mirror of
https://github.com/discourse/discourse.git
synced 2025-02-25 18:55:32 -06:00
FIX: use crawler layout when saving url in Wayback Machine (#7667)
This commit is contained in:
@@ -1,6 +1,7 @@
|
||||
# frozen_string_literal: true
|
||||
|
||||
module CrawlerDetection
|
||||
WAYBACK_MACHINE_URL = "web.archive.org"
|
||||
|
||||
def self.to_matcher(string, type: nil)
|
||||
escaped = string.split('|').map { |agent| Regexp.escape(agent) }.join('|')
|
||||
@@ -13,8 +14,8 @@ module CrawlerDetection
|
||||
Regexp.new(escaped, Regexp::IGNORECASE)
|
||||
end
|
||||
|
||||
def self.crawler?(user_agent)
|
||||
return true if user_agent.nil?
|
||||
def self.crawler?(user_agent, via_header = nil)
|
||||
return true if user_agent.nil? || via_header&.include?(WAYBACK_MACHINE_URL)
|
||||
|
||||
# this is done to avoid regenerating regexes
|
||||
@non_crawler_matchers ||= {}
|
||||
|
||||
Reference in New Issue
Block a user