Merge pull request #1674 from github/rework-heuristics

Rework heuristics

lib/linguist/blob_helper.rb

@@ -146,6 +146,13 @@ module Linguist
       end
     end
 
+    # Public: Is the blob empty?
+    #
+    # Return true or false
+    def empty?
+      data.nil? || data == ""
+    end
+
     # Public: Is the blob text?
     #
     # Return true or false

lib/linguist/heuristics.rb

@@ -13,26 +13,28 @@ module Linguist
     # Returns an array of Languages or []
     def self.find_by_heuristics(data, languages)
       if active?
+        result = []
+
         if languages.all? { |l| ["Perl", "Prolog"].include?(l) }
-          result = disambiguate_pl(data, languages)
+          result = disambiguate_pl(data)
         end
         if languages.all? { |l| ["ECL", "Prolog"].include?(l) }
-          result = disambiguate_ecl(data, languages)
+          result = disambiguate_ecl(data)
         end
         if languages.all? { |l| ["IDL", "Prolog"].include?(l) }
-          result = disambiguate_pro(data, languages)
+          result = disambiguate_pro(data)
         end
         if languages.all? { |l| ["Common Lisp", "OpenCL"].include?(l) }
-          result = disambiguate_cl(data, languages)
+          result = disambiguate_cl(data)
         end
         if languages.all? { |l| ["Hack", "PHP"].include?(l) }
-          result = disambiguate_hack(data, languages)
+          result = disambiguate_hack(data)
         end
         if languages.all? { |l| ["Scala", "SuperCollider"].include?(l) }
-          result = disambiguate_sc(data, languages)
+          result = disambiguate_sc(data)
         end
         if languages.all? { |l| ["AsciiDoc", "AGS Script"].include?(l) }
-          result = disambiguate_asc(data, languages)
+          result = disambiguate_asc(data)
         end
         return result
       end
@@ -42,28 +44,37 @@ module Linguist
     # We want to shortcut look for Objective-C _and_ now C++ too!
     #
     # Returns an array of Languages or []
-    def self.disambiguate_c(data, languages)
+    def self.disambiguate_c(data)
       matches = []
-      matches << Language["Objective-C"] if data.include?("@interface")
-      matches << Language["C++"] if data.include?("#include <cstdint>")
+      if data.include?("@interface")
+        matches << Language["Objective-C"]
+      elsif data.include?("#include <cstdint>")
+        matches << Language["C++"]
+      end
       matches
     end
 
-    def self.disambiguate_pl(data, languages)
+    def self.disambiguate_pl(data)
       matches = []
-      matches << Language["Prolog"] if data.include?(":-")
-      matches << Language["Perl"] if data.include?("use strict")
+      if data.include?("use strict")
+        matches << Language["Perl"]
+      elsif data.include?(":-")
+        matches << Language["Prolog"]
+      end
       matches
     end
 
-    def self.disambiguate_ecl(data, languages)
+    def self.disambiguate_ecl(data)
       matches = []
-      matches << Language["Prolog"] if data.include?(":-")
-      matches << Language["ECL"] if data.include?(":=")
+      if data.include?(":-")
+        matches << Language["Prolog"]
+      elsif data.include?(":=")
+        matches << Language["ECL"]
+      end
       matches
     end
 
-    def self.disambiguate_pro(data, languages)
+    def self.disambiguate_pro(data)
       matches = []
       if (data.include?(":-"))
         matches << Language["Prolog"]
@@ -73,7 +84,7 @@ module Linguist
       matches
     end
 
-    def self.disambiguate_ts(data, languages)
+    def self.disambiguate_ts(data)
       matches = []
       if (data.include?("</translation>"))
         matches << Language["XML"]
@@ -83,21 +94,24 @@ module Linguist
       matches
     end
 
-    def self.disambiguate_cl(data, languages)
+    def self.disambiguate_cl(data)
       matches = []
-      matches << Language["Common Lisp"] if data.include?("(defun ")
-      matches << Language["OpenCL"] if /\/\* |\/\/ |^\}/.match(data)
+      if data.include?("(defun ")
+        matches << Language["Common Lisp"]
+      elsif /\/\* |\/\/ |^\}/.match(data)
+        matches << Language["OpenCL"]
+      end
       matches
     end
 
-    def self.disambiguate_r(data, languages)
+    def self.disambiguate_r(data)
       matches = []
       matches << Language["Rebol"] if /\bRebol\b/i.match(data)
       matches << Language["R"] if data.include?("<-")
       matches
     end
 
-    def self.disambiguate_hack(data, languages)
+    def self.disambiguate_hack(data)
       matches = []
       if data.include?("<?hh")
         matches << Language["Hack"]
@@ -107,7 +121,7 @@ module Linguist
       matches
     end
 
-    def self.disambiguate_sc(data, languages)
+    def self.disambiguate_sc(data)
       matches = []
       if (/\^(this|super)\./.match(data) || /^\s*(\+|\*)\s*\w+\s*{/.match(data) || /^\s*~\w+\s*=\./.match(data))
         matches << Language["SuperCollider"]
@@ -118,7 +132,7 @@ module Linguist
       matches
     end
 
-    def self.disambiguate_asc(data, languages)
+    def self.disambiguate_asc(data)
       matches = []
       matches << Language["AsciiDoc"] if /^=+(\s|\n)/.match(data)
       matches

lib/linguist/language.rb

@@ -100,12 +100,8 @@ module Linguist
     def self.detect(blob)
       name = blob.name.to_s
 
-      # Check if the blob is possibly binary and bail early; this is a cheap
-      # test that uses the extension name to guess a binary binary mime type.
-      #
-      # We'll perform a more comprehensive test later which actually involves
-      # looking for binary characters in the blob
-      return nil if blob.likely_binary? || blob.binary?
+      # Bail early if the blob is binary or empty.
+      return nil if blob.likely_binary? || blob.binary? || blob.empty?
 
