url_canonicalize 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 3dc466b71c5b337e8b92565b92356360e03b4292
-  data.tar.gz: c45bcc67c37dccf7d6d63eac37fa465b308990e3
+  metadata.gz: ad637f9ccbf0f81bdaf1d3d240c633d9b588c23d
+  data.tar.gz: 018fb65d2c324e8fd85e0a48016713de1585ea21
 SHA512:
-  metadata.gz: 518f8de7767b41a1898988a54ef4f27e3a296131a1961375f47fff5d0d3d1e24129c2645353d8bd997de1420c66af5b8bd2171d17242f59f5aa1c2a20b8be50b
-  data.tar.gz: da8f99769e69a3603a7665fe29b5d6b82d16e255cb7734b56f347aba3af159dc6a6a040ffa66775f7a3c8e297e2a5162182e773af488d5d1f5c59cbc7233d878
+  metadata.gz: 0f6174381673a8b8e6ff050956d6804fe4d6d5011a590ad5d1aeaf6b147ddb0f2064f98372e4a70d79418b3eeae483d154d522ae964beafce4dd90347a4bf589
+  data.tar.gz: 93cef8c77e155af01698830a9fbfb905d6e0db9182a8f603af0e2bbe760145d584545af16372e5c3a46097d35db3246eaa0c764fbae3296333db048b8e4f0dba

data/lib/url_canonicalize/http.rb

@@ -20,7 +20,7 @@ module URLCanonicalize
 
     private
 
-    attr_reader :last_known_good
+    attr_accessor :last_known_good
 
     def initialize(raw_url)
       @raw_url = raw_url
@@ -55,6 +55,8 @@ module URLCanonicalize
     end
 
     def redirect_loop_detected?
+      puts 'Redirect' # debug
+
       if redirect_list.include?(response.url)
         return true if last_known_good
         raise URLCanonicalize::Exception::Redirect, 'Redirect loop detected'
@@ -85,7 +87,9 @@ module URLCanonicalize
     end
 
     def handle_canonical_found
-      @last_known_good = response.response
+      puts 'Canonical found' # debug
+
+      self.last_known_good = response.response
       return true if response.url == url || redirect_list.include?(response.url)
       set_url_from_response
       false
@@ -105,7 +109,9 @@ module URLCanonicalize
     end
 
     def handle_success
-      @last_known_good = response
+      puts 'Success' # debug
+
+      self.last_known_good = response
       true
     end
 

data/lib/url_canonicalize/request.rb

@@ -23,6 +23,8 @@ module URLCanonicalize
     end
 
     def handle_response
+      log_response
+
       case response
       when Net::HTTPSuccess
         handle_success
@@ -52,8 +54,11 @@ module URLCanonicalize
         self.http_method = :get
         handle_success
       else
-        location = relative_to_absolute(response['location'])
-        URLCanonicalize::Response::Redirect.new(location)
+        if location
+          URLCanonicalize::Response::Redirect.new(location)
+        else
+          URLCanonicalize::Response::Failure.new(::URI::InvalidURIError, response['location'])
+        end
       end
     end
 
@@ -63,6 +68,7 @@ module URLCanonicalize
 
     def enhanced_response
       if canonical_url
+        puts "  * canonical_url:\t#{canonical_url}" if ENV['DEBUG']
         response_plus = URLCanonicalize::Response::Success.new(canonical_url, response, html)
         URLCanonicalize::Response::CanonicalFound.new(canonical_url, response_plus)
       else
@@ -98,6 +104,10 @@ module URLCanonicalize
       @host ||= uri.host
     end
 
+    def location
+      @location ||= relative_to_absolute(response['location'])
+    end
+
     def request_for_method
       r = base_request
       headers.each { |header_key, header_value| r[header_key] = header_value }
@@ -147,10 +157,16 @@ module URLCanonicalize
       if partial_uri.host
         partial_url # It's already absolute
       else
-        base_uri = uri.dup || ::URI.parse(url)
-        base_uri.path = partial_url
-        base_uri.to_s
+        ::URI.join((uri || url), partial_url).to_s
       end
+    rescue ::URI::InvalidURIError
+      nil
+    end
+
+    def log_response
+      puts "#{http_method.upcase} #{url} #{response.code} #{response.message}" if ENV['DEBUG']
+      return unless ENV['DEBUG'].casecmp('headers')
+      response.each { |k, v| puts "  #{k}:\t#{v}"}
     end
 
     NETWORK_EXCEPTIONS = [
@@ -165,7 +181,9 @@ module URLCanonicalize
       Net::ReadTimeout,
       OpenSSL::SSL::SSLError,
       SocketError,
-      Timeout::Error
+      Timeout::Error,
+      Zlib::BufError,
+      Zlib::DataError
     ].freeze
   end
 end

data/lib/url_canonicalize/version.rb

@@ -1,3 +1,3 @@
 module URLCanonicalize
-  VERSION = '0.1.0'.freeze
+  VERSION = '0.1.1'.freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: url_canonicalize
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Dominic Sayers
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-10-20 00:00:00.000000000 Z
+date: 2016-10-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: addressable