summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorDiego Elio Pettenò <flameeyes@google.com>2019-01-21 14:00:09 +0000
committerDiego Elio Pettenò <flameeyes@google.com>2019-01-21 14:00:09 +0000
commit8b3926ba609fce70cd94719b68590d73fdf876b7 (patch)
tree2f5dd6ff9f3102b786d63cbd10ff34e16c6fcdd1
parent42aad6110a02a86a9a51a77c75c2d638a50656b7 (diff)
downloadblinker-8b3926ba609fce70cd94719b68590d73fdf876b7.tar.gz
Fix typo to maintain consistency.
The word is "bookkeeping" (double-k in the middle), and it's already in use in the function name and part of the description.
-rw-r--r--blinker/base.py8
1 files changed, 4 insertions, 4 deletions
diff --git a/blinker/base.py b/blinker/base.py
index 3fd77d8..c9f65de 100644
--- a/blinker/base.py
+++ b/blinker/base.py
@@ -354,9 +354,9 @@ class Signal(object):
self._by_receiver[receiver_id].discard(sender_id)
def _cleanup_bookkeeping(self):
- """Prune unused sender/receiver bookeeping. Not threadsafe.
+ """Prune unused sender/receiver bookkeeping. Not threadsafe.
- Connecting & disconnecting leave behind a small amount of bookeeping
+ Connecting & disconnecting leave behind a small amount of bookkeeping
for the receiver and sender values. Typical workloads using Blinker,
for example in most web apps, Flask, CLI scripts, etc., are not
adversely affected by this bookkeeping.
@@ -364,10 +364,10 @@ class Signal(object):
With a long-running Python process performing dynamic signal routing
with high volume- e.g. connecting to function closures, "senders" are
all unique object instances, and doing all of this over and over- you
- may see memory usage will grow due to extraneous bookeeping. (An empty
+ may see memory usage will grow due to extraneous bookkeeping. (An empty
set() for each stale sender/receiver pair.)
- This method will prune that bookeeping away, with the caveat that such
+ This method will prune that bookkeeping away, with the caveat that such
pruning is not threadsafe. The risk is that cleanup of a fully
disconnected receiver/sender pair occurs while another thread is
connecting that same pair. If you are in the highly dynamic, unique