deadfinder 1.5.1 → 1.6.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8ddb8f1182cbb94fdb1c270ff6d0af2ae176fdb2376a52496fc535c69a1c576c
-  data.tar.gz: 5f49781224441d738fade165e0fae6c851c6739de2959e8c32601ebd9495e16c
+  metadata.gz: 5a9dba525c92f09cb14019fe890f73dc85eaa975179ac19efca534f0fdc76b14
+  data.tar.gz: fee3168a2912dacd9a2740b1aaf2c495b0b68b48b4d024c208523d8a8a5bb5e1
 SHA512:
-  metadata.gz: 5875750e01d1b909390a566b94a4daeeee5e2df3da81528d2c2cf4f831ef92c5e1d46732bbe4425cf5b0ba751136faa9ac7cb0dffeb6beefa0a4fabfa805e2b7
-  data.tar.gz: bc62403a89307870b0166f530928f8ce1c951e131818b3ac55bc6d879e5e8273189dc120ee51740fccb3eaf2f473bc5d9bcfbb3441ae17c1793fe83f9c0de716
+  metadata.gz: be806719c1ed5ca905885222d485f376c7e1a61dff0388f812e05d52e279e21315af08fdcfb5b66e9773c88be2f1a5c8c434eba62e6340f91db76ea29f4643db
+  data.tar.gz: d966f9237d2a4f62ba98684346b045c49bd14b024723eda22ac5f0c9c2fb0e62fc358d6607d924c0cdd80cd10e8e5f4eb37a13a067188d8394d142fe0476001b

data/lib/deadfinder/utils.rb

@@ -4,41 +4,24 @@ require 'uri'
 
 def generate_url(text, base_url)
   node = text.to_s
+  return node if node.start_with?('http://', 'https://')
+
   begin
-    unless node.start_with?('http://', 'https://')
-      uri = URI(base_url)
-      if node.start_with? '//'
-        return "#{uri.scheme}:#{node}"
-      elsif node.start_with? '/'
-        return "#{uri.scheme}://#{uri.host}#{node}"
-      elsif ignore_scheme? node
-        return nil
-      else
-        return "#{extract_directory(uri)}#{node}"
-      end
+    uri = URI(base_url)
+    if node.start_with?('//')
+      "#{uri.scheme}:#{node}"
+    elsif node.start_with?('/')
+      "#{uri.scheme}://#{uri.host}#{node}"
+    elsif ignore_scheme?(node)
+      nil
+    else
+      URI.join(base_url, node).to_s
     end
   rescue StandardError
-    # puts e
+    nil
   end
-  node
 end
 
 def ignore_scheme?(url)
   url.start_with?('mailto:', 'tel:', 'sms:', 'data:', 'file:')
 end
-
-def extract_directory(uri)
-  return "#{uri.scheme}://#{uri.host}#{uri.path}" if uri.path.end_with?('/')
-
-  path_components = uri.path.split('/')
-  path_components.last
-  path_components.pop
-
-  directory_path = path_components.join('/')
-
-  if directory_path.start_with?('/')
-    "#{uri.scheme}://#{uri.host}#{directory_path}/"
-  else
-    "#{uri.scheme}://#{uri.host}/#{directory_path}/"
-  end
-end

data/lib/deadfinder/version.rb

@@ -1,3 +1,3 @@
 # frozen_string_literal: true
 
-VERSION = '1.5.1'
+VERSION = '1.6.0'

data/lib/deadfinder.rb

@@ -10,6 +10,8 @@ require 'deadfinder/version'
 require 'concurrent-edge'
 require 'sitemap-parser'
 require 'json'
+require 'yaml'
+require 'csv'
 
 Channel = Concurrent::Channel
 CACHE_SET = Concurrent::Map.new
@@ -22,6 +24,7 @@ class DeadFinderRunner
       'concurrency' => 50,
       'timeout' => 10,
       'output' => '',
+      'output_format' => 'json',
       'headers' => [],
       'worker_headers' => [],
       'silent' => true,
@@ -188,19 +191,40 @@ def run_sitemap(sitemap_url, options)
 end
 
 def gen_output(options)
+  return if options['output'].empty?
+
   output_data = OUTPUT.to_h
-  File.write(options['output'], JSON.pretty_generate(output_data)) unless options['output'].empty?
+  format = options['output_format'].to_s.downcase
+
+  content = case format
+            when 'yaml', 'yml'
+              output_data.to_yaml
+            when 'csv'
+              CSV.generate do |csv|
+                csv << ['target', 'url']
+                output_data.each do |target, urls|
+                  Array(urls).each do |url|
+                    csv << [target, url]
+                  end
+                end
+              end
+            else
+              JSON.pretty_generate(output_data)
+            end
+
+  File.write(options['output'], content)
 end
 
 class DeadFinder < Thor
   class_option :include30x, aliases: :r, default: false, type: :boolean, desc: 'Include 30x redirections'
   class_option :concurrency, aliases: :c, default: 50, type: :numeric, desc: 'Number of concurrency'
   class_option :timeout, aliases: :t, default: 10, type: :numeric, desc: 'Timeout in seconds'
-  class_option :output, aliases: :o, default: '', type: :string, desc: 'File to write JSON result'
+  class_option :output, aliases: :o, default: '', type: :string, desc: 'File to write result (e.g., json, yaml, csv)'
+  class_option :output_format, aliases: :f, default: 'json', type: :string, desc: 'Output format'
   class_option :headers, aliases: :H, default: [], type: :array,
                          desc: 'Custom HTTP headers to send with initial request'
   class_option :worker_headers, default: [], type: :array, desc: 'Custom HTTP headers to send with worker requests'
-  class_option :user_agent, default: 'Mozilla/5.0 (compatible; DeadFinder/1.5.1;)', type: :string,
+  class_option :user_agent, default: 'Mozilla/5.0 (compatible; DeadFinder/1.6.0;)', type: :string,
                             desc: 'User-Agent string to use for requests'
   class_option :proxy, aliases: :p, default: '', type: :string, desc: 'Proxy server to use for requests'
   class_option :silent, aliases: :s, default: false, type: :boolean, desc: 'Silent mode'

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: deadfinder
 version: !ruby/object:Gem::Version
-  version: 1.5.1
+  version: 1.6.0
 platform: ruby
 authors:
 - hahwul
 bindir: bin
 cert_chain: []
-date: 2024-12-26 00:00:00.000000000 Z
+date: 2025-02-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: colorize
@@ -203,7 +203,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.6.2
+rubygems_version: 3.6.3
 specification_version: 4
 summary: Find dead-links (broken links)
 test_files: []