optic14n 2.0.0

data/.gitignore

@@ -0,0 +1,17 @@
+*.gem
+*.rbc
+.bundle
+.config
+.yardoc
+Gemfile.lock
+InstalledFiles
+_yardoc
+coverage
+doc/
+lib/bundler/man
+pkg
+rdoc
+spec/reports
+test/tmp
+test/version_tmp
+tmp

data/Gemfile

@@ -0,0 +1,8 @@
+source 'https://rubygems.org'
+
+# Specify your gem's dependencies in optic14n.gemspec
+gemspec
+
+group :test do
+  gem 'rspec'
+end

data/LICENSE.txt

@@ -0,0 +1,22 @@
+Copyright (c) 2013 Government Digital Service
+
+MIT License
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/README.md

@@ -0,0 +1,60 @@
+# Optic14n
+
+Canonicalises URLs.
+
+## Installation
+
+Add this line to your application's Gemfile:
+
+    gem 'optic14n'
+
+And then execute:
+
+    $ bundle
+
+Or install it yourself as:
+
+    $ gem install optic14n
+
+## Usage
+
+Parse a `BLURI` like this:
+
+```ruby
+  bluri = BLURI('http://somewhere.com/?a=1&b=2&c=3')
+```
+
+Canonicalize it according to the [Previously-Established Rules](#the-previously-established-rules) thusly:
+
+```ruby
+  bluri.canonicalize!
+```
+
+You can also do site-specific stuff if you know some of the querystring will be valuable
+```ruby
+  bluri.canonicalize!(allow_query: :all)
+```
+
+```ruby
+  bluri.canonicalize!(allow_query: [:a, :c])
+  # or
+  bluri.canonicalize!(allow_query: ['a', 'c'])
+```
+
+### The previously-established rules
+
+This is a gem for canonicalising HTTP URIs such that we can boil our input set of URIs down to something that is much
+smaller than it would otherwise be. We do this aggressively by:
+
+* lowercasing URIs
+* removing query strings (unless told otherwise)
+* removing fragments
+* escaping and unescaping various characters and escape sequences according to RFC3986
+
+## Contributing
+
+1. Fork it
+2. Create your feature branch (`git checkout -b my-new-feature`)
+3. Commit your changes (`git commit -am 'Add some feature'`)
+4. Push to the branch (`git push origin my-new-feature`)
+5. Create new Pull Request

data/Rakefile

@@ -0,0 +1,19 @@
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+
+require 'bundler/gem_tasks'
+require 'rspec/core/rake_task'
+require 'optic14n'
+Dir.glob('lib/tasks/*.rake').each { |r| import r }
+
+
+require 'gem_publisher'
+desc 'Publish gem to Rubygems'
+task :publish_gem do
+  gem = GemPublisher.publish_if_updated('optic14n.gemspec', :rubygems)
+  puts "Published #{gem}" if gem
+end
+
+RSpec::Core::RakeTask.new(:spec)
+
+task default: :spec
+task test: :spec

data/jenkins.sh

@@ -0,0 +1,10 @@
+#!/bin/bash -x
+export RAILS_ENV=test
+export DISPLAY=":99"
+
+set -e
+rm -f Gemfile.lock
+bundle install --path "${HOME}/bundles/${JOB_NAME}"
+export GOVUK_APP_DOMAIN=dev.gov.uk
+bundle exec rake
+bundle exec rake publish_gem

data/lib/optic14n/canonicalized_urls.rb

@@ -0,0 +1,45 @@
+module Optic14n
+  ##
+  # Canonicalizes a set of URLs
+  class CanonicalizedUrls
+    attr_reader :output_set, :seen, :failures, :each
+
+    extend Forwardable
+
+    def_delegators :@output_set, :size
+
+    def initialize(urls, options)
+      @urls = urls
+      @options = options
+    end
+
+    def canonicalize!
+      @seen = 0
+      @failures = {}
+      @output_set = Set.new
+
+      @urls.each do |url|
+        begin
+          @output_set.add(BLURI(url).canonicalize!(@options))
+        rescue Exception => e
+          failures[url] = e
+        end
+        @seen += 1
+      end
+    end
+
+    def write(filename)
+      File.open(filename, 'w') do |file|
+        @output_set.each do |url|
+          file.puts url
+        end
+      end
+    end
+
+    ##
+    # Canonicalize given urls. +options+ will be passed to +BLURI.parse+
+    def self.from_urls(urls, options = {})
+      CanonicalizedUrls.new(urls, options).tap { |c| c.canonicalize! }
+    end
+  end
+end

