - {title}.ogg",
+                                { "title"  : title,
+                                  "artist" : artist,
+                                  "album"  : album,
+                                  "genre"  : f"Test Genre {albumno}",
+                                  "date"   : str(1970 + (albumno * 3)),
+                                  "tracknumber" : f"{trackno:02}" })