Merge branch 'master' into go-vendor

* master: (168 commits)
  ruby for example
  Bumping version
  Updating grammars
  Grammar for Less from Atom package
  Remove Less grammar
  Updating to latest perl6 grammar
  Adding Perl6-specific grammar.
  Grammar for YANG from Atom package
  Support for YANG language
  Add detection of GrammarKit-generated files
  Add .xproj to list of XML extensions
  Test submodules are using HTTPS links
  Improved vim modeline detection
  Heuristic for Pod vs. Perl
  Bumping to v4.7.4
  Grammar update
  Support .rs.in as a file extension for Rust files.
  HTTPS links for submodules
  Add the LFE lexer as an example of erlang .xrl
  Add the Elixir parser as an example of erlang .yrl
  ...

lib/linguist/blob.rb

@@ -0,0 +1,73 @@
+require 'linguist/blob_helper'
+
+module Linguist
+  # A Blob is a wrapper around the content of a file to make it quack
+  # like a Grit::Blob. It provides the basic interface: `name`,
+  # `data`, `path` and `size`.
+  class Blob
+    include BlobHelper
+
+    # Public: Initialize a new Blob.
+    #
+    # path    - A path String (does not necessarily exists on the file system).
+    # content - Content of the file.
+    #
+    # Returns a Blob.
+    def initialize(path, content)
+      @path = path
+      @content = content
+    end
+
+    # Public: Filename
+    #
+    # Examples
+    #
+    #   Blob.new("/path/to/linguist/lib/linguist.rb", "").path
+    #   # =>  "/path/to/linguist/lib/linguist.rb"
+    #
+    # Returns a String
+    attr_reader :path
+
+    # Public: File name
+    #
+    # Returns a String
+    def name
+      File.basename(@path)
+    end
+
+    # Public: File contents.
+    #
+    # Returns a String.
+    def data
+      @content
+    end
+
+    # Public: Get byte size
+    #
+    # Returns an Integer.
+    def size
+      @content.bytesize
+    end
+
+    # Public: Get file extension.
+    #
+    # Returns a String.
+    def extension
+      extensions.last || ""
+    end
+
+    # Public: Return an array of the file extensions
+    #
+    #     >> Linguist::Blob.new("app/views/things/index.html.erb").extensions
+    #     => [".html.erb", ".erb"]
+    #
+    # Returns an Array
+    def extensions
+      basename, *segments = name.downcase.split(".")
+
+      segments.map.with_index do |segment, index|
+        "." + segments[index..-1].join(".")
+      end
+    end
+  end
+end

lib/linguist/file_blob.rb

@@ -1,10 +1,11 @@
 require 'linguist/blob_helper'
+require 'linguist/blob'
 
 module Linguist
   # A FileBlob is a wrapper around a File object to make it quack
   # like a Grit::Blob. It provides the basic interface: `name`,
   # `data`, `path` and `size`.
-  class FileBlob
+  class FileBlob < Blob
     include BlobHelper
 
     # Public: Initialize a new FileBlob from a path
@@ -18,20 +19,6 @@ module Linguist
       @path = base_path ? path.sub("#{base_path}/", '') : path
     end
 
-    # Public: Filename
-    #
-    # Examples
-    #
-    #   FileBlob.new("/path/to/linguist/lib/linguist.rb").path
-    #   # =>  "/path/to/linguist/lib/linguist.rb"
-    #
-    #   FileBlob.new("/path/to/linguist/lib/linguist.rb",
-    #                "/path/to/linguist").path
-    #   # =>  "lib/linguist.rb"
-    #
-    # Returns a String
-    attr_reader :path
-
     # Public: Read file permissions
     #
     # Returns a String like '100644'
@@ -39,13 +26,6 @@ module Linguist
       File.stat(@fullpath).mode.to_s(8)
     end
 