data/lib/optic14n/version.rb

@@ -0,0 +1,3 @@
+module Optic14n
+  VERSION = '2.0.0'
+end

data/lib/optic14n.rb

@@ -0,0 +1,10 @@
+require 'optic14n/version'
+
+require 'uri'
+require 'addressable/uri'
+require 'cgi'
+require 'forwardable'
+require 'uri/query_hash'
+require 'uri/bluri'
+
+require 'optic14n/canonicalized_urls'

data/lib/tasks/measure_reduction.rake

@@ -0,0 +1,15 @@
+require 'set'
+
+namespace :opt do
+  desc 'Measure reduction from canonicalisation'
+  task :measure, [:filename, :output_file] do |_, args|
+    filename = args[:filename]
+    output_file = args[:output_file]
+
+    Optic14n::CanonicalizedUrls.from_urls(File.read(filename).each_line).tap do |urls|
+      urls.write(output_file) if output_file
+
+      puts "#{urls.seen} urls seen, #{urls.size} after canonicalisation"
+    end
+  end
+end

data/lib/uri/bluri.rb

@@ -0,0 +1,116 @@
+# encoding: utf-8
+
+module URI
+  ##
+  # A URI class with a bit extra for canonicalising query strings
+  #
+  class BLURI < URI::HTTP
+    PATH_ESCAPE_MAPPINGS = {
+      '[' => '%5b',
+      ']' => '%5d',
+      ',' => '%2c',
+      '"' => '%22',
+      "'" => '%27',
+      '|' => '%7c',
+      '!' => '%21',
+      '£' => '%c2%a3'
+    }
+
+    PATH_UNESCAPE_MAPPINGS = {
+      '%7e' => '~',
+      '%21' => '!'
+    }
+
+    REQUIRE_REGEX_ESCAPE = %w<. | ( ) [ ] { } + \ ^ $ * ?> & PATH_ESCAPE_MAPPINGS.keys
+
+    extend Forwardable
+
+    def_delegators :@uri, :scheme, :path, :host, :host=, :query, :fragment, :to_s
+
+    def initialize(uri_str)
+      @uri = ::Addressable::URI.parse(uri_str)
+      raise URI::InvalidURIError, "'#{uri_str}' not a valid URI" unless @uri
+    end
+
+    def query_hash
+      @query_hash ||= CGI::parse(self.query || '').tap do |query_hash|
+        # By default, CGI::parse produces lots of arrays. Usually they have a single element
+        # in them. That's correct but not terribly usable. Fix it here.
+        query_hash.each_pair { |k, v| query_hash[k] = v[0] if v.length == 1 }
+        query_hash.extend QueryHash
+      end
+    end
+
+    def query_hash=(value)
+      @query_hash = value
+      @uri.query = @query_hash.to_s == '' ? nil : @query_hash.to_s
+    end
+
+    def query=(query_str)
+      @query_hash = nil
+      @uri.query = query_str == '' ? nil : query_str
+    end
+
+    def self.parse(uri_str)
+      # Deal with known URI spec breaks - leading/trailing spaces and unencoded entities
+      if uri_str.is_a? String
+        uri_str = uri_str.strip.downcase.gsub(' ', '%20')
+        uri_str.gsub!('&', '%26') if uri_str =~ /^mailto:.*&.*/
+      end
+      BLURI.new(uri_str)
+    end
+
+    def has_query?
+      %w(http https).include?(@uri.scheme) && query
+    end
+
+    def canonicalize!(options = {})
+      @uri.scheme = 'http' if @uri.scheme == 'https'
+
+      @uri.path = @uri.path.sub(/\/*$/, '') if @uri.path =~ /^*\/$/
+      @uri.path.gsub!(BLURI.path_escape_char_regex,   PATH_ESCAPE_MAPPINGS)
+      @uri.path.gsub!(BLURI.path_unescape_code_regex, PATH_UNESCAPE_MAPPINGS)
+
+      canonicalize_query!(options)
+
+      @uri.fragment = nil
+      self
+    end
+
+    def canonicalize_query!(options)
+      allow_all = (options[:allow_query] == :all)
+      allowed_keys = [options[:allow_query]].flatten.compact unless allow_all
+
+      query_hash.keep_if do |k, _|
+        allow_all || (allowed_keys.include?(k) || allowed_keys.include?(k.to_sym))
+      end
+
+      self.query_hash = QueryHash[query_hash.sort_by { |k, _| k }]
+    end
+
+    ##
+    # Generate a regex which matches all characters in PATH_ESCAPE_MAPPINGS
+    def self.path_escape_char_regex
+      @path_escape_char_regex ||=
+          Regexp.new('[' + PATH_ESCAPE_MAPPINGS.keys.map do |char|
+            REQUIRE_REGEX_ESCAPE.include?(char) ? "\\#{char}" : char
+          end.join + ']')
+    end
+
+    ##
+    # Generate a regex which matches all escape sequences in PATH_UNESCAPE_MAPPINGS
+    def self.path_unescape_code_regex
+      @path_unescape_code_regex ||= Regexp.new(
+        PATH_UNESCAPE_MAPPINGS.keys.map { |code| "(?:#{code})" }.join('|')
+      )
+    end
+  end
+end
+
+module Kernel
+  def BLURI(uri_str)
+    ::URI::BLURI.parse(uri_str)
+  end
+
+  module_function :BLURI
+end

