normalize_url 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8028e7b8023b9a5bd1eedb26cc0a3f827573b0b5
-  data.tar.gz: abfdc72f52b71bcdb14d654f6f5c42fd9df3e682
+  metadata.gz: 540321091211ecaa6d08afffbd12cd43437b648b
+  data.tar.gz: 61ecab0b4e64c33edfe120666f05f7636fbe51b7
 SHA512:
-  metadata.gz: 8c9d689c0e501cdc118c026842eb40e6f3ff5a546e61e0cbb333f628c6f385af20fa5eba3b0088026ed21d3f7692b3e60e5fa6b0195f798294ddfad4f049da02
-  data.tar.gz: 43cfa103f3f30a8c5c8b5ea2f11ba2677950c2a19b7d9f5b10a3f883117a6830cfc2ec12352e81a5e4d955e055cc95af07d383438995ae05b747405cede52deb
+  metadata.gz: f4234b44bc0e43f256885814b39ee38d3a39b37a0953bf14a18c75fb9a0e166789d8877fffdf4e244f8f0686d19276b8aabdf0e77f97c625d5a7bc0c0b77d3e9
+  data.tar.gz: 4de7be9fb6120d2d7a83aa244e32d88510b7fa03d18e15c0c4e8dfbbfb20fadc539c55de21b645e24557c11c8fcd06f210e333b28814c37db622705cd785cf5e

data/README.md

@@ -1,4 +1,7 @@
 # NormalizeUrl
+[![Gem Version](https://img.shields.io/gem/v/normalize_url.svg)](https://rubygems.org/gems/normalize_url)
+[![Build Status](https://img.shields.io/travis/rwz/normalize_url.svg)](http://travis-ci.org/rwz/normalize_url)
+[![Code Climate](https://img.shields.io/codeclimate/github/rwz/normalize_url.svg)](https://codeclimate.com/github/rwz/normalize_url)
 
 This gem can normalize HTTP(S) URLs by applying a certain set of
 transformations. After normalization, two different URLs that point to the same
@@ -45,7 +48,7 @@ Each tranformation could be skipped by demand. All you need is to pass it as a
 optional value to `normalize` method:
 
 ```ruby
-NormalizeUrl.normalize("http://example.com/foo/", remove_trailing_slash: false) # => "http://example.com/foo/"
+NormalizeUrl.process("http://example.com/foo/", remove_trailing_slash: false) # => "http://example.com/foo/"
 ```
 
 ## Transformations

data/lib/normalize_url/normalizer.rb

@@ -1,18 +1,36 @@
+require "set"
 require "addressable/uri"
 
 module NormalizeUrl
   class Normalizer
     attr_reader :uri, :options
 
+    TRACKING_QUERY_PARAMS = %w[
+      utm_source
+      utm_medium
+      utm_term
+      utm_content
+      utm_campaign
+      sms_ss
+      awesm
+      xtor
+      PHPSESSID
+    ].to_set
+
     def initialize(original_uri, options={})
       @uri = Addressable::URI.parse(original_uri).normalize
       @options = options
+      fail_uri "only absolute URLs can be normalized" unless uri.absolute?
+      fail_uri "only HTTP/HTTPS URLs can be normalized" unless uri.scheme =~ /https?/
+    rescue Addressable::URI::InvalidURIError
+      fail_uri "#{original_uri.inspect} is not a URL"
     end
 
     def normalize
       process :remove_trailing_slash
       process :remove_repeating_slashes
       process :remove_hash
+      process :remove_tracking
       process :sort_query
       uri.to_s
     end
@@ -42,5 +60,21 @@ module NormalizeUrl
     def process_remove_repeating_slashes
       uri.path = uri.path.squeeze(?/) if uri.host
     end
+
+    def process_remove_tracking
+      return unless uri.query_values
+      original = uri.query_values
+      cleaned = original.reject{ |key, _| TRACKING_QUERY_PARAMS.include?(key) }
+
+      if cleaned.empty?
+        uri.query_values = nil
+      elsif cleaned != original
+        uri.query_values = cleaned
+      end
+    end
+
+    def fail_uri(message)
+      fail ArgumentError, message
+    end
   end
 end

data/lib/normalize_url/version.rb

@@ -1,3 +1,3 @@
 module NormalizeUrl
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: normalize_url
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Pavel Pravosud
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-10-17 00:00:00.000000000 Z
+date: 2014-10-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: addressable