-    # Public: File name
-    #
-    # Returns a String
-    def name
-      File.basename(@fullpath)
-    end
-
     # Public: Read file contents.
     #
     # Returns a String.
@@ -59,26 +39,5 @@ module Linguist
     def size
       File.size(@fullpath)
     end
-
-    # Public: Get file extension.
-    #
-    # Returns a String.
-    def extension
-      extensions.last || ""
-    end
-
-    # Public: Return an array of the file extensions
-    #
-    #     >> Linguist::FileBlob.new("app/views/things/index.html.erb").extensions
-    #     => [".html.erb", ".erb"]
-    #
-    # Returns an Array
-    def extensions
-      basename, *segments = name.downcase.split(".")
-
-      segments.map.with_index do |segment, index|
-        "." + segments[index..-1].join(".")
-      end
-    end
   end
 end

lib/linguist/generated.rb

@@ -72,7 +72,10 @@ module Linguist
       generated_jni_header? ||
       vcr_cassette? ||
       generated_module? ||
-      generated_unity3d_meta?
+      generated_unity3d_meta? ||
+      generated_racc? ||
+      generated_jflex? ||
+      generated_grammarkit?
     end
 
     # Internal: Is the blob an Xcode file?
@@ -242,22 +245,26 @@ module Linguist
       return lines[0].include?("Code generated by")
     end
 
+    PROTOBUF_EXTENSIONS = ['.py', '.java', '.h', '.cc', '.cpp']
+
     # Internal: Is the blob a C++, Java or Python source file generated by the
     # Protocol Buffer compiler?
     #
     # Returns true of false.
     def generated_protocol_buffer?
-      return false unless ['.py', '.java', '.h', '.cc', '.cpp'].include?(extname)
+      return false unless PROTOBUF_EXTENSIONS.include?(extname)
       return false unless lines.count > 1
 
       return lines[0].include?("Generated by the protocol buffer compiler.  DO NOT EDIT!")
     end
 
+    APACHE_THRIFT_EXTENSIONS = ['.rb', '.py', '.go', '.js', '.m', '.java', '.h', '.cc', '.cpp']
+
     # Internal: Is the blob generated by Apache Thrift compiler?
     #
     # Returns true or false
     def generated_apache_thrift?
-      return false unless ['.rb', '.py', '.go', '.js', '.m', '.java', '.h', '.cc', '.cpp'].include?(extname)
+      return false unless APACHE_THRIFT_EXTENSIONS.include?(extname)
       return false unless lines.count > 1
 
       return lines[0].include?("Autogenerated by Thrift Compiler") || lines[1].include?("Autogenerated by Thrift Compiler")
@@ -366,5 +373,45 @@ module Linguist
       return false unless lines.count > 1
       return lines[0].include?("fileFormatVersion: ")
     end
+
+    # Internal: Is this a Racc-generated file?
+    #
+    # A Racc-generated file contains:
+    # # This file is automatically generated by Racc x.y.z
+    # on the third line.
+    #
+    # Return true or false
+    def generated_racc?
+      return false unless extname == '.rb'
+      return false unless lines.count > 2
+      return lines[2].start_with?("# This file is automatically generated by Racc")
+    end
+
+    # Internal: Is this a JFlex-generated file?
+    #
+    # A JFlex-generated file contains:
+    # /* The following code was generated by JFlex x.y.z on d/at/e ti:me */
+    # on the first line.
+    #
+    # Return true or false
+    def generated_jflex?
+      return false unless extname == '.java'
+      return false unless lines.count > 1
+      return lines[0].start_with?("/* The following code was generated by JFlex ")
+    end
+
+    # Internal: Is this a GrammarKit-generated file?
+    #
+    # A GrammarKit-generated file typically contain:
+    # // This is a generated file. Not intended for manual editing.
+    # on the first line. This is not always the case, as it's possible to
+    # customize the class header.
+    #
+    # Return true or false
+    def generated_grammarkit?
+      return false unless extname == '.java'
+      return false unless lines.count > 1
+      return lines[0].start_with?("// This is a generated file. Not intended for manual editing.")
+    end
   end
 end

lib/linguist/heuristics.rb

@@ -56,7 +56,8 @@ module Linguist
 
     # Internal: Check if this heuristic matches the candidate languages.
     def matches?(filename)
-      @extensions.any? { |ext| filename.downcase.end_with?(ext) }
+      filename = filename.downcase
+      @extensions.any? { |ext| filename.end_with?(ext) }
     end
 
