From df5ac352f916c1118fba0ddd61cf6cea2b4a1b70 Mon Sep 17 00:00:00 2001
From: Sergey Fedotov <sergey89@gmail.com>
Date: Thu, 23 Feb 2017 14:32:42 +0300
Subject: [PATCH] Allow escape sequences in Apache access log

---
 lib/fluent/plugin/parser_apache2.rb | 2 +-
 test/plugin/test_parser_apache2.rb  | 8 ++++++++
 2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/lib/fluent/plugin/parser_apache2.rb b/lib/fluent/plugin/parser_apache2.rb
index fd1a69e9b6..8c18c56bc7 100644
--- a/lib/fluent/plugin/parser_apache2.rb
+++ b/lib/fluent/plugin/parser_apache2.rb
@@ -21,7 +21,7 @@ module Plugin
     class Apache2Parser < Parser
       Plugin.register_parser('apache2', self)
 
-      REGEXP = /^(?<host>[^ ]*) [^ ]* (?<user>[^ ]*) \[(?<time>[^\]]*)\] "(?<method>\S+)(?: +(?<path>[^\"]*?)(?: +\S*)?)?" (?<code>[^ ]*) (?<size>[^ ]*)(?: "(?<referer>[^\"]*)" "(?<agent>[^\"]*)")?$/
+      REGEXP = /^(?<host>[^ ]*) [^ ]* (?<user>[^ ]*) \[(?<time>[^\]]*)\] "(?<method>\S+)(?: +(?<path>(?:[^\"]|\\.)*?)(?: +\S*)?)?" (?<code>[^ ]*) (?<size>[^ ]*)(?: "(?<referer>(?:[^\"]|\\.)*)" "(?<agent>(?:[^\"]|\\.)*)")?$/
       TIME_FORMAT = "%d/%b/%Y:%H:%M:%S %z"
 
       def initialize
diff --git a/test/plugin/test_parser_apache2.rb b/test/plugin/test_parser_apache2.rb
index 6513775001..f2e30a73fd 100644
--- a/test/plugin/test_parser_apache2.rb
+++ b/test/plugin/test_parser_apache2.rb
@@ -35,4 +35,12 @@ def test_parse_without_http_version
       assert_equal(@expected, record)
     }
   end
+
+  def test_parse_with_escape_sequence
+    @parser.instance.parse('192.168.0.1 - - [28/Feb/2013:12:00:00 +0900] "GET /\" HTTP/1.1" 200 777 "referer \\\ \"" "user agent \\\ \""') { |_, record|
+      assert_equal('/\"', record['path'])
+      assert_equal('referer \\\ \"', record['referer'])
+      assert_equal('user agent \\\ \"', record['agent'])
+    }
+  end
 end