twitter-text 1.13.0 → 1.13.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: bba5fb63e86ec7388a267dd85e07dd82b343cff3
-  data.tar.gz: a469ff1a40ca5bf59a8c1e58c4c4faf51420dced
+  metadata.gz: 93fd2b0a12b6a70531556e0d2d23386009900f42
+  data.tar.gz: 5fd5a7c55403f171491e492b964300adf4b19b2f
 SHA512:
-  metadata.gz: a35474299563dd6d64573b3289f5de44608d62070c33bf92900e161eb0739e06a08f54bc076bb4bdbb957fd25b0cdd7bf486e4e24fe2154a47842ea9bb864866
-  data.tar.gz: 88a929e5568e581954d670b90e4e5cb29fdc2b8008f811e737c7d383689b14c2501e84e39264b8554f1b9e19a084ab51537a0bd3abdd162aa0de0680bafd54e1
+  metadata.gz: 5d7bed341ec7a7e43be70d49bdc9ba2702350aa1981d75893eb52275e91eb7dee16daeb756372cf03115c6fc3dcd60bb2a1af9d666cd66699607aba64212537b
+  data.tar.gz: 2a4fe2a37fbfec1ef9fa6aeb00e25a42db5f8f141fc08ef1af40298c6f813b5b36f9410b53b3aa9a0de2c6062a61e8a88889d05cb542babedd4a0da38e5ce4c8

data/lib/twitter-text/regex.rb

@@ -105,10 +105,10 @@ module Twitter
 
     # A hashtag must contain at least one unicode letter or mark, as well as numbers, underscores, and select special characters.
     HASHTAG_ALPHA = /[\p{L}\p{M}]/
-    HASHTAG_ALPHANUMERIC = /[\p{L}\p{M}\p{Nd}_\u200c\u200d\u0482\ua673\ua67e\u05be\u05f3\u05f4\u309b\u309c\u30a0\u30fb\u3003\u0f0b\u0f0c\u00b7]/
+    HASHTAG_ALPHANUMERIC = /[\p{L}\p{M}\p{Nd}_\u200c\u200d\u0482\ua673\ua67e\u05be\u05f3\u05f4\uff5e\u301c\u309b\u309c\u30a0\u30fb\u3003\u0f0b\u0f0c\u00b7]/
     HASHTAG_BOUNDARY = /\A|\z|[^&\p{L}\p{M}\p{Nd}_\u200c\u200d\u0482\ua673\ua67e\u05be\u05f3\u05f4\u309b\u309c\u30a0\u30fb\u3003\u0f0b\u0f0c\u00b7]/
 
-    HASHTAG = /(#{HASHTAG_BOUNDARY})(#|＃)(#{HASHTAG_ALPHANUMERIC}*#{HASHTAG_ALPHA}#{HASHTAG_ALPHANUMERIC}*)/io
+    HASHTAG = /(#{HASHTAG_BOUNDARY})(#|＃)(?!\ufe0f|\u20e3)(#{HASHTAG_ALPHANUMERIC}*#{HASHTAG_ALPHA}#{HASHTAG_ALPHANUMERIC}*)/io
 
     REGEXEN[:valid_hashtag] = /#{HASHTAG}/io
     # Used in Extractor for final filtering
@@ -175,7 +175,7 @@ module Twitter
 
     REGEXEN[:valid_port_number] = /[0-9]+/
 
-    REGEXEN[:valid_general_url_path_chars] = /[a-z0-9!\*';:=\+\,\.\$\/%#\[\]\-_~&|@#{LATIN_ACCENTS}]/io
+    REGEXEN[:valid_general_url_path_chars] = /[a-z\p{Cyrillic}0-9!\*';:=\+\,\.\$\/%#\[\]\-_~&\|@#{LATIN_ACCENTS}]/io
     # Allow URL paths to contain up to two nested levels of balanced parens
     #  1. Used in Wikipedia URLs like /Primer_(film)
     #  2. Used in IIS sessions like /S(dfd346)/
@@ -198,7 +198,7 @@ module Twitter
     /iox
     # Valid end-of-path chracters (so /foo. does not gobble the period).
     #   1. Allow =&# for empty URL parameters and other URL-join artifacts
-    REGEXEN[:valid_url_path_ending_chars] = /[a-z0-9=_#\/\+\-#{LATIN_ACCENTS}]|(?:#{REGEXEN[:valid_url_balanced_parens]})/io
+    REGEXEN[:valid_url_path_ending_chars] = /[a-z\p{Cyrillic}0-9=_#\/\+\-#{LATIN_ACCENTS}]|(?:#{REGEXEN[:valid_url_balanced_parens]})/io
     REGEXEN[:valid_url_path] = /(?:
       (?:
         #{REGEXEN[:valid_general_url_path_chars]}*
@@ -226,7 +226,7 @@ module Twitter
     REGEXEN[:valid_cashtag] = /(^|#{REGEXEN[:spaces]})(\$)(#{REGEXEN[:cashtag]})(?=$|\s|[#{PUNCTUATION_CHARS}])/i
 
     # These URL validation pattern strings are based on the ABNF from RFC 3986
-    REGEXEN[:validate_url_unreserved] = /[a-z0-9\-._~]/i
+    REGEXEN[:validate_url_unreserved] = /[a-z\p{Cyrillic}0-9\-._~]/i
     REGEXEN[:validate_url_pct_encoded] = /(?:%[0-9a-f]{2})/i
     REGEXEN[:validate_url_sub_delims] = /[!$&'()*+,;=]/i
     REGEXEN[:validate_url_pchar] = /(?:

data/spec/validation_spec.rb

@@ -1,4 +1,4 @@
-# encoding: binary
+# encoding: utf-8
 require File.dirname(__FILE__) + '/spec_helper'
 
 class TestValidation

data/twitter-text.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name = "twitter-text"
-  s.version = "1.13.0"
+  s.version = "1.13.3"
   s.authors = ["Matt Sanford", "Patrick Ewing", "Ben Cherry", "Britt Selvitelle",
                "Raffi Krikorian", "J.P. Cummins", "Yoshimasa Niwa", "Keita Fujii", "James Koval"]
   s.email = ["matt@twitter.com", "patrick.henry.ewing@gmail.com", "bcherry@gmail.com", "bs@brittspace.com",

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: twitter-text
 version: !ruby/object:Gem::Version
-  version: 1.13.0
+  version: 1.13.3
 platform: ruby
 authors:
 - Matt Sanford
@@ -16,7 +16,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-08-26 00:00:00.000000000 Z
+date: 2016-02-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: multi_json
@@ -184,7 +184,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.7
+rubygems_version: 2.4.3
 signing_key: 
 specification_version: 4
 summary: Twitter text handling library