     # Internal: Perform the heuristic
@@ -65,7 +66,7 @@ module Linguist
     end
 
     # Common heuristics
-    ObjectiveCRegex = /^[ \t]*@(interface|class|protocol|property|end|synchronized|selector|implementation)\b/
+    ObjectiveCRegex = /^\s*(@(interface|class|protocol|property|end|synchronised|selector|implementation)\b|#import\s+.+\.h[">])/
 
     disambiguate ".asc" do |data|
       if /^(----[- ]BEGIN|ssh-(rsa|dss)) /.match(data)
@@ -130,7 +131,7 @@ module Linguist
     disambiguate ".for", ".f" do |data|
       if /^: /.match(data)
         Language["Forth"]
-      elsif /^([c*][^a-z]|      (subroutine|program)\s|\s*!)/i.match(data)
+      elsif /^([c*][^abd-z]|      (subroutine|program|end)\s|\s*!)/i.match(data)
         Language["FORTRAN"]
       end
     end
@@ -237,8 +238,10 @@ module Linguist
     disambiguate ".ms" do |data|
       if /^[.'][a-z][a-z](\s|$)/i.match(data)
         Language["Groff"]
-      elsif /((^|\s)move?[. ])|\.(include|globa?l)\s/.match(data)
+      elsif /(?<!\S)\.(include|globa?l)\s/.match(data) || /(?<!\/\*)(\A|\n)\s*\.[A-Za-z]/.match(data.gsub(/"([^\\"]|\\.)*"|'([^\\']|\\.)*'|\\\s*(?:--.*)?\n/, ""))
         Language["GAS"]
+      else
+        Language["MAXScript"]
       end
     end
 
@@ -273,19 +276,27 @@ module Linguist
     end
 
     disambiguate ".pl" do |data|
-      if /^(use v6|(my )?class|module)/.match(data)
-        Language["Perl6"]
+      if /^[^#]+:-/.match(data)
+        Language["Prolog"]
       elsif /use strict|use\s+v?5\./.match(data)
         Language["Perl"]
-      elsif /^[^#]+:-/.match(data)
-        Language["Prolog"]
+      elsif /^(use v6|(my )?class|module)/.match(data)
+        Language["Perl6"]
       end
     end
 
     disambiguate ".pm", ".t" do |data|
-      if /^(use v6|(my )?class|module)/.match(data)
+      if /use strict|use\s+v?5\./.match(data)
+        Language["Perl"]
+      elsif /^(use v6|(my )?class|module)/.match(data)
         Language["Perl6"]
-      elsif /use strict|use\s+v?5\./.match(data)
+      end
+    end
+
+    disambiguate ".pod" do |data|
+      if /^=\w+$/.match(data)
+        Language["Pod"]
+      else
         Language["Perl"]
       end
     end

lib/linguist/languages.yml

@@ -8,7 +8,8 @@
 #                     Use "text" if a mode does not exist.
 # wrap              - Boolean wrap to enable line wrapping (default: false)
 # extensions        - An Array of associated extensions (the first one is
-#                     considered the primary extension)
+#                     considered the primary extension, the others should be
+#                     listed alphabetically)
 # interpreters      - An Array of associated interpreters
 # searchable        - Boolean flag to enable searching (defaults to true)
 # search_term       - Deprecated: Some languages maybe indexed under a
@@ -141,7 +142,7 @@ Agda:
 
 Alloy:
   type: programming  # 'modeling' would be more appropiate
-  color: "#cc5c24"
+  color: "#64C800"
   extensions:
   - .als
   ace_mode: text
@@ -182,6 +183,7 @@ AppleScript:
   interpreters:
   - osascript
   ace_mode: applescript
+  color: "#101F1F"
 
 Arc:
   type: programming
@@ -289,6 +291,7 @@ Batchfile:
   - .cmd
   tm_scope: source.dosbatch
   ace_mode: batchfile
+  color: "#C1F12E"
 
 Befunge:
   type: programming
@@ -303,6 +306,7 @@ Bison:
   extensions:
   - .bison
   ace_mode: text
+  color: "#6A463F"
 
 BitBake:
   type: programming
@@ -392,6 +396,7 @@ C#:
   - csharp
   extensions:
   - .cs
+  - .cake
   - .cshtml
   - .csx
 
@@ -539,6 +544,14 @@ Clean:
   tm_scope: none
   ace_mode: text
 
+Click:
+  type: programming
+  color: "#E4E6F3"
+  extensions:
+  - .click
+  tm_scope: source.click
+  ace_mode: text
+
 Clojure:
   type: programming
   ace_mode: clojure
@@ -567,6 +580,7 @@ CoffeeScript:
   extensions:
   - .coffee
   - ._coffee
+  - .cake
   - .cjsx
   - .cson
   - .iced
@@ -627,7 +641,7 @@ Common Lisp:
 
