Merge pull request #1787 from github/move-shebang

Move shebang (updated)

lib/linguist.rb

@@ -4,4 +4,5 @@ require 'linguist/heuristics'
 require 'linguist/language'
 require 'linguist/repository'
 require 'linguist/samples'
+require 'linguist/shebang'
 require 'linguist/version'

lib/linguist/language.rb

@@ -11,7 +11,7 @@ require 'linguist/samples'
 require 'linguist/file_blob'
 require 'linguist/blob_helper'
 require 'linguist/strategy/filename'
-require 'linguist/strategy/shebang'
+require 'linguist/shebang'
 
 module Linguist
   # Language names that are recognizable by GitHub. Defined languages
@@ -95,7 +95,7 @@ module Linguist
 
     STRATEGIES = [
       Linguist::Strategy::Filename,
-      Linguist::Strategy::Shebang,
+      Linguist::Shebang,
       Linguist::Heuristics,
       Linguist::Classifier
     ]
@@ -199,20 +199,26 @@ module Linguist
       @extension_index[extname]
     end
 
-    # Public: Look up Languages by shebang line.
+    # DEPRECATED
+    def self.find_by_shebang(data)
+      @interpreter_index[Shebang.interpreter(data)]
+    end
+
+    # Public: Look up Languages by interpreter.
     #
-    # data - Array of tokens or String data to analyze.
+    # interpreter - String of interpreter name
     #
     # Examples
     #
-    #   Language.find_by_shebang("#!/bin/bash\ndate;")
+    #   Language.find_by_interpreter("bash")
     #   # => [#<Language name="Bash">]
     #
     # Returns the matching Language
-    def self.find_by_shebang(data)
-      @interpreter_index[Linguist.interpreter_from_shebang(data)]
+    def self.find_by_interpreter(interpreter)
+      @interpreter_index[interpreter]
     end
 
+
     # Public: Look up Language by its name or lexer.
     #
     # name - The String name of the Language

lib/linguist/samples.rb

@@ -6,6 +6,7 @@ end
 
 require 'linguist/md5'
 require 'linguist/classifier'
+require 'linguist/shebang'
 
 module Linguist
   # Model for accessing classifier training data.
@@ -61,7 +62,7 @@ module Linguist
             yield({
               :path     => path,
               :language => category,
-              :interpreter => Linguist.interpreter_from_shebang(File.read(path)),
+              :interpreter => Shebang.interpreter(File.read(path)),
               :extname  => File.extname(filename)
             })
           end
@@ -114,41 +115,4 @@ module Linguist
       db
     end
   end
-
-  # Used to retrieve the interpreter from the shebang line of a file's
-  # data.
-  def self.interpreter_from_shebang(data)
-    lines = data.lines.to_a
-
-    if lines.any? && (match = lines[0].match(/(.+)\n?/)) && (bang = match[0]) =~ /^#!/
-      bang.sub!(/^#! /, '#!')
-      tokens = bang.split(' ')
-      pieces = tokens.first.split('/')
-
-      if pieces.size > 1
-        script = pieces.last
-      else
-        script = pieces.first.sub('#!', '')
-      end
-
-      script = script == 'env' ? tokens[1] : script
-
-      # If script has an invalid shebang, we might get here
-      return unless script
-
-      # "python2.6" -> "python2"
-      script.sub! $1, '' if script =~ /(\.\d+)$/
-
-      # Check for multiline shebang hacks that call `exec`
-      if script == 'sh' &&
-        lines[0...5].any? { |l| l.match(/exec (\w+).+\$0.+\$@/) }
-        script = $1
-      end
-      
-      File.basename(script)
-    else
-      nil
-    end
-  end
-
 end

lib/linguist/shebang.rb

@@ -0,0 +1,44 @@
+module Linguist
+  class Shebang
+    # Public: Use shebang to detect language of the blob.
+    #
+    # blob               - An object that quacks like a blob.
+    #
+    # Examples
+    #
+    #   Shebang.call(FileBlob.new("path/to/file"))
+    #
+    # Returns an Array with one Language if the blob has a shebang with a valid
+    # interpreter, or empty if there is no shebang.
+    def self.call(blob, _ = nil)
+      Language.find_by_interpreter interpreter(blob.data)
+    end
+
+    # Public: Get the interpreter from the shebang
+    #
+    # Returns a String or nil
+    def self.interpreter(data)
+      lines = data.lines
+      return unless match = /^#! ?(.*)$/.match(lines.first)
+
+      tokens = match[1].split(' ')
+      script = tokens.first.split('/').last
+
+      script = tokens[1] if script == 'env'
+
+      # If script has an invalid shebang, we might get here
+      return unless script
+
+      # "python2.6" -> "python2"
+      script.sub! $1, '' if script =~ /(\.\d+)$/
+
+      # Check for multiline shebang hacks that call `exec`
+      if script == 'sh' &&
+        lines.first(5).any? { |l| l.match(/exec (\w+).+\$0.+\$@/) }
+        script = $1
+      end
+
+      File.basename(script)
+    end
+  end
+end

lib/linguist/strategy/shebang.rb

@@ -1,10 +0,0 @@
-module Linguist
-  module Strategy
-    # Check if there's a shebang line and use that as authoritative
-    class Shebang
-      def self.call(blob, _)
-        Language.find_by_shebang(blob.data)
-      end
-    end
-  end
-end

test/test_language.rb

@@ -223,34 +223,21 @@ class TestLanguage < Test::Unit::TestCase
     assert_equal [Language['Chapel']], Language.find_by_filename('examples/hello.chpl')
   end
 