data/lib/uri/query_hash.rb

@@ -0,0 +1,33 @@
+module URI
+  ##
+  # Extends a hash with query string rendering/semi-indifferent access
+  module QueryHash
+    def [](key)
+      item = super key
+      item = super(key.to_s) if item.nil? || item.length == 0
+      item.class == Array && item.length == 0 ? nil : item
+    end
+
+    def to_s
+      keys.map { |key| render_value(key, self[key]) }.join('&')
+    end
+
+    ##
+    # Creates a new hash populated with the given objects.
+    def self.[](value)
+      Hash[value].tap do |hash|
+        hash.extend(QueryHash)
+      end
+    end
+
+    private
+
+    def render_value(key, value)
+      case value
+        when nil   then key
+        when Array then value.map { |el| render_value(key, el) }.join('&')
+        else            URI.encode_www_form_component(key) << '=' << URI.encode_www_form_component(value)
+      end
+    end
+  end
+end

data/optic14n.gemspec

@@ -0,0 +1,25 @@
+# coding: utf-8
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require 'optic14n/version'
+
+Gem::Specification.new do |spec|
+  spec.name          = 'optic14n'
+  spec.version       = Optic14n::VERSION
+  spec.authors       = ['Russell Garner']
+  spec.email         = %w(rgarner@zephyros-systems.co.uk)
+  spec.description   = %q{Canonicalises URLs.}
+  spec.summary       = %q{Specifically, HTTP URLs, for a limited purpose}
+  spec.homepage      = ''
+  spec.license       = 'MIT'
+
+  spec.files         = `git ls-files`.split($/)
+  spec.executables   = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
+  spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
+  spec.require_paths = %w(lib)
+
+  spec.add_dependency 'addressable', '~> 2.3'
+
+  spec.add_development_dependency 'rake'
+  spec.add_development_dependency 'gem_publisher', '~> 1.3.0'
+end