       # A bit of an elegant hack. If the file is executable but extensionless,
       # append a "magic" extension so it can be classified with other
@@ -124,16 +120,18 @@ module Linguist
       if possible_languages.length > 1
         data = blob.data
         possible_language_names = possible_languages.map(&:name)
+        heuristic_languages = Heuristics.find_by_heuristics(data, possible_language_names)
+
+        if heuristic_languages.size > 1
+          possible_language_names = heuristic_languages.map(&:name)
+        end
 
-        # Don't bother with binary contents or an empty file
-        if data.nil? || data == ""
-          nil
         # Check if there's a shebang line and use that as authoritative
-        elsif (result = find_by_shebang(data)) && !result.empty?
+        if (result = find_by_shebang(data)) && !result.empty?
           result.first
         # No shebang. Still more work to do. Try to find it with our heuristics.
-        elsif (determined = Heuristics.find_by_heuristics(data, possible_language_names)) && !determined.empty?
-          determined.first
+        elsif heuristic_languages.size == 1
+          heuristic_languages.first
         # Lastly, fall back to the probabilistic classifier.
         elsif classified = Classifier.classify(Samples.cache, data, possible_language_names).first
           # Return the actual Language object based of the string language name (i.e., first element of `#classify`)

samples/JavaScript/empty.js

@@ -1,3 +0,0 @@
-(function() {
-
-}).call(this);

test/test_blob.rb

@@ -226,7 +226,6 @@ class TestBlob < Test::Unit::TestCase
     assert !blob("PostScript/sierpinski.ps").generated?
 
     # These examples are too basic to tell
-    assert !blob("JavaScript/empty.js").generated?
     assert !blob("JavaScript/hello.js").generated?
 
     assert blob("JavaScript/intro-old.js").generated?
@@ -469,4 +468,13 @@ class TestBlob < Test::Unit::TestCase
   def test_minified_files_not_safe_to_highlight
     assert !blob("JavaScript/jquery-1.6.1.min.js").safe_to_colorize?
   end
+
+  def test_empty
+    blob = Struct.new(:data) { include Linguist::BlobHelper }
+
+    assert blob.new("").empty?
+    assert blob.new(nil).empty?
+    refute blob.new(" ").empty?
+    refute blob.new("nope").empty?
+  end
 end

test/test_heuristics.rb

@@ -20,18 +20,18 @@ class TestHeuristcs < Test::Unit::TestCase
     Dir.glob("#{samples_path}/#{language_name}/#{file}")
   end
 
+  # Candidate languages = ["C++", "Objective-C"]
   def test_obj_c_by_heuristics
-    languages = ["C++", "Objective-C"]
     # Only calling out '.h' filenames as these are the ones causing issues
     all_fixtures("Objective-C", "*.h").each do |fixture|
-      results = Heuristics.disambiguate_c(fixture("Objective-C/#{File.basename(fixture)}"), languages)
+      results = Heuristics.disambiguate_c(fixture("Objective-C/#{File.basename(fixture)}"))
       assert_equal Language["Objective-C"], results.first
     end
   end
 
+  # Candidate languages = ["C++", "Objective-C"]
   def test_cpp_by_heuristics
-    languages = ["C++", "Objective-C"]
-    results = Heuristics.disambiguate_c(fixture("C++/render_adapter.cpp"), languages)
+    results = Heuristics.disambiguate_c(fixture("C++/render_adapter.cpp"))
     assert_equal Language["C++"], results.first
   end
 
@@ -41,57 +41,57 @@ class TestHeuristcs < Test::Unit::TestCase
     assert_equal Language["Objective-C"], match
   end
 
+  # Candidate languages = ["Perl", "Prolog"]
   def test_pl_prolog_by_heuristics
-    languages = ["Perl", "Prolog"]
-    results = Heuristics.disambiguate_pl(fixture("Prolog/turing.pl"), languages)
+    results = Heuristics.disambiguate_pl(fixture("Prolog/turing.pl"))
     assert_equal Language["Prolog"], results.first
   end
 
+  # Candidate languages = ["Perl", "Prolog"]
   def test_pl_perl_by_heuristics
-    languages = ["Perl", "Prolog"]
-    results = Heuristics.disambiguate_pl(fixture("Perl/perl-test.t"), languages)
+    results = Heuristics.disambiguate_pl(fixture("Perl/perl-test.t"))
     assert_equal Language["Perl"], results.first
   end
 
+  # Candidate languages = ["ECL", "Prolog"]
   def test_ecl_prolog_by_heuristics
-    languages = ["ECL", "Prolog"]
-    results = Heuristics.disambiguate_ecl(fixture("Prolog/or-constraint.ecl"), languages)
+    results = Heuristics.disambiguate_ecl(fixture("Prolog/or-constraint.ecl"))
     assert_equal Language["Prolog"], results.first
   end
 
+  # Candidate languages = ["ECL", "Prolog"]
   def test_ecl_ecl_by_heuristics
-    languages = ["ECL", "Prolog"]
-    results = Heