Mirrors/rspamd
1
0
Fork 0
mirror of https://github.com/rspamd/rspamd.git synced 2024-09-13 23:56:50 +02:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

[Minor] Allow to filter urls

Browse Source
This commit is contained in:
Vsevolod Stakhov 2018-05-12 14:22:33 +01:00
parent d003c95fcb
commit e39900d3ee
1 changed files with 12 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
lualib/lua_util.lua
View File

@ -476,14 +476,21 @@ end
exports.override_defaults = override_defaults
--[[[
-- @function lua_util.extract_specific_urls(task, limit)
-- @function lua_util.extract_specific_urls(task, limit, [need_emails[, filter[, prefix])
-- Apply heuristic in extracting of urls from task, this function
-- tries its best to extract specific number of urls from a task based on
-- their characteristics
--]]
exports.extract_specific_urls = function(task, lim, need_emails)
exports.extract_specific_urls = function(task, lim, need_emails, filter, prefix)
local fun = require "fun"
local cache_key = string.format('sp_urls_%d%s', lim, need_emails)
local cache_key
if prefix then
cache_key = prefix
else
cache_key = string.format('sp_urls_%d%s', lim, need_emails)
end
local cached = task:cache_get(cache_key)
@ -493,9 +500,10 @@ exports.extract_specific_urls = function(task, lim, need_emails)
local urls = task:get_urls(need_emails)
if not urls then return {} end
if filter then urls = fun.totable(fun.filter(filter, urls)) end
if #urls <= lim then
task:cache_set(cache_key, urls)