From a59f2599bd55ed0e35a206e9bb09d0d6ff7414c6 Mon Sep 17 00:00:00 2001
From: Lukas Winkler <git@lw1.at>
Date: Sat, 4 Jul 2020 12:20:02 +0200
Subject: [PATCH] initial commit

---
 .gitignore      |  1 +
 README.md       | 26 ++++++++++++++++++
 spotifystats.py | 72 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 99 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 README.md
 create mode 100644 spotifystats.py

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..9f11b75
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+.idea/
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..e809073
--- /dev/null
+++ b/README.md
@@ -0,0 +1,26 @@
+# Spotify Stats
+
+a very simple python script to analyse hours listened by day and most played songs
+
+### How-To
+
+- Request your account data from https://www.spotify.com/us/account/privacy/
+- confirm the E-Mail
+- up to 30 days later you will get a zip file containing a `MyData` folder containing files similar to those:
+  - Follow.json
+  - Payments.json
+  - Playlist1.json
+  - Read_Me_First.pdf
+  - SearchQueries.json
+  - StreamingHistory0.json
+  - StreamingHistory1.json
+  - Userdata.json
+  - YourLibrary.json
+- download the `spotifystats.py` file
+- edit the `BASEDIR` path to your `MyData` folder
+- run `spotifystats.py`
+
+### Dependencies
+
+- python 3.7 or newer
+- matplotlib
diff --git a/spotifystats.py b/spotifystats.py
new file mode 100644
index 0000000..746c448
--- /dev/null
+++ b/spotifystats.py
@@ -0,0 +1,72 @@
+import json
+from collections import Counter
+from dataclasses import dataclass
+from datetime import datetime
+from pathlib import Path
+
+from matplotlib import pyplot as plt
+
+# minumum time a song needs to be played to count as "played"
+MIN_PLAY_TIME = 60  # seconds
+
+# only show songs that were played at least this often in the top list
+MIN_PLAY_NUM = 3
+
+# path to Spotify export
+BASEDIR = Path("/home/lukas/Nextcloud/MyData/")
+
+
+@dataclass
+class Song:
+    end_time: str
+    artist_name: str
+    track_name: str
+    ms_played: int
+
+    def datetime(self) -> datetime:
+        return datetime.strptime(self.end_time, "%Y-%m-%d %H:%M")
+
+    def date(self):
+        return self.datetime().date()
+
+    @property
+    def min_played(self) -> float:
+        return self.ms_played / 1000 / 60
+
+
+history = []
+
+for file in BASEDIR.glob("StreamingHistory*.json"):
+    with file.open() as f:
+        data = json.load(f)
+    for song in data:
+        try:
+            song_obj = Song(end_time=song["endTime"], artist_name=song["artistName"],
+                            track_name=song["trackName"], ms_played=song["msPlayed"])
+            history.append(song_obj)
+        except KeyError:
+            print("this entry doesn't seem to be a song")
+            continue
+
+print(f"{len(history)} plays found")
+
+bins = {}
+for song in history:
+    if song.date() in bins:
+        bins[song.date()] += song.min_played
+    else:
+        bins[song.date()] = song.min_played
+
+plt.bar(bins.keys(), bins.values())
+plt.ylabel("minutes")
+
+played_songs = []
+for song in history:
+    if song.ms_played > MIN_PLAY_TIME * 1000:
+        played_songs.append((song.artist_name, song.track_name))
+
+for (artist, title), num_played in sorted(Counter(played_songs).most_common(), key=lambda c: [-c[1], c[0][0], c[0][1]]):
+    if num_played >= MIN_PLAY_NUM:
+        print(f"{num_played:>4} {artist}: {title}")
+
+plt.show()