reddit: cache user profile API fetches for 5m to avoid repeating them

snarfed · snarfed · commit 85ccb5a6ef3e · 2021-04-01T07:27:07.000-07:00
...when fetching multiiple comments or posts from the same author. snarfed/bridgy#1021
diff --git a/README.md b/README.md
@@ -338,7 +338,9 @@ Changelog
   * Bug fix: URL-encode list names in API calls.
   * Bug fix: propagate alt text into AS1 `photo.displayName` so that it gets all the way into microformats2 JSON and HTML ([#183](https://github.com/snarfed/granary/issues/183)).
 * Reddit:
-  * Fix `post_id()`.
+  * Implement `post_id()`.
+  * Cache user data fetched from the API for 5m to avoid repeating user profile API requests ([bridgy#1021](https://github.com/snarfed/bridgy/issues/1021)).
+    when fetching multiple comments or posts from the same author
   * Bug fix: use 'displayName' instead of 'name' in AS1 objects for submissions.
   * Bug fix: use tag URIs for activity ids.
 * ActivityStreams 2:
diff --git a/granary/reddit.py b/granary/reddit.py
@@ -9,17 +9,24 @@
 PRAW API docs:
 https://praw.readthedocs.io/
 """
-
-from . import source
 import logging
-from oauth_dropins import reddit
-from oauth_dropins.webutil import appengine_info, util
+import operator
 import re
 import urllib.parse, urllib.request
+import threading
 
+from cachetools import cachedmethod, TTLCache
+from oauth_dropins import reddit
+from oauth_dropins.webutil import appengine_info, util
 import praw
 from prawcore.exceptions import NotFound
 
+from . import source
+
+USER_CACHE_TIME = 5 * 60  # 5 minute expiration, in seconds
+user_cache = TTLCache(1000, USER_CACHE_TIME)
+user_cache_lock = threading.RLock()
+
 
 class Reddit(source.Source):
   """Reddit source class. See file docstring and Source class for details."""
@@ -57,13 +64,23 @@ def post_id(self, url):
     if len(path_parts) >= 2:
       return path_parts[-2]
 
+  @cachedmethod(lambda self: user_cache, lock=lambda self: user_cache_lock,
+                key=lambda user: getattr(user, 'name', None))
   def praw_to_actor(self, praw_user):
     """Converts a PRAW Redditor to an actor.
 
     Makes external calls to fetch data from the Reddit API.
 
     https://praw.readthedocs.io/en/latest/code_overview/models/redditor.html
 
+    Caches fetched user data for 5m to avoid repeating user profile API requests
+    when fetching multiple comments or posts from the same author. Background:
+    https://github.com/snarfed/bridgy/issues/1021
+
+    Ideally this would be part of PRAW, but they seem uninterested:
+    https://github.com/praw-dev/praw/issues/131
+    https://github.com/praw-dev/praw/issues/1140
+
     Args:
       user: PRAW Redditor object
 
diff --git a/granary/tests/test_reddit.py b/granary/tests/test_reddit.py
@@ -175,6 +175,8 @@ def setUp(self):
     self.submission_link.selftext = ''
     self.submission_link.url = 'https://reddit.com/ezv3f2'
 
+    reddit.user_cache.clear()
+
   def test_user_url(self):
     self.assert_equals('https://reddit.com/user/foo', self.reddit.user_url('foo'))
 