 Component Pascal:
   type: programming
-  color: "#b0ce4e"
+  color: "#B0CE4E"
   extensions:
   - .cp
   - .cps
@@ -690,6 +704,7 @@ Cucumber:
   aliases:
   - gherkin
   ace_mode: text
+  color: "#5B2063"
 
 Cuda:
   type: programming
@@ -698,6 +713,7 @@ Cuda:
   - .cuh
   tm_scope: source.cuda-c++
   ace_mode: c_cpp
+  color: "#3A4E3A"
 
 Cycript:
   type: programming
@@ -719,7 +735,7 @@ Cython:
 
 D:
   type: programming
-  color: "#fcd46d"
+  color: "#ba595e"
   extensions:
   - .d
   - .di
@@ -790,7 +806,6 @@ Dart:
 
 Diff:
   type: data
-  color: "#88dddd"
   extensions:
   - .diff
   - .patch
@@ -884,6 +899,8 @@ Elixir:
   ace_mode: elixir
   filenames:
   - mix.lock
+  interpreters:
+  - elixir
 
 Elm:
   type: programming
@@ -926,6 +943,8 @@ Erlang:
   - .es
   - .escript
   - .hrl
+  - .xrl
+  - .yrl
   filenames:
   - rebar.config
   - rebar.config.lock
@@ -1011,6 +1030,7 @@ Formatted:
   type: data
   extensions:
   - .for
+  - .eam.fs
   tm_scope: none
   ace_mode: text
 
@@ -1028,6 +1048,16 @@ Forth:
   - .fs
   ace_mode: forth
 
+FreeMarker:
+  type: programming
+  color: "#0050b2"
+  aliases:
+  - ftl
+  extensions:
+  - .ftl
+  tm_scope: text.html.ftl
+  ace_mode: ftl
+
 Frege:
   type: programming
   color: "#00cafe"
@@ -1317,11 +1347,22 @@ HTML+Django:
   - .mustache
   - .jinja
   aliases:
+  - django
   - html+django/jinja
   - html+jinja
   - htmldjango
   ace_mode: django
 
+HTML+EEX:
+  type: markup
+  tm_scope: text.html.elixir
+  group: HTML
+  aliases:
+  - eex
+  extensions:
+  - .eex
+  ace_mode: text
+
 HTML+ERB:
   type: markup
   tm_scope: text.html.erb
@@ -1331,7 +1372,7 @@ HTML+ERB:
   extensions:
   - .erb
   - .erb.deface
-  ace_mode: html_ruby
+  ace_mode: text
 
 HTML+PHP:
   type: markup
@@ -1355,6 +1396,7 @@ Hack:
   - .hh
   - .php
   tm_scope: text.html.php
+  color: "#878787"
 
 Haml:
   group: HTML
@@ -1363,10 +1405,12 @@ Haml:
   - .haml
   - .haml.deface
   ace_mode: haml
+  color: "#ECE2A9"
 
 Handlebars:
   type: markup
   color: "#01a9d6"
+  group: HTML
   aliases:
   - hbs
   - htmlbars
@@ -1474,7 +1518,7 @@ Inform 7:
   extensions:
   - .ni
   - .i7x
-  tm_scope: source.Inform7
+  tm_scope: source.inform7
   aliases:
   - i7
   - inform7
@@ -1575,12 +1619,20 @@ JSONiq:
   - .jq
   tm_scope: source.jq
 
+JSX:
+  type: programming
+  group: JavaScript
+  extensions:
+  - .jsx
+  tm_scope: source.js.jsx
+  ace_mode: javascript
+
 Jade:
   group: HTML
   type: markup
   extensions:
   - .jade
-  tm_scope: source.jade
+  tm_scope: text.jade
   ace_mode: jade
 
 Jasmin:
@@ -1625,10 +1677,10 @@ JavaScript:
   - .gs
   - .jake
   - .jsb
+  - .jscad
   - .jsfl
   - .jsm
   - .jss
-  - .jsx
   - .njs
   - .pac
   - .sjs
@@ -1660,6 +1712,18 @@ Julia:
   color: "#a270ba"
   ace_mode: julia
 
