linkify-it-rb 1.0.0.3 → 1.1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 296b9d57ee2e27c227876ed70cf4ea5d24764e1f
-  data.tar.gz: 12b466103e584672eb59e4842dcdfb8c1b558580
+  metadata.gz: 5f47c7eb9a5b7f156b03e9b1ab2ca228bbcddd02
+  data.tar.gz: b5a2df2ecb0fad2c059e461da03b12a97641c975
 SHA512:
-  metadata.gz: 7489a676fa2bd5cb364fb8c0152b1b33b47cc11bbe34f03a8604ed901d63c7ecee09075c2ee60901de9c27c1a410b763b6f44a233499b3f2656fc28a86cc7457
-  data.tar.gz: 1f1c175f4be65ed8f98480703073788031d5a135fae59226b9140d0efa40e600ea6d44be8195f5ca715555b7e8c0c8ef37c67e6a14d1027746545121a401d285
+  metadata.gz: 57d20b3c3c4aedad32e3958496bae64cd6959c2e4a2b04f90fa1dfbefb5da7ecd6099f3f5e892bbceb680eca029b418c8c02b8f68415fd2dfe8d6616c8189a08
+  data.tar.gz: 978dc051818bd21fc1dcc788df060fe647a7efcd5c61033c4603c36f076fd07b3596e8c984e162dbf11278066ab0430701698fbae25dc02358ce427b908409ab

data/README.md

@@ -46,10 +46,11 @@ Usage examples
 ```ruby
 linkify = Linkify.new
 
-# add unoffocial `.mydomain` domain.
-linkify.tlds('.mydomain', true)  # Add unofficial `.mydomain` domain
+# add unoffocial `.onion` domain.
+linkify.tlds('.onion', true)     # Add unofficial `.onion` domain
 linkify.add('git:', 'http:')     # Add `git:` ptotocol as "alias"
 linkify.add('ftp:', null)        # Disable `ftp:` ptotocol
+linkify.set({fuzzyIP: true})     # Enable IPs in fuzzy links (without schema)
 
 linkify.test('Site github.com!'))
 => true
@@ -83,7 +84,7 @@ linkify.add('@', {
 API
 ---
 
-### LinkifyIt.new(schemas)
+### LinkifyIt.new(schemas, options)
 
 Creates new linkifier instance with optional additional schemas.
 
@@ -105,6 +106,13 @@ By default understands:
     - _normalize_ - optional block to normalize text & url of matched result
       (for example, for twitter mentions).
 
+`options`:
+
+- __fuzzyLink__ - recognige URL-s without `http(s):` prefix. Default `true`.
+- __fuzzyIP__ - allow IPs in fuzzy links above. Can conflict with some texts
+  like version numbers. Default `false`.
+- __fuzzyEmail__ - recognize emails without `mailto:` prefix.
+
 
 ### .test(text)
 
@@ -157,6 +165,9 @@ If list is replaced, then exact match for 2-chars root zones will be checked.
 Add new rule with `schema` prefix. For definition details see constructor
 description. To disable existing rule use `.add(name, nil)`
 
+### .set(options)
+
+Override default options. Missed properties will not be changed.
 
 ## License
 

data/lib/linkify-it-rb/index.rb

@@ -4,9 +4,18 @@ class Linkify
   attr_accessor   :__index__, :__last_index__, :__text_cache__, :__schema__, :__compiled__
   attr_accessor   :re, :bypass_normalizer
   
+  # RE pattern for 2-character tlds (autogenerated by ./support/tlds_2char_gen.js)
+  TLDS_2CH_SRC_RE = 'a[cdefgilmnoqrstuwxz]|b[abdefghijmnorstvwyz]|c[acdfghiklmnoruvwxyz]|d[ejkmoz]|e[cegrstu]|f[ijkmor]|g[abdefghilmnpqrstuwy]|h[kmnrtu]|i[delmnoqrst]|j[emop]|k[eghimnprwyz]|l[abcikrstuvy]|m[acdeghklmnopqrstuvwxyz]|n[acefgilopruz]|om|p[aefghklmnrstwy]|qa|r[eosuw]|s[abcdeghijklmnortuvxyz]|t[cdfghjklmnortvwz]|u[agksyz]|v[aceginu]|w[fs]|y[et]|z[amw]'
+
   # DON'T try to make PRs with changes. Extend TLDs with LinkifyIt.tlds() instead
   TLDS_DEFAULT = 'biz|com|edu|gov|net|org|pro|web|xxx|aero|asia|coop|info|museum|name|shop|рф'.split('|')
   
+  DEFAULT_OPTIONS = {
+    fuzzyLink: true,
+    fuzzyEmail: true,
+    fuzzyIP: false
+  }
+
   DEFAULT_SCHEMAS = {
     'http:' => {
       validate: lambda do |text, pos, obj|
@@ -56,6 +65,14 @@ class Linkify
     }
   }
 
+  # not needed
+  #------------------------------------------------------------------------------
+  # def isOptionsObj(obj)
+  #   return Object.keys(obj || {}).reduce(function (acc, k) {
+  #     return acc || defaultOptions.hasOwnProperty(k);
+  #   }, false);
+  # end
+
   #------------------------------------------------------------------------------
   def escapeRE(str)
     return str.gsub(/[\.\?\*\+\^\$\[\]\\\(\)\{\}\|\-]/, "\\$&")
@@ -91,12 +108,13 @@ class Linkify
 
