responding to @bfalling CR, comments

openlibrary/core/cache.py

@@ -25,7 +25,7 @@
 HOUR = 60 * 60
 DAY = HOUR * 24
 HALF_DAY = HOUR * 12
-ONE_WEEK = HALF_DAY * 2 * 7
+ONE_WEEK = DAY * 7
 
 
 class memcache_memoize:

openlibrary/plugins/openlibrary/code.py

@@ -772,6 +772,13 @@ def GET(self):
         return result
 
 def is_bot():
+    """Generated on ol-www1 within /var/log/nginx with:
+
+    cat access.log | grep -oh "; \w*[bB]ot" | sort --unique | awk '{print tolower($2)}'
+    cat access.log | grep -oh "; \w*[sS]pider" | sort --unique | awk '{print tolower($2)}'
+
+    Manually removed singleton `bot` (to avoid overly complex grep regex)
+    """
     user_agent_bots = [
         'sputnikbot', 'dotbot', 'semrushbot',
         'googlebot', 'yandexbot', 'monsidobot', 'kazbtbot',

openlibrary/plugins/upstream/code.py

@@ -132,16 +132,21 @@ def cached_get_amazon_metadata(*args, **kwargs):
     """If the cached data is `None`, likely a 503 throttling occurred on
     Amazon's side. Try again to fetch the value instead of using the
     cached value. It may 503 again, in which case the next access of
-    this page will trigger another re-cache. If the book has no price
-    data, then {"price": None} will be cached will will not trigger a
-    re-cache (only the value `None` will)
-
+    this page will trigger another re-cache. If the amazon API call
+    succeeds but the book has no price data, then {"price": None} will
+    be cached as to not trigger a re-cache (only the value `None`
+    will cause re-cache)
     """
+    # fetch/compose a cache controller obj for
+    # "upstream.code._get_amazon_metadata"
     memoized_get_amazon_metadata = cache.memcache_memoize(
         _get_amazon_metadata, "upstream.code._get_amazon_metadata",
         timeout=cache.ONE_WEEK)
+    # fetch cached value from this controller
     result = memoized_get_amazon_metadata(*args, **kwargs)
     if result is None:
+        # recache / update this controller's cached value
+        # (corresponding to these input args)
         result = memoized_get_amazon_metadata.update(*args, **kwargs)[0]
     return result