+Jupyter Notebook:
+  type: markup
+  ace_mode: json
+  tm_scope: source.json
+  color: "#DA5B0B"
+  extensions:
+  - .ipynb
+  filenames:
+  - Notebook
+  aliases:
+  - IPython Notebook
+
 KRL:
   type: programming
   color: "#28431f"
@@ -1672,6 +1736,7 @@ KiCad:
   type: programming
   extensions:
   - .sch
+  - .brd
   - .kicad_pcb
   tm_scope: none
   ace_mode: text
@@ -1707,6 +1772,7 @@ LLVM:
   extensions:
   - .ll
   ace_mode: text
+  color: "#185619"
 
 LOLCODE:
   type: programming
@@ -1770,6 +1836,7 @@ Less:
   - .less
   tm_scope: source.css.less
   ace_mode: less
+  color: "#A1D9A1"
 
 Lex:
   type: programming
@@ -1921,6 +1988,15 @@ M:
   tm_scope: source.lisp
   ace_mode: lisp
 
+MAXScript:
+  type: programming
+  color: "#00a6a6"
+  extensions:
+  - .ms
+  - .mcr
+  tm_scope: source.maxscript
+  ace_mode: text
+
 MTML:
   type: markup
   color: "#b7e1f4"
@@ -1979,6 +2055,7 @@ Markdown:
   - .mkdown
   - .ron
   tm_scope: source.gfm
+  color: "#083FA1"
 
 Mask:
   type: markup
@@ -1995,6 +2072,7 @@ Mathematica:
   - .cdf
   - .m
   - .ma
+  - .mt
   - .nb
   - .nbp
   - .wl
@@ -2006,6 +2084,8 @@ Mathematica:
 Matlab:
   type: programming
   color: "#bb92ac"
+  aliases:
+  - octave
   extensions:
   - .matlab
   - .m
@@ -2039,6 +2119,7 @@ MediaWiki:
   wrap: true
   extensions:
   - .mediawiki
+  - .wiki
   tm_scope: text.html.mediawiki
   ace_mode: text
 
@@ -2054,6 +2135,14 @@ Mercury:
   tm_scope: source.mercury
   ace_mode: prolog
 
+Metal:
+  type: programming
+  color: "#8f14e9"
+  extensions:
+  - .metal
+  tm_scope: source.c++
+  ace_mode: c_cpp
+
 MiniD: # Legacy
   type: programming
   searchable: false
@@ -2206,6 +2295,7 @@ Nginx:
   aliases:
   - nginx configuration file
   ace_mode: text
+  color: "#9469E9"
 
 Nimrod:
   type: programming
@@ -2264,6 +2354,7 @@ NumPy:
   - .numsc
   tm_scope: none
   ace_mode: text
+  color: "#9C8AF9"
 
 OCaml:
   type: programming
@@ -2432,6 +2523,7 @@ PHP:
   - .php3
   - .php4
   - .php5
+  - .phps
   - .phpt
   filenames:
   - Phakefile
@@ -2448,6 +2540,7 @@ PLSQL:
   color: "#dad8d8"
   extensions:
   - .pls
+  - .pck
   - .pkb
   - .pks
   - .plb
@@ -2512,7 +2605,7 @@ Parrot Internal Representation:
 
 Pascal:
   type: programming
-  color: "#b0ce4e"
+  color: "#E3F171"
   extensions:
   - .pas
   - .dfm
@@ -2561,9 +2654,16 @@ Perl6:
   - Rexfile
   interpreters:
   - perl6
-  tm_scope: source.perl.6
+  tm_scope: source.perl6fe
   ace_mode: perl
 
+Pickle:
+  type: data
+  extensions:
+  - .pkl
+  tm_scope: none
+  ace_mode: text
+
 PicoLisp:
   type: programming
   extensions:
@@ -2608,6 +2708,13 @@ PogoScript:
   tm_scope: source.pogoscript
   ace_mode: text
 
+Pony:
+  type: programming
+  extensions:
+  - .pony
+  tm_scope: source.pony
+  ace_mode: text
+
 PostScript:
   type: markup
   extensions:
@@ -2766,7 +2873,7 @@ QMake:
 
 R:
   type: programming
-  color: "#198ce7"
+  color: "#198CE7"
   aliases:
   - R
   - Rscript
@@ -2796,6 +2903,7 @@ RDoc:
   extensions:
   - .rdoc
   tm_scope: text.rdoc
+  color: "#8E84BF"
 
 REALbasic:
   type: programming
@@ -2891,6 +2999,17 @@ Redcode:
   tm_scope: none
   ace_mode: text
 
