typogrowth 0.9.3 → 0.9.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 !binary "U0hBMQ==":
-  metadata.gz: d96c37d2c10f0fc9f73caaeda6be11910a725c7d
-  data.tar.gz: abc183314746913a04adca10b88e380e75981a61
+  metadata.gz: b2fec1cd3faf12f3d9572b3ce6212e656426276c
+  data.tar.gz: ecc1a510f6fda9e9f40b71a93c61a2c03ecae9e7
 !binary "U0hBNTEy":
-  metadata.gz: 9277400dec9a0f43219676988fdac4dfc6e1b257537a9bdf8ba2218abcdf35c473d223f9ed9a07c02b14838687ca9aee923bf380bf3f5047e50534809ef016b2
-  data.tar.gz: 5fd258973bb34862e20c77a0ec8a6b203bb5d60be67ee9736e3a5b90107b21d63625507b5dfd8bac2bb9c7acd9cc40d944c57895de06b533fa9caba2cdba613b
+  metadata.gz: 421b2b311fe83738d9be9498334abf69005be6085ba30d47be3f41afece162d850f71340c88857c4a0b8ba998cb5c4261192d3013cd69c8b05b354fb570dd758
+  data.tar.gz: c4dd0579e140938e3b63177998ad6bdb5b5e3ea2ac821207c023a4bc90aa5f4eecbf06c0f22d93ba05eda6b36c7ce9870b6068b06d8c77229a3035a39d065fbc

data/features/step_definitions/typogrowth_steps.rb

@@ -19,6 +19,10 @@ When(/^input string is modified inplace with typo!$/) do
   @typoed.typo!
 end
 
+When(/^input string language is determined$/) do
+  @lang = @content.is_ru? ? "ru" : "us"
+end
+
 Then(/^neither single nor double quotes are left in the string$/) do
   @typo.scan(/"|'/).count.should == 0
 end
@@ -38,3 +42,7 @@ end
 Then(/^typoed result should equal to "(.*?)"$/) do |str|
   @typoed.should == str
 end
+
+Then(/^the language should equal to "(.*?)"$/) do |lang|
+  @lang.should == lang
+end

data/features/typogrowth.feature

@@ -67,3 +67,22 @@ Feature: Text is to be typographed (spacing and pubctuation are to be sanitized)
         | input                                  | output                                |
         | "This is a cat."                       | "This is a cat."                      |
 
+  Scenario Outline: Shadows handling
+    Given the input string is <input>
+    When input string is processed with Typogrowl’s typography parser
+    Then the typoed result should equal to <output>
+    And the call to string’s typo should equal to <output>
+
+    Examples:
+        | input                                  | output                                |
+        | "<p><img src="http://mudasobwa.ru/i/self.jpg">Here: http://wikipedia.ru</p>" | "<p><img src="http://mudasobwa.ru/i/self.jpg">Here: http://wikipedia.ru</p>"                      |
+
+  Scenario Outline: Language determining
+    Given the input string is <input>
+    When input string language is determined
+    Then the language should equal to <output>
+
+    Examples:
+        | input                                  | output                                |
+        | "<p><img src="http://mudasobwa.ru/i/self.jpg">Here: http://wikipedia.ru</p>" | "us" |
+        | "<p><img src="http://mudasobwa.ru/i/self.jpg">Здесь: http://wikipedia.ru</p>" | "ru" |

data/lib/typogrowth/string.rb

@@ -7,11 +7,15 @@ class String
   # Typographyes the string and returns a result
   # See Typogrowth::Parser#parse
   def typo lang = nil
-    Typogrowth::Parser.parse(self, lang: lang ? lang : I18n.locale)
+    Typogrowth.parse(self, lang: lang ? lang : I18n.locale)
   end
   # Typographyes the string inplace
   # See Typogrowth::Parser#parse!
   def typo! lang = nil
-    Typogrowth::Parser.parse!(self, lang: lang ? lang : I18n.locale)
+    Typogrowth.parse!(self, lang: lang ? lang : I18n.locale)
+  end
+
+  def is_ru? shadows = []
+    Typogrowth.is_ru? self, shadows: shadows
   end
 end

data/lib/typogrowth/version.rb

@@ -1,3 +1,3 @@
 module Typogrowth
-  VERSION = "0.9.3"
+  VERSION = "0.9.5"
 end

data/lib/typogrowth.rb

@@ -122,7 +122,16 @@ module Typogrowth
       }.join(%Q(
 
 ))
-      .gsub(/#{delims.first}(.*)#{delims.last}/m) { |m| Base64.decode64 m }
+      .gsub(/#{delims.first}(.*?)#{delims.last}/m) { |m|
+        Base64.decode64(m).force_encoding('UTF-8')
+      }
+    end
+
+    def is_ru? str, shadows: []
+      clean = @shadows.concat([*shadows]).uniq.inject(str) { |memo, re|
+        memo.gsub(re, '')
+      }
+      clean.scan(/[А-Яа-я]/).size > clean.length / 3
     end
 
     def add_shadows re
@@ -143,14 +152,19 @@ module Typogrowth
       str.replace self.parse str, lang: lang, shadows: shadows
     end
 
+    # Out-of-place version of `String` typographing. See #parse!
+    def self.is_ru? str, shadows: []
+      Parser.new.is_ru? str, shadows: shadows
+    end
+
     DEFAULT_SET = 'typogrowth'
-    HTML_TAG_RE = /<[A-Za-z]+(.*?)>/
+    HTML_TAG_RE = /<[^>]*>/
 
     def initialize file = nil
       file = DEFAULT_SET unless file
       @yaml = YAML.load_file "#{File.dirname(__FILE__)}/config/#{file}.yaml"
       @yaml.delete(:placeholder)
-      @shadows = [URI.regexp, HTML_TAG_RE]
+      @shadows = [HTML_TAG_RE, URI.regexp(['ftp', 'http', 'https', 'mailto'])]
     end
 
   end
@@ -162,5 +176,9 @@ module Typogrowth
   def self.parse! str, lang: :default, shadows: []
     Parser.parse! str, lang: lang, shadows: shadows
   end
+
+  def self.is_ru? str, shadows: []
+    Parser.is_ru? str, shadows: shadows
+  end
 end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: typogrowth
 version: !ruby/object:Gem::Version
-  version: 0.9.3
+  version: 0.9.5
 platform: ruby
 authors:
 - Alexei Matyushkin