data/spec/bluri_spec.rb

@@ -0,0 +1,88 @@
+require 'spec_helper'
+
+describe URI::BLURI do
+  it 'should be an HTTP URI' do
+    bluri = BLURI('http://some.where.com')
+    bluri.should be_a URI::HTTP
+  end
+
+  it 'should not allow other schemes' do
+    lambda { BLURI('ftp://foo').should raise_error(ArgumentError) }
+  end
+
+  it 'should not allow nil' do
+    lambda { BLURI(nil) }.should raise_error(URI::InvalidURIError)
+  end
+
+  it 'supports scheme' do
+    BLURI('http://foo').scheme.should == 'http'
+  end
+  it 'supports host' do
+    BLURI('http://foo').host.should == 'foo'
+  end
+  it 'supports path' do
+    BLURI('http://foo/a/path').path.should == '/a/path'
+  end
+  it 'supports query' do
+    BLURI('http://foo?to=you&you=foo').query.should == 'to=you&you=foo'
+  end
+  it 'supports fragment' do
+    BLURI('http://foo#fragment').fragment.should == 'fragment'
+  end
+  it 'supports mailto:someone@somewhere' do
+    BLURI('mailto:me@there.com').to_s.should == 'mailto:me@there.com'
+  end
+  it 'corrects unencoded ampersands ins mailto' do # http://www.faqs.org/rfcs/rfc2368.html
+    BLURI('mailto:fruit&veg.newcastle@rpa.gsi.gov.uk').to_s.should == 'mailto:fruit%26veg.newcastle@rpa.gsi.gov.uk'
+  end
+  it 'corrects trailing spaces' do
+    BLURI('http://www.newspapersoc.org.uk ').to_s.should == 'http://www.newspapersoc.org.uk'
+  end
+  it 'corrects leading spaces' do
+    BLURI('  http://www.newspapersoc.org.uk').to_s.should == 'http://www.newspapersoc.org.uk'
+  end
+
+  describe 'Query string parsing' do
+    context 'the query string is of HTML-encoded form k=v&q=p' do
+      before do
+        @bluri = BLURI('http://some.com/a/path?itemid=1&type=RESOURCE')
+      end
+
+      it 'indexes the query string' do
+        @bluri.query_hash['itemid'].should == '1'
+      end
+
+      it 'allows indexing by symbol' do
+        @bluri.query_hash[:itemid].should == '1'
+      end
+
+      it 'shows nil for absent items' do
+        @bluri.query_hash[:eerie_flash].should == nil
+      end
+
+      it 'indexes the second query string item' do
+        @bluri.query_hash['type'].should == 'resource'
+      end
+
+      it 'allows setting of the query' do
+        @bluri.query = 'furry=really'
+        @bluri.to_s.should == 'http://some.com/a/path?furry=really'
+      end
+    end
+
+    context 'the querystring is not an HTML-encoded thing' do
+      before do
+        @bluri = BLURI('http://some.com/a/path?foo&bar')
+      end
+
+      it 'retains the query string' do
+        @bluri.query.should == 'foo&bar'
+      end
+
+      it 'has a query hash with empty elements' do
+        @bluri.query_hash['foo'].should == nil
+        @bluri.query_hash['foo'].should == nil
+      end
+    end
+  end
+end

