bad_link_finder 0.3.2 → 0.3.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: eb79f23c7fabe10bbd9654a4eadfffde34ed2cac
-  data.tar.gz: bf8da7c36a356b4bf2281d78a872a897afc2d94a
+  metadata.gz: 7336841a6833dbf1369e624892d2c59fc75a782c
+  data.tar.gz: 744b7a4514a0ced554df8d9ae5030ad058ac882f
 SHA512:
-  metadata.gz: 97e222a53897c800715cec1c16c322213c9c563cde70aa483abda47aaf99e1ae8c8eefa2659969fc4260f9c8cc9999affdac0d2ff89ad18ff5f0de430653b664
-  data.tar.gz: fb7db07dbf3b04e1804c9e3ea555796fa47734dc2e1c2613bab2d7df3a338c9221dae8ce059a253e1f907217a18fdc5f7fade1e15f6736252b0dd01b07c9b10b
+  metadata.gz: f4646a14c3f8960d4cbb38e5aa9cebfe0552b5c0a516b4696b785db99beb393584fb88a3ee963dee6af5b15a82638870ca71aca56beda1eab75cee36b680c4ea
+  data.tar.gz: 798a38b30d5874655b16e64df255dccc50878eb4f9d15c8290794edf29794498e5661423556627bbfda5b41a558df9da9931b4d596a012ae40964ecb35f2f806

data/lib/bad_link_finder.rb

@@ -1,9 +1,10 @@
 require 'bad_link_finder/site_checker'
 require 'bad_link_finder/csv_builder'
 require 'pathname'
+require 'logger'
 
 module BadLinkFinder
-  def self.run
+  def self.run(logger = NullLogger.new)
     ['MIRROR_DIR', 'REPORT_OUTPUT_FILE', 'SITE_HOST'].each do |var|
       raise EnvironmentVariableError.new("Missing environment variable #{var}") unless ENV.has_key?(var)
     end
@@ -16,7 +17,7 @@ module BadLinkFinder
     csv_file = report_path.open('w')
     csv_builder = BadLinkFinder::CSVBuilder.new(csv_file)
 
-    BadLinkFinder::SiteChecker.new(ENV['MIRROR_DIR'], ENV['SITE_HOST'], csv_builder, ENV['START_FROM']).run
+    BadLinkFinder::SiteChecker.new(ENV['MIRROR_DIR'], ENV['SITE_HOST'], csv_builder, ENV['START_FROM'], logger).run
 
     csv_file.close
 
@@ -24,4 +25,12 @@ module BadLinkFinder
   end
 
   class EnvironmentVariableError < ArgumentError; end
+
+  class NullLogger < Logger
+    def initialize(*args)
+    end
+
+    def add(*args, &block)
+    end
+  end
 end

data/lib/bad_link_finder/link.rb

@@ -5,7 +5,8 @@ module BadLinkFinder
   class Link
     attr_reader :link, :url, :error_message, :exception
 
-    def initialize(page_url, link)
+    def initialize(page_url, link, logger = BadLinkFinder::NullLogger.new)
+      @logger = logger
       @page_url = page_url
       @link = link
       @url = get_url_from_link(link)
@@ -45,7 +46,7 @@ module BadLinkFinder
   protected
 
     def validate_with_request
-      puts "-- testing link #{@link} using #{@url}"
+      @logger.info "-- testing link #{@link} using #{@url}"
       sleep 0.1 # Recommended pause for gov.uk rate limiting
 
       browser = Mechanize.new
@@ -76,7 +77,7 @@ module BadLinkFinder
       @error_message = message
       @exception = exception
 
-      puts "---- found broken link #{@url}: #{message}: #{exception.message if exception}"
+      @logger.info "---- found broken link #{@url}: #{message}: #{exception.message if exception}"
     end
   end
 end

data/lib/bad_link_finder/page_checker.rb

@@ -2,28 +2,24 @@ require 'bad_link_finder/link'
 
 module BadLinkFinder
   class PageChecker
-    def initialize(host, page, result_cache)
+    def initialize(host, page, result_cache, logger = BadLinkFinder::NullLogger.new)
       host = host.chomp('/') + '/'
       @page = page
       @page_url = URI.join(host, page.path).to_s
       @result_cache = result_cache
+      @logger = logger
     end
 
     attr_reader :page_url
 
-    def each_bad_link(&block)
-      if @bad_links
-        @bad_links.each(&block)
-      else
-        @bad_links = @page.links.map do |raw_link|
-          link = @result_cache.fetch(raw_link) || @result_cache.store(raw_link, BadLinkFinder::Link.new(@page_url, raw_link))
+    def bad_links
+      @bad_links ||= @page.links.map { |link| fetch_or_build(link) }.reject(&:valid?)
+    end
+
+  private
 
-          unless link.valid?
-            yield link
-            next link
-          end
-        end.compact
-      end
+    def fetch_or_build(link)
+      @result_cache.fetch(link) || @result_cache.store(link, BadLinkFinder::Link.new(@page_url, link, @logger))
     end
   end
 end

data/lib/bad_link_finder/site_checker.rb

@@ -4,20 +4,21 @@ require 'bad_link_finder/page_checker'
 
 module BadLinkFinder
   class SiteChecker
-    def initialize(mirror_dir, host, csv_builder, start_from = nil)
+    def initialize(mirror_dir, host, csv_builder, start_from = nil, logger = BadLinkFinder::NullLogger.new)
       @mirror_dir = File.expand_path(mirror_dir)
       @host = host
       @csv_builder = csv_builder
       @start_from = start_from
       @result_cache = BadLinkFinder::ResultCache.new
+      @logger = logger
     end
 
     def run
       BadLinkFinder::Site.new(@mirror_dir, @start_from).each do |page|
-        page_checker = BadLinkFinder::PageChecker.new(@host, page, @result_cache)
-        puts "Checking page #{page.path} as #{page_checker.page_url}"
+        page_checker = BadLinkFinder::PageChecker.new(@host, page, @result_cache, @logger)
+        @logger.info "Checking page #{page.path} as #{page_checker.page_url}"
 
-        page_checker.each_bad_link do |link|
+        page_checker.bad_links.each do |link|
           @csv_builder << {
             url: page_checker.page_url,
             id: page.id,

data/lib/bad_link_finder/version.rb

@@ -1,3 +1,3 @@
 module BadLinkFinder
-  VERSION = "0.3.2"
+  VERSION = "0.3.3"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: bad_link_finder
 version: !ruby/object:Gem::Version
-  version: 0.3.2
+  version: 0.3.3
 platform: ruby
 authors:
 - Elliot Crosby-McCullough
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-31 00:00:00.000000000 Z
+date: 2014-05-15 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mechanize