aboutsummaryrefslogtreecommitdiffstats
path: root/lualib/redis_scripts
Commit message (Collapse)AuthorAgeFilesLines
* [Minor] Make names more informativeLeftTry2024-09-291-4/+4
|
* [Minor] Change wrong logic of LRU cache to the right oneLeftTry2024-09-291-1/+1
|
* [Feature] Add LRU cache for last filled ratelimit bucketsLeftTry2024-09-291-0/+5
|
* [Project] Enable compatibility with the existing bucketsVsevolod Stakhov2024-07-092-40/+56
|
* [Fix] Increase/decrease hash_key value for tokens depending on is_unlearn aduernberger2024-04-181-2/+2
| | | | | | | | When relearning a message, this script is called twice: 1. With `is_unlearn` true and the old `is_spam` value 2. With `is_unlearn` false and the new `is_spam` value If `is_unlearn` is true, the code should not increase the value of the hash_key. Decreasing it ensures that it behaves like a fresh learn with the new is_spam value.
* [Fix] Update is_spam in learned_ids when relearningaduernberger2024-04-181-2/+3
|
* [Fix] Change expiration logic for redis_cacheVsevolod Stakhov2024-03-091-1/+1
| | | | Noticed by: @jeremy
* [Fix] Fix storing of the bayes tokensVsevolod Stakhov2024-01-221-1/+1
|
* [Minor] Remove special handlingVsevolod Stakhov2024-01-171-3/+1
|
* [Project] Some more work with regard to configurationVsevolod Stakhov2024-01-172-9/+19
|
* [Project] Final things on redis cache reworkVsevolod Stakhov2024-01-172-0/+73
|
* [Project] Initial implementation of the lua counterpartVsevolod Stakhov2024-01-122-0/+0
|
* [Project] Implement text tokens storage on C++ sideVsevolod Stakhov2023-12-311-6/+8
|
* [Project] Add store tokens support in new bayes learnVsevolod Stakhov2023-12-311-1/+19
|
* [Project] Try to do strings concatenation in CVsevolod Stakhov2023-12-142-5/+3
|
* [Project] Add some statsVsevolod Stakhov2023-12-081-0/+19
|
* [Project] Various fixesVsevolod Stakhov2023-12-081-2/+2
|
* [Rework] Use strings for int64_tVsevolod Stakhov2023-12-083-5/+5
| | | | | | It seems there is no easy way to use int64 in Redis Lua, hence, we have to use strings. It's much more expensive but still some advantage over the previous schema.
* [Minor] Fix keys nameVsevolod Stakhov2023-12-081-2/+2
|
* [Project] Add some basic learningVsevolod Stakhov2023-12-071-2/+2
|
* [Project] Add bayes learn scriptVsevolod Stakhov2023-12-071-0/+25
|
* [Project] Optimise classify scriptVsevolod Stakhov2023-12-021-13/+25
|
* [Project] Add classify redis scriptVsevolod Stakhov2023-12-021-0/+26
|
* [Minor] Reformat all Lua code, no functional changesVsevolod Stakhov2023-08-076-22/+48
|
* [Minor] Minor fixes for previous ratelimit changesAndrew Lewis2023-07-241-1/+1
|
* [Fix] Try harder to clean pending bucketVsevolod Stakhov2023-06-171-0/+29
| | | | Issue: #4467
* fix: avoid double-dip of nrcpt when calculating ratelimit conditionAndreas Simbuerger2023-03-291-1/+1
| | | | | Before this commit, rspamd would consider the number of recipients of the current message twice when calculating the ratelimit condition.
* [Minor] Neural: Extract lua scriptsVsevolod Stakhov2023-03-254-0/+88
|
* [Minor] Add some more comments as we now strip themVsevolod Stakhov2023-03-252-29/+42
|
* [Minor] Improve style in the scriptsVsevolod Stakhov2023-03-252-40/+45
|
* [Minor] Move redis scripts from ratelimit file to simplify checksVsevolod Stakhov2023-03-252-0/+142
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-31 23:33:29 +0000