     # Define dynamic patterns
     tlds = @__tlds__.dup
-    tlds.push('[a-z]{2}') if (!@__tlds_replaced__)
+    tlds.push(TLDS_2CH_SRC_RE) if (!@__tlds_replaced__)
     tlds.push(@re[:src_xn])
 
     @re[:src_tlds] = tlds.join('|')
     @re[:email_fuzzy]      = Regexp.new(LinkifyRe::TPL_EMAIL_FUZZY.gsub('%TLDS%', @re[:src_tlds]), true)
     @re[:link_fuzzy]       = Regexp.new(LinkifyRe::TPL_LINK_FUZZY.gsub('%TLDS%', @re[:src_tlds]), true)
+    @re[:link_no_ip_fuzzy] = Regexp.new(LinkifyRe::TPL_LINK_NO_IP_FUZZY.gsub('%TLDS%', @re[:src_tlds]), true)
     @re[:host_fuzzy_test]  = Regexp.new(LinkifyRe::TPL_HOST_FUZZY_TEST.gsub('%TLDS%', @re[:src_tlds]), true)
 
     #
@@ -237,7 +255,7 @@ class Linkify
 
 
 
-  # new LinkifyIt(schemas)
+  # new LinkifyIt(schemas, options)
   # - schemas (Object): Optional. Additional schemas to validate (prefix/validator)
   #
   # Creates new linkifier instance with optional additional schemas.
@@ -260,12 +278,30 @@ class Linkify
   #       or `RegExp`.
   #     - _normalize_ - optional function to normalize text & url of matched result
   #       (for example, for @twitter mentions).
+  #
+  # `options`:
+  #
+  # - __fuzzyLink__ - recognige URL-s without `http(s):` prefix. Default `true`.
+  # - __fuzzyIP__ - allow IPs in fuzzy links above. Can conflict with some texts
+  #   like version numbers. Default `false`.
+  # - __fuzzyEmail__ - recognize emails without `mailto:` prefix.
+  #
   #------------------------------------------------------------------------------
-  def initialize(schemas = {})
+  def initialize(schemas = {}, options = {})
     # if (!(this instanceof LinkifyIt)) {
-    #   return new LinkifyIt(schemas);
+    #   return new LinkifyIt(schemas, options);
     # }
 
+    #--- not needed, if you want to pass options, then must also pass schemas
+    # if options.empty?
+    #   if (isOptionsObj(schemas)) {
+    #     options = schemas;
+    #     schemas = {};
+    #   }
+    # }
+
+    @__opts__           = DEFAULT_OPTIONS.merge(options)
+
     # Cache last tested result. Used to skip repeating steps on next `match` call.
     @__index__          = -1
     @__last_index__     = -1 # Next scan position
@@ -299,6 +335,16 @@ class Linkify
     return self
   end
 
+  # chainable
+  # LinkifyIt#set(options)
+  # - options (Object): { fuzzyLink|fuzzyEmail|fuzzyIP: true|false }
+  #
+  # Set recognition options for links without schema.
+  #------------------------------------------------------------------------------
+  def set(options)
+    @__opts__.merge!(options)
+    return self
+  end
 
   # LinkifyIt#test(text) -> Boolean
   #
@@ -328,12 +374,12 @@ class Linkify
     end
 
     # guess schemaless links
-    if (@__compiled__['http:'])
+    if (@__opts__[:fuzzyLink] && @__compiled__['http:'])
       tld_pos = text.index(@re[:host_fuzzy_test])
       if !tld_pos.nil?
         # if tld is located after found link - no need to check fuzzy pattern
         if (@__index__ < 0 || tld_pos < @__index__)
-          if ((ml = text.match(@re[:link_fuzzy])) != nil)
+          if ((ml = text.match(@__opts__[:fuzzyIP] ? @re[:link_fuzzy] : @re[:link_no_ip_fuzzy])) != nil)
 
             shift = ml.begin(0) + ml[1].length
 
@@ -348,7 +394,7 @@ class Linkify
     end
 
     # guess schemaless emails
-    if (@__compiled__['mailto:'])
+    if (@__opts__[:fuzzyEmail] && @__compiled__['mailto:'])
       at_pos = text.index('@')
       if !at_pos.nil?
         # We can't skip this check, because this cases are possible:

data/lib/linkify-it-rb/re.rb

@@ -39,8 +39,11 @@ module LinkifyRe
             '\\"(?:(?!' + SRC_Z_CC + '|["]).)+\\"|' +
             "\\'(?:(?!" + SRC_Z_CC + "|[']).)+\\'|" +
             "\\'(?=" + SRC_PSEUDO_LETTER + ').|' +  # allow `I'm_king` if no pair found
-            '\\.{2,3}[a-zA-Z0-9%]|' + # github has ... in commit range links. Restrict to
-                                      # english & percent-encoded only, until more examples found.