data/spec/c14n.t

@@ -0,0 +1,71 @@
+# Here for reference, see original at
+# https://github.com/alphagov/redirector/blob/master/tests/lib/c14n.t
+
+use strict;
+use Test::More;
+require 'lib/c14n.pl';
+
+#
+#  case
+#
+is(c14n_url("http://www.EXAMPLE.COM/Foo/Bar/BAZ"), "http://www.example.com/foo/bar/baz", "c14n URL is lower-case");
+
+#
+#  protocol
+#
+is(c14n_url("https://www.example.com"), "http://www.example.com", "translates protocol to http");
+
+#
+#  slashes
+#
+is(c14n_url("http://www.example.com/"), "http://www.example.com", "drops trailing slash");
+is(c14n_url("http://www.example.com////"), "http://www.example.com", "drops multiple trailing slashes");
+
+#
+#  fragment identifier
+#
+is(c14n_url("http://www.example.com#foo"), "http://www.example.com", "drops fragment identifier");
+is(c14n_url("http://www.example.com/#foo"), "http://www.example.com", "drops fragment identifier and slashes");
+
+#
+#  encoding
+#
+is(c14n_url("http://www.example.com/:colon:"), "http://www.example.com/:colon:", "colons");
+is(c14n_url("http://www.example.com/~tide"), "http://www.example.com/~tide", "tide");
+is(c14n_url("http://www.example.com/_underscore_"), "http://www.example.com/_underscore_", "underscore");
+is(c14n_url("http://www.example.com/*asterisk*"), "http://www.example.com/*asterisk*", "asterisk");
+is(c14n_url("http://www.example.com/(parens)"), "http://www.example.com/(parens)", "parens");
+is(c14n_url("http://www.example.com/[square-brackets]"), "http://www.example.com/%5bsquare-brackets%5d", "square-brackets");
+
+is(c14n_url("http://www.example.com/commas,and-\"quotes\"-make-CSV-harder-to-'awk'"), 'http://www.example.com/commas%2cand-%22quotes%22-make-csv-harder-to-%27awk%27', "commas and quotes");
+is(c14n_url("http://www.example.com/problematic-in-curl[]||[and-regexes]"), "http://www.example.com/problematic-in-curl%5b%5d%7c%7c%5band-regexes%5d", "square brackets and pipes");
+is(c14n_url("http://www.example.com/%7eyes%20I%20have%20now%20read%20%5brfc%203986%5d%2C%20%26%20I%27m%20a%20%3Dlot%3D%20more%20reassured%21%21"),
+            'http://www.example.com/~yes%20i%20have%20now%20read%20%5brfc%203986%5d%2c%20%26%20i%27m%20a%20%3dlot%3d%20more%20reassured!!',
+            "non-reserved character percent decoding");
+
+is(c14n_url("https://www.example.com/pound-sign-£"), "http://www.example.com/pound-sign-%c2%a3", "pound sign");
+
+#
+#  query_strings
+#
+is(c14n_url("http://www.example.com?q=foo"), "http://www.example.com", "drops disallowed query-string");
+is(c14n_url("http://www.example.com/?q=foo"), "http://www.example.com", "drops disallowed query-string after slash");
+is(c14n_url("http://www.example.com/?q=foo#bar"), "http://www.example.com", "drops disallowed query-string after a slash with fragid");
+
+is(c14n_url("http://www.example.com?a=1&c=3&b=2", '*'), "http://www.example.com?a=1&b=2&c=3", "query string wildcard value");
+
+is(c14n_url("http://www.example.com/?q=foo", "q"), "http://www.example.com?q=foo", "allow named query_string parameter");
+
+is(c14n_url("http://www.example.com?c=23&d=1&b=909&e=33&a=1", "b,e,c,d,a"), "http://www.example.com?a=1&b=909&c=23&d=1&e=33", "sorts query_string values");
+is(c14n_url("http://www.example.com?c=23&d=1&b=909&e=33&a=1", "  b e,c:d, a  "), "http://www.example.com?a=1&b=909&c=23&d=1&e=33", "accept colon and space separated allowed values");
+is(c14n_url("http://www.example.com?c=23;d=1;b=909;e=33;a=1", "b,e,c,d,a"), "http://www.example.com?a=1&b=909&c=23&d=1&e=33", "converts matrix URI to query_string");
+
+is(c14n_url("http://www.example.com?a=2322sdfsf&topic=334499&q=909&item=23444", "topic,item"), "http://www.example.com?item=23444&topic=334499", "allows cherry-picked  query_string");
+is(c14n_url("http://www.example.com?a=2322sdfsf&topic=334499&q=909&item=23444", "foo,bar,baz"), "http://www.example.com", "no ? for empty query_string values");
+
+is(c14n_url("http://www.example.com?a=you're_dangerous", '*'), "http://www.example.com?a=you%27re_dangerous", "escape query string values");
+
+#
+#  normalise url
+#
+is(normalise_url("http://www.example.com/commas,and-\"quotes\"-make-CSV-harder-to-'awk'"), 'http://www.example.com/commas%2cand-%22quotes%22-make-CSV-harder-to-%27awk%27', "commas and quotes");