-  def test_find_by_shebang
-    assert_equal 'ruby', Linguist.interpreter_from_shebang("#!/usr/bin/ruby\n# baz")
-    { []         => ["",
-                     "foo",
-                     "#bar",
-                     "#baz",
-                     "///",
-                     "\n\n\n\n\n",
-                     " #!/usr/sbin/ruby",
-                     "\n#!/usr/sbin/ruby"],
-      ['Ruby']   => ["#!/usr/bin/env ruby\n# baz",
-                     "#!/usr/sbin/ruby\n# bar",
-                     "#!/usr/bin/ruby\n# foo",
-                     "#!/usr/sbin/ruby",
-                     "#!/usr/sbin/ruby foo bar baz\n"],
-      ['R']      => ["#!/usr/bin/env Rscript\n# example R script\n#\n"],
-      ['Shell']  => ["#!/usr/bin/bash\n", "#!/bin/sh"],
-      ['Python'] => ["#!/bin/python\n# foo\n# bar\n# baz",
-                     "#!/usr/bin/python2.7\n\n\n\n",
-                     "#!/usr/bin/python3\n\n\n\n"],
-      ["Common Lisp"] => ["#!/usr/bin/sbcl --script\n\n"]
-    }.each do |languages, bodies|
-      bodies.each do |body|
-        assert_equal([body, languages.map{|l| Language[l]}],
-                     [body, Language.find_by_shebang(body)])
+  def test_find_by_interpreter
+    {
+      "ruby" => "Ruby",
+      "Rscript" => "R",
+      "sh" => "Shell",
+      "bash" => "Shell",
+      "python" => "Python",
+      "python2" => "Python",
+      "python3" => "Python",
+      "sbcl" => "Common Lisp"
+    }.each do |interpreter, language|
+      assert_equal [Language[language]], Language.find_by_interpreter(interpreter)
+    end
 
-      end
-    end
+    assert_equal [], Language.find_by_interpreter(nil)
   end
 
   def test_find

test/test_samples.rb

@@ -82,9 +82,4 @@ class TestSamples < Test::Unit::TestCase
       end
     end
   end
-
-  def test_shebang
-    assert_equal "crystal", Linguist.interpreter_from_shebang("#!/usr/bin/env bin/crystal")
-    assert_equal "python2", Linguist.interpreter_from_shebang("#!/usr/bin/python2.4")
-  end
 end

test/test_shebang.rb

@@ -0,0 +1,38 @@
+require_relative "./helper"
+
+class TestShebang < Test::Unit::TestCase
+  include Linguist
+
+  def assert_interpreter(interpreter, body)
+    assert_equal interpreter, Shebang.interpreter(body)
+  end
+
+  def test_shebangs
+    assert_interpreter nil, ""
+    assert_interpreter nil, "foo"
+    assert_interpreter nil, "#bar"
+    assert_interpreter nil, "#baz"
+    assert_interpreter nil, "///"
+    assert_interpreter nil, "\n\n\n\n\n"
+    assert_interpreter nil, " #!/usr/sbin/ruby"
+    assert_interpreter nil, "\n#!/usr/sbin/ruby"
+
+    assert_interpreter "ruby", "#!/usr/sbin/ruby\n# bar"
+    assert_interpreter "ruby", "#!/usr/bin/ruby\n# foo"
+    assert_interpreter "ruby", "#!/usr/sbin/ruby"
+    assert_interpreter "ruby", "#!/usr/sbin/ruby foo bar baz\n"
+
+    assert_interpreter "Rscript", "#!/usr/bin/env Rscript\n# example R script\n#\n"
+    assert_interpreter "crystal", "#!/usr/bin/env bin/crystal"
+    assert_interpreter "ruby", "#!/usr/bin/env ruby\n# baz"
+
+    assert_interpreter "bash", "#!/usr/bin/bash\n"
+    assert_interpreter "sh", "#!/bin/sh"
+    assert_interpreter "python", "#!/bin/python\n# foo\n# bar\n# baz"
+    assert_interpreter "python2", "#!/usr/bin/python2.7\n\n\n\n"
+    assert_interpreter "python3", "#!/usr/bin/python3\n\n\n\n"
+    assert_interpreter "sbcl", "#!/usr/bin/sbcl --script\n\n"
+    assert_interpreter "perl", "#! perl"
+  end
+
+end