+            '\\.{2,3}[a-zA-Z0-9%/]|' + # github has ... in commit range links. Restrict to
+                                       # - english
+                                       # - percent-encoded
+                                       # - parts of file path
+                                       # until more examples found.
             '\\.(?!' + SRC_Z_CC + '|[.]).|' +
             '\\-(?!' + SRC_Z_CC + '|--(?:[^-]|$))(?:[-]+|.)|' +  # `---` => long dash, terminate
             '\\,(?!' + SRC_Z_CC + ').|' +      # allow `,,,` in paths
@@ -90,11 +93,15 @@ module LinkifyRe
         '(?:(?:(?:' + SRC_DOMAIN + ')\\.)+(?:%TLDS%))' +
       ')'
 
+    TPL_HOST_NO_IP_FUZZY =
+      '(?:(?:(?:' + SRC_DOMAIN + ')\\.)+(?:%TLDS%))'
+
     SRC_HOST_STRICT            = SRC_HOST + SRC_HOST_TERMINATOR
     TPL_HOST_FUZZY_STRICT      = TPL_HOST_FUZZY + SRC_HOST_TERMINATOR
     SRC_HOST_PORT_STRICT       = SRC_HOST + SRC_PORT + SRC_HOST_TERMINATOR
     TPL_HOST_PORT_FUZZY_STRICT = TPL_HOST_FUZZY + SRC_PORT + SRC_HOST_TERMINATOR
-
+    TPL_HOST_PORT_NO_IP_FUZZY_STRICT = TPL_HOST_NO_IP_FUZZY + SRC_PORT + SRC_HOST_TERMINATOR
+      
     #------------------------------------------------------------------------------
     # Main rules
 
@@ -107,4 +114,9 @@ module LinkifyRe
         '(^|(?![.:/\\-_@])(?:[$+<=>^`|]|' + SRC_Z_P_CC + '))' +
         '((?![$+<=>^`|])' + TPL_HOST_PORT_FUZZY_STRICT + SRC_PATH + ')'
 
+    TPL_LINK_NO_IP_FUZZY =
+        # Fuzzy link can't be prepended with .:/\- and non punctuation.
+        # but can start with > (markdown blockquote)
+        '(^|(?![.:/\\-_@])(?:[$+<=>^`|]|' + SRC_Z_P_CC + '))' +
+        '((?![$+<=>^`|])' + TPL_HOST_PORT_NO_IP_FUZZY_STRICT + SRC_PATH + ')'
 end

data/lib/linkify-it-rb/version.rb

@@ -1,5 +1,5 @@
 module LinkifyIt
 
-  VERSION = '1.0.0.3'
+  VERSION = '1.1.1.0'
 
 end

data/spec/linkify-it-rb/test_spec.rb

@@ -3,7 +3,7 @@ fixture_dir  = File.join(File.dirname(__FILE__), 'fixtures')
 #------------------------------------------------------------------------------
 describe 'links' do
 
-  l = Linkify.new
+  l = Linkify.new({}, {fuzzyIP: true})
   l.bypass_normalizer = true    # kill the normalizer
 
   skipNext  = false
@@ -218,4 +218,41 @@ describe 'API' do
     expect(l.test('@@invalid')).to eq false
   end
 
+  #------------------------------------------------------------------------------
+  it 'set option: fuzzyLink' do
+    l = Linkify.new({}, { fuzzyLink: false })
+
+    expect(l.test('google.com.')).to eq false
+
+    l.set({ fuzzyLink: true })
+
+    expect(l.test('google.com.')).to eq true
+    expect(l.match('google.com.')[0].text).to eq 'google.com'
+  end
+
+
+  #------------------------------------------------------------------------------
+  it 'set option: fuzzyEmail' do
+    l = Linkify.new({}, { fuzzyEmail: false })
+
+    expect(l.test('foo@bar.com.')).to eq false
+
+    l.set({ fuzzyEmail: true })
+
+    expect(l.test('foo@bar.com.')).to eq true
+    expect(l.match('foo@bar.com.')[0].text).to eq 'foo@bar.com'
+  end
+
+  #------------------------------------------------------------------------------
+  it 'set option: fuzzyIP' do
+    l = Linkify.new
+
+    expect(l.test('1.1.1.1.')).to eq false
+
+    l.set({ fuzzyIP: true })
+
+    expect(l.test('1.1.1.1.')).to eq true
+    expect(l.match('1.1.1.1.')[0].text).to eq '1.1.1.1'
+  end
+
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: linkify-it-rb
 version: !ruby/object:Gem::Version
-  version: 1.0.0.3
+  version: 1.1.1.0
 platform: ruby
 authors:
 - Brett Walker
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-04-01 00:00:00.000000000 Z
+date: 2015-06-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: uc.micro-rb