data/spec/c14n_spec.rb

@@ -0,0 +1,163 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+describe "Paul's tests, translated from Perl" do
+  it 'lowercases URLs' do
+    BLURI('http://www.EXAMPLE.COM/Foo/Bar/BAZ').canonicalize!.to_s.should == 'http://www.example.com/foo/bar/baz'
+  end
+
+  describe 'protocol' do
+    it 'translates protocol to http', reason: 'Reduces our input space, everything public anyway' do
+      BLURI('https://www.example.com').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+  end
+
+  describe 'slashes' do
+    it 'drops single trailing slashes' do
+      BLURI('http://www.example.com/').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+
+    it 'drops multiple trailing slashes' do
+      BLURI('http://www.example.com////').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+
+    it 'drops multiple trailing slashes on the path' do
+      BLURI('http://www.example.com/foo///').canonicalize!.to_s.should == 'http://www.example.com/foo'
+    end
+  end
+
+  describe 'fragments' do
+    it 'drops fragment identifier', reason: 'They won''t be mapped, so are redundant' do
+      BLURI('http://www.example.com#foo').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+    it 'drops fragment identifier and slashes' do
+      BLURI('http://www.example.com/#foo').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+  end
+
+  describe 'Things to keep verbatim or encode', reason: 'http://tools.ietf.org/html/rfc3986' do
+    it 'retains colons' do
+      BLURI('http://www.example.com/:colon:').canonicalize!.to_s.should == 'http://www.example.com/:colon:'
+    end
+    it 'retains tilde' do
+      BLURI('http://www.example.com/~tilde').canonicalize!.to_s.should == 'http://www.example.com/~tilde'
+    end
+    it 'retains underscores' do
+      BLURI('http://www.example.com/_underscore_').canonicalize!.to_s.should == 'http://www.example.com/_underscore_'
+    end
+    it 'retains asterisks' do
+      BLURI('http://www.example.com/*asterisk*').canonicalize!.to_s.should == 'http://www.example.com/*asterisk*'
+    end
+    it 'retains parens' do
+      BLURI('http://www.example.com/(parens)').canonicalize!.to_s.should == 'http://www.example.com/(parens)'
+    end
+    it 'escapes square brackets' do
+      BLURI('http://www.example.com/[square-brackets]').canonicalize!.to_s.should == 'http://www.example.com/%5bsquare-brackets%5d'
+    end
+    it 'encodes commas and quotes', reason: 'They make csv harder to awk' do
+      BLURI("http://www.example.com/commas,and-\"quotes\"-make-CSV-harder-to-'awk'").canonicalize!.to_s.should ==
+          'http://www.example.com/commas%2cand-%22quotes%22-make-csv-harder-to-%27awk%27'
+    end
+    it 'encodes square brackets and pipes', reason: "It's problematic in curl and regexes" do
+      BLURI('http://www.example.com/problematic-in-curl[]||[and-regexes]').canonicalize!.to_s.should ==
+          'http://www.example.com/problematic-in-curl%5b%5d%7c%7c%5band-regexes%5d'
+    end
+    it 'decodes non-reserved characters (! and ~)' do
+      # My god, it's full of stars
+      BLURI('http://www.example.com/%7eyes%20I%20have%20now%20read%20%5brfc%203986%5d%2C%20%26%20I%27m%20a%20%3Dlot%3D%20more%20reassured%21%21').
+          canonicalize!.to_s.should == 'http://www.example.com/~yes%20i%20have%20now%20read%20%5brfc%203986%5d%2c%20%26%20i%27m%20a%20%3dlot%3d%20more%20reassured!!'
+    end
+    it 'encodes pound signs' do
+      BLURI('https://www.example.com/pound-sign-£').canonicalize!.to_s.should == 'http://www.example.com/pound-sign-%c2%a3'
+    end
+  end
+
+  describe 'query strings' do
+    it 'disallows all query string params by default' do
+      BLURI('http://www.example.com?q=foo').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+    it 'disallows all params when there''s a slash' do