+Ren'Py:
+  type: programming
+  group: Python
+  aliases:
+  - renpy
+  color: "#ff7f7f"
+  extensions:
+  - .rpy
+  tm_scope: source.renpy
+  ace_mode: python
+
 RenderScript:
   type: programming
   extensions:
@@ -2976,6 +3095,7 @@ Rust:
   color: "#dea584"
   extensions:
   - .rs
+  - .rs.in
   ace_mode: rust
 
 SAS:
@@ -2993,6 +3113,7 @@ SCSS:
   ace_mode: scss
   extensions:
   - .scss
+  color: "#CF649A"
 
 SMT:
   type: programming
@@ -3095,11 +3216,12 @@ Sass:
   extensions:
   - .sass
   ace_mode: sass
+  color: "#CF649A"
 
 Scala:
   type: programming
   ace_mode: scala
-  color: "#7dd3b0"
+  color: "#DC322F"
   extensions:
   - .scala
   - .sbt
@@ -3251,6 +3373,14 @@ Squirrel:
   tm_scope: source.c++
   ace_mode: c_cpp
 
+Stan:
+  type: programming
+  color: "#b2011d"
+  extensions:
+  - .stan
+  ace_mode: text
+  tm_scope: source.stan
+
 Standard ML:
   type: programming
   color: "#dc566d"
@@ -3288,9 +3418,12 @@ SuperCollider:
   type: programming
   color: "#46390b"
   extensions:
-  - .scd
   - .sc
-  tm_scope: none
+  - .scd
+  interpreters:
+  - sclang
+  - scsynth
+  tm_scope: source.supercollider
   ace_mode: text
 
 Swift:
@@ -3433,6 +3566,7 @@ TypeScript:
   - ts
   extensions:
   - .ts
+  - .tsx
   tm_scope: source.ts
   ace_mode: typescript
 
@@ -3448,7 +3582,6 @@ Unified Parallel C:
 Unity3D Asset:
   type: data
   ace_mode: yaml
-  color: "#ab69a1"
   extensions:
   - .anim
   - .asset
@@ -3466,6 +3599,17 @@ UnrealScript:
   tm_scope: source.java
   ace_mode: java
 
+UrWeb:
+  type: programming
+  aliases:
+  - Ur/Web
+  - Ur
+  extensions:
+  - .ur
+  - .urs
+  tm_scope: source.ur
+  ace_mode: text
+
 VCL:
   group: Perl
   type: programming
@@ -3603,6 +3747,7 @@ XML:
   - .ccxml
   - .clixml
   - .cproject
+  - .csl
   - .csproj
   - .ct
   - .dita
@@ -3649,6 +3794,7 @@ XML:
   - .tmSnippet
   - .tmTheme
   - .ts
+  - .tsx
   - .ui
   - .urdf
   - .vbproj
@@ -3667,6 +3813,7 @@ XML:
   - .xliff
   - .xmi
   - .xml.dist
+  - .xproj
   - .xsd
   - .xul
   - .zcml
@@ -3723,6 +3870,7 @@ XSLT:
   - .xsl
   tm_scope: text.xml.xsl
   ace_mode: xml
+  color: "#EB8CEB"
 
 Xojo:
   type: programming
@@ -3756,6 +3904,13 @@ YAML:
   - .yaml-tmlanguage
   ace_mode: yaml
 
+YANG:
+  type: data
+  extensions:
+  - .yang
+  tm_scope: source.yang
+  ace_mode: text
+
 Yacc:
   type: programming
   extensions:
@@ -3764,6 +3919,7 @@ Yacc:
   - .yy
   tm_scope: source.bison
   ace_mode: text
+  color: "#4B6C4B"
 
 Zephir:
   type: programming
@@ -3803,7 +3959,6 @@ eC:
 edn:
   type: data
   ace_mode: clojure
-  color: "#db5855"
   extensions:
   - .edn
   tm_scope: source.clojure
@@ -3845,7 +4000,10 @@ reStructuredText:
   extensions:
   - .rst
   - .rest
+  - .rest.txt
+  - .rst.txt
   ace_mode: text
+  color: "#B3BCBC"
 
 wisp:
   type: programming

lib/linguist/lazy_blob.rb

@@ -79,11 +79,15 @@ module Linguist
       @size
     end
 
+    def cleanup!
+      @data.clear if @data
+    end
+
     protected
 
     # Returns true if the attribute is present and not the string "false".
-    def boolean_attribute(attr)
-      attr != "false"
+    def boolean_attribute(attribute)
+      attribute != "false"
     end
 
     def load_blob!

