fluent-plugin-filter-list 0.4.3 → 0.5.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 61d9d3071404bc65e8aa1cb85c1375436f8fce088e54ed118e1b73698c0d73de
-  data.tar.gz: ba810440dc189c6ea8a0cc94142dc8b5f87636756ee80e24c57d560fdad5af08
+  metadata.gz: 17bee9e79c216bfbdf746e8d7d089d229b2997bccf376d24c5e7311f1931f2b4
+  data.tar.gz: b00d3238b9a8d775f5f1dc2e7fd03beedde536bc6894d38f22091395d3dde1bf
 SHA512:
-  metadata.gz: bbc45bdceb06bba6907dc229e139e72b5013bc20addd935f51f15e1975fa7e023f24a5edd4ead2ba1db0cee0adb87a16f25b456fcbbb97b3eaf8c5612cbbbc4c
-  data.tar.gz: 88014dbb02b3afc0d485d7f7a0a628d9e0fe3c7a3a146001a95c7535c957a3aeecaa3f42ddcb065aa615f429e6cb70a5814d88f1c9d132879aaeca83da8ad5a2
+  metadata.gz: c578ea80823b2c26e70a37cf63b3df24832bef79013fcd7e1c28cbf937d220130581bd2a382a6ef06b5c9d1b96838a12d154b1263512920214041cfdfe467ee5
+  data.tar.gz: 513142ca73ebf949f94d3f22fea2deff6e5147ffea92d65687ce33c80c6217e9d70d47dc3ec53b47858a5edb725c1b43ba70334e0feece51971978ca9fe886f4

data/.rubocop.yml

@@ -17,7 +17,7 @@ Style/FrozenStringLiteralComment:
 Style/MutableConstant:
   Enabled: false
 Metrics/AbcSize:
-  Max: 25
+  Max: 30
 Metrics/BlockLength:
   Exclude:
     - 'Rakefile'
@@ -29,9 +29,9 @@ Metrics/BlockLength:
     - 'config/routes/**/*.rb'
     - '*.gemspec'
 Metrics/CyclomaticComplexity:
-  Max: 10
+  Max: 15
 Metrics/PerceivedComplexity:
-  Max: 10
+  Max: 15
 Style/GlobalVars:
   Exclude:
     - 'test/test_helper.rb'

data/README.md

@@ -34,8 +34,9 @@ Use the `filter_list` filter. Configure fluentd as follows.
   @type filter_list
 
   filter AC
-  key_to_filter xyz
+  key_to_filter x
   patterns_file_path blacklist.txt
+  filter_empty true
 </filter>
 ```
 
@@ -50,13 +51,28 @@ buzz
 The following message is discarded since its `x` field contains the sequence of characters _bar_, contained in the list.
 
 ```json
-{"x":"halbart","y":1}
+{
+  "x": "halbart",
+  "y": 1
+}
 ```
 
-While the following message is passed through as the target field specified in the config is not _y_ but _x_ .
+While the following message is passed through as the target field specified in the config is not _y_ but _x_.
 
 ```json
-{"x":1,"y":"halbart"}
+{
+  "x": 1,
+  "y": "halbart"
+}
+```
+
+Additionally, the following message is also omitted since `filter_empty` is `true`. The value is determined to be empty when the trimed value is empty.
+
+```json
+{
+  "x": "   ",
+  "y": "halbart"
+}
 ```
 
 #### IPMatcher
@@ -81,19 +97,28 @@ Given the `blacklist.txt` is as follows.
 The following message is discarded since its `ip` field is the IP address in the list (exact IP).
 
 ```json
-{"ip":"255.255.0.0","y":1}
+{
+  "ip": "255.255.0.0",
+  "y": 1
+}
 ```
 
 Also the following message is discarded since its `ip` field is the IP address in the list (CIDR-notated IP).
 
 ```json
-{"ip":"192.168.1.255","y":1}
+{
+  "ip": "192.168.1.255",
+  "y": 1
+}
 ```
 
 While the following message is passed through.
 
 ```json
-{"ip":"192.168.2.0","y":1}
+{
+  "ip": "192.168.2.0",
+  "y": 1
+}
 ```
 
 ### Output plugin
@@ -106,6 +131,7 @@ The other use case is to filter messages likewise, but process the filtered mess
 
   key_to_filter field_name_you_want_to_filter
   patterns_file_path file_including_patterns_separated_by_new_line
+  filter_empty true
 
   <retag>
     add_prefix x # retag non-filtered messages whose tag will be "x.your_tag"

data/lib/fluent/plugin/filter_filter_list.rb

@@ -14,6 +14,7 @@ module Fluent
       config_param :filter, :string, default: 'AC'
       config_param :key_to_filter, :string, default: nil
       config_param :patterns_file_path, :string, default: ''
+      config_param :filter_empty, :bool, default: false
 
       def configure(conf)
         super
@@ -23,7 +24,7 @@ module Fluent
 
       def filter(_tag, _time, record)
         target = record[@key_to_filter]
-        return nil if target && @matcher.matches?(target)
+        return nil if target && (@matcher.matches?(target) || (@filter_empty && target.strip.empty?))
         record
       end
     end

data/lib/fluent/plugin/out_filter_list.rb

@@ -16,6 +16,7 @@ module Fluent
       config_param :filter, :string, default: 'AC'
       config_param :key_to_filter, :string, default: nil
       config_param :patterns_file_path, :string, default: ''
+      config_param :filter_empty, :bool, default: false
 
       config_section :retag, required: true, multi: false do
         config_param :tag, :string, default: nil
@@ -71,7 +72,7 @@ module Fluent
           target = record[@key_to_filter]
           log.debug "target: #{target}"
           # Do filter
-          if target && @matcher.matches?(target)
+          if target && (@matcher.matches?(target) || (@filter_empty && target.strip.empty?))
             if @retag_for_filtered
               t = @retag_for_filtered.tag || ((tag && !tag.empty?) ? @prefix_for_filtered_tag + tag : @retag_for_filtered.add_prefix)
               log.debug "re-emit with the tag: '#{t}', originally: '#{tag}'"

data/lib/fluent/plugin/out_filter_list/version.rb

@@ -1,7 +1,7 @@
 module Fluent
   module Plugin
     module FilterList
-      VERSION = "0.4.3"
+      VERSION = "0.5.0"
     end
   end
 end

data/lib/matcher.rb

@@ -12,7 +12,7 @@ module Matchers
 
     def matches?(text)
       node = @trie.root
-      text.to_s.split('').each do |char|
+      text.to_s.chars.each do |char|
         node = node.failure while node.children[char].nil? # Follow failure if it exists in case pattern doesn't match
         node = node.children[char]
         return true unless node.output.nil?
@@ -50,7 +50,7 @@ module Matchers
 
     def insert(pattern = '')
       current_node = @root
-      pattern.split('').each_with_index do |char, i|
+      pattern.chars.each_with_index do |char, i|
         current_node = current_node.insert(char)
         current_node.output = pattern if i == pattern.length - 1
       end
@@ -83,7 +83,7 @@ module Matchers
     def forward_match(pattern)
       return false if @root.children.empty?
       cur_node = @root
-      pattern.split('').each do |char|
+      pattern.chars.each do |char|
         return true if cur_node.children.empty?
         return false unless cur_node.children.key?(char)
         cur_node = cur_node.children[char]

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: fluent-plugin-filter-list
 version: !ruby/object:Gem::Version
-  version: 0.4.3
+  version: 0.5.0
 platform: ruby
 authors:
 - Shun Yanaura