+      BLURI('http://www.example.com/?q=foo').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+    it 'disallows all params after a slash with fragid' do
+      BLURI('http://www.example.com/?q=foo#bar').canonicalize!.to_s.should == 'http://www.example.com'
+    end
+
+    describe 'allowing some or all query string values' do
+      it 'allows named query_string parameters' do
+        BLURI('http://www.example.com/?q=foo&r=bar').canonicalize!(allow_query: 'q').to_s.should ==
+            'http://www.example.com?q=foo'
+      end
+      it 'sorts query string values' do
+        BLURI('http://www.example.com?c=23&d=1&b=909&e=33&a=1').
+          canonicalize!(allow_query: [:b,:e,:c,:d,:a]).to_s.should == 'http://www.example.com?a=1&b=909&c=23&d=1&e=33'
+      end
+      it 'encodes querystring values' do
+        BLURI("http://www.example.com?a=you're_dangerous").canonicalize!(allow_query: :all).to_s.should ==
+          'http://www.example.com?a=you%27re_dangerous'
+      end
+      it 'whitelists and sorts query strings' do
+        BLURI('http://www.example.com?a=1&c=3&b=2').canonicalize!(allow_query: :all).to_s.should ==
+          'http://www.example.com?a=1&b=2&c=3'
+      end
+      it 'converts matrix URI to query_string' do
+        BLURI('http://www.example.com?c=23;d=1;b=909;e=33;a=1').
+          canonicalize!(allow_query: [:b,:e,:c,:d,:a]).to_s.should == 'http://www.example.com?a=1&b=909&c=23&d=1&e=33'
+      end
+      it 'sorts cherry-picked query string arguments' do
+        BLURI('http://www.example.com?a=2322sdfsf&topic=334499&q=909&item=23444').
+          canonicalize!(allow_query: [:topic, :item]).to_s.should == 'http://www.example.com?item=23444&topic=334499'
+      end
+      it 'ignores empty querystring values' do
+        BLURI('http://www.example.com?a=2322sdfsf&topic=334499&q=909&item=23444').
+          canonicalize!(allow_query: %w(foo bar baz)).to_s.should == 'http://www.example.com'
+      end
+
+      describe 'querystrings that are not an HTML-encoded thing' do
+        before do
+          @bluri = BLURI('http://some.com/a/path?foo&bar').canonicalize!(allow_query: :all)
+        end
+
+        it 'retains the query string' do
+          @bluri.query.should == 'bar&foo'
+        end
+
+        it 'has a query hash with empty elements' do
+          @bluri.query_hash['foo'].should == nil
+          @bluri.query_hash['bar'].should == nil
+        end
+
+        it 'renders the string properly' do
+          @bluri.query_hash.to_s.should == 'bar&foo'
+        end
+      end
+    end
+
+    describe 'degenerate cases' do
+      describe 'the treatment of query strings when there are query string octets that unescape to '\
+               'invalid UTF-8 sequences (we no longer treat these as failures)' do
+        it 'no longer raises exceptions when there are bad things in query values' do
+          BLURI('http://example.com/path?view=%ED').
+            canonicalize!(allow_query: :all).
+            to_s.should eql('http://example.com/path?view=%ED')
+        end
+
+        it 're-encodes correctly when there are bad things in query keys' do
+          BLURI('http://example.com/path?%ED=view').
+            canonicalize!(allow_query: :all).
+            to_s.should eql('http://example.com/path?%ED=view')
+        end
+      end
+
+      describe 'failure to canonicalize paths correctly' do
+        # see https://www.pivotaltracker.com/s/projects/860575/stories/54502932
+
+        subject { BLURI('http://www.voa.gov.uk/stuff/?query=thing').canonicalize!(allow_query: :all) }
+
+        its(:path) { should eql('/stuff') }
+        its(:query) { should eql('query=thing') }
+      end
+    end
+  end
+end

data/spec/canonicalized_urls_spec.rb

@@ -0,0 +1,59 @@
+require 'spec_helper'
+
+describe Optic14n::CanonicalizedUrls do
+  describe 'c14nize' do
+    let(:test_urls) do
+      %w(
+        http://www.qhm.mod.uk/portsmouth/leisure/fuel
+        http://www.qhm.mod.uk/portsmouth/leisure/lntm?
+        http://www.qhm.mod.uk/portsmouth/leisure/lntm?action=view
+        http://www.qhm.mod.uk/portsmouth/leisure/lntm?action=view&id=199
+        http://unistats.direct.gov.uk/searchResults.do?pname=institutesearchresults&level3Subjects=L3.90%AC10007761%ACFIRSTDEGREE%ACFulltime%AC430%ACNo%AC60%ACYes%AC83%ACNo%ACYes
+        1234://123
+      )
+    end
+
+    context 'options[:allow_query] is false' do
+      subject(:c14nizer) { Optic14n::CanonicalizedUrls.from_urls(test_urls, allow_query: false) }
+
+      it { should be_a(Optic14n::CanonicalizedUrls) }
+
+      its(:seen) { should eql(6) }
+
+      describe 'the output set' do
+        subject(:output_set) { c14nizer.output_set }
+
+        its(:size) { should eql(3) }
+
+        describe 'the items' do
+          subject { output_set.map(&:to_s) }
+
+          it { should include('http://www.qhm.mod.uk/portsmouth/leisure/fuel') }
+          it { should include('http://www.qhm.mod.uk/portsmouth/leisure/lntm') }
+          it { should include('http://unistats.direct.gov.uk/searchresults.do') }
+        end
+      end
+    end
+
+    context 'options[:allow_query] is :all' do
+      subject(:c14nizer) { Optic14n::CanonicalizedUrls.from_urls(test_urls, allow_query: :all) }
+
+      describe 'the output set' do
+        subject(:output_set) { c14nizer.output_set }
+
+        its(:size) { should eql(5) }
+      end
+
+      describe 'failures' do
+        subject(:failures) { c14nizer.failures }
+
+        it { should be_a(Hash) }
+
+        it 'has our last URL and an error' do
+          e = failures[test_urls.last]
+          e.should be_an(Addressable::URI::InvalidURIError)
+        end
+      end
+    end
+  end
+end

data/spec/query_hash_spec.rb

@@ -0,0 +1,15 @@
+require 'spec_helper'
+
+describe URI::QueryHash do
+  subject(:hash) { {}.extend URI::QueryHash }
+
+  its(:to_s) { should eql('') }
+
+  describe 'setting a value by symbol' do
+    before { hash['x'] = '1' }
+
+    its([:x])    { should eql('1') }
+    its(['x'])   { should eql('1') }
+    its(:to_s)   { should eql('x=1') }
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1 @@
+require 'optic14n'

data/spec/uri/query_hash_spec.rb

@@ -0,0 +1,11 @@
+require 'spec_helper'
+
+describe URI::QueryHash do
+  describe 'non-HTML encoded query strings' do
+    subject { { 'foo' => nil, 'bar' => nil }.extend URI::QueryHash }
+
+    its(['foo']) { should be_nil }
+    its(['bar']) { should be_nil }
+    its(:to_s)   { should eql('foo&bar') }
+  end
+end

metadata

@@ -0,0 +1,127 @@
+--- !ruby/object:Gem::Specification
+name: optic14n
+version: !ruby/object:Gem::Version
+  version: 2.0.0
+  prerelease: 
+platform: ruby
+authors:
+- Russell Garner
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2014-03-27 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: addressable
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '2.3'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '2.3'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: gem_publisher
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+description: Canonicalises URLs.
+email:
+- rgarner@zephyros-systems.co.uk
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- LICENSE.txt
+- README.md
+- Rakefile
+- jenkins.sh
+- lib/optic14n.rb
+- lib/optic14n/canonicalized_urls.rb
+- lib/optic14n/version.rb
+- lib/tasks/measure_reduction.rake
+- lib/uri/bluri.rb
+- lib/uri/query_hash.rb
+- optic14n.gemspec
+- spec/bluri_spec.rb
+- spec/c14n.t
+- spec/c14n_spec.rb
+- spec/canonicalized_urls_spec.rb
+- spec/query_hash_spec.rb
+- spec/spec_helper.rb
+- spec/uri/query_hash_spec.rb
+homepage: ''
+licenses:
+- MIT
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+      segments:
+      - 0
+      hash: 2602697415991458495
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+      segments:
+      - 0
+      hash: 2602697415991458495
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.23
+signing_key: 
+specification_version: 3
+summary: Specifically, HTTP URLs, for a limited purpose
+test_files:
+- spec/bluri_spec.rb
+- spec/c14n.t
+- spec/c14n_spec.rb
+- spec/canonicalized_urls_spec.rb
+- spec/query_hash_spec.rb
+- spec/spec_helper.rb
+- spec/uri/query_hash_spec.rb