lib/linguist/repository.rb

@@ -126,12 +126,13 @@ module Linguist
     end
 
     protected
+    MAX_TREE_SIZE = 100_000
 
     def compute_stats(old_commit_oid, cache = nil)
+      return {} if current_tree.count_recursive(MAX_TREE_SIZE) >= MAX_TREE_SIZE
+
       old_tree = old_commit_oid && Rugged::Commit.lookup(repository, old_commit_oid).tree
-
       read_index
-
       diff = Rugged::Tree.diff(repository, old_tree, current_tree)
 
       # Clear file map and fetch full diff if any .gitattributes files are changed
@@ -157,8 +158,11 @@ module Linguist
 
           blob = Linguist::LazyBlob.new(repository, delta.new_file[:oid], new, mode.to_s(8))
 
-          next unless blob.include_in_language_stats?
-          file_map[new] = [blob.language.group.name, blob.size]
+          if blob.include_in_language_stats?
+            file_map[new] = [blob.language.group.name, blob.size]
+          end
+
+          blob.cleanup!
         end
       end
 

lib/linguist/strategy/modeline.rb

@@ -1,8 +1,19 @@
 module Linguist
   module Strategy
     class Modeline
-      EmacsModeline = /-\*-\s*(?:(?!mode)[\w-]+\s*:\s*(?:[\w+-]+)\s*;?\s*)*(?:mode\s*:)?\s*([\w+-]+)\s*(?:;\s*(?!mode)[\w-]+\s*:\s*[\w+-]+\s*)*;?\s*-\*-/i
-      VimModeline = /vim:\s*set.*\s(?:ft|filetype)=(\w+)\s?.*:/i
+      EMACS_MODELINE = /-\*-\s*(?:(?!mode)[\w-]+\s*:\s*(?:[\w+-]+)\s*;?\s*)*(?:mode\s*:)?\s*([\w+-]+)\s*(?:;\s*(?!mode)[\w-]+\s*:\s*[\w+-]+\s*)*;?\s*-\*-/i
+
+      # First form vim modeline
+      # [text]{white}{vi:|vim:|ex:}[white]{options}
+      # ex: 'vim: syntax=ruby'
+      VIM_MODELINE_1 = /(?:vim|vi|ex):\s*(?:ft|filetype|syntax)=(\w+)\s?/i
+
+      # Second form vim modeline (compatible with some versions of Vi)
+      # [text]{white}{vi:|vim:|Vim:|ex:}[white]se[t] {options}:[text]
+      # ex: 'vim set syntax=ruby:'
+      VIM_MODELINE_2 = /(?:vim|vi|Vim|ex):\s*se(?:t)?.*\s(?:ft|filetype|syntax)=(\w+)\s?.*:/i
+
+      MODELINES = [EMACS_MODELINE, VIM_MODELINE_1, VIM_MODELINE_2]
 
       # Public: Detects language based on Vim and Emacs modelines
       #
@@ -22,7 +33,7 @@ module Linguist
       #
       # Returns a String or nil
       def self.modeline(data)
-        match = data.match(EmacsModeline) || data.match(VimModeline)
+        match = MODELINES.map { |regex| data.match(regex) }.reject(&:nil?).first
         match[1] if match
       end
     end

lib/linguist/tokenizer.rb

@@ -86,13 +86,13 @@ module Linguist
           if s.peek(1) == "\""
             s.getch
           else
-            s.skip_until(/[^\\]"/)
+            s.skip_until(/(?<!\\)"/)
           end
         elsif s.scan(/'/)
           if s.peek(1) == "'"
             s.getch
           else
-            s.skip_until(/[^\\]'/)
+            s.skip_until(/(?<!\\)'/)
           end
 
         # Skip number literals

lib/linguist/vendor.yml

@@ -95,7 +95,7 @@
 - jquery.fn.gantt.js
 
 # jQuery fancyBox
-- jquery.fancybox.js
+- jquery.fancybox.(js|css)
 
 # Fuel UX
 - fuelux.js

lib/linguist/version.rb

@@ -1,3 +1,3 @@
 module Linguist
-  VERSION = "4.5.15"
+  VERSION = "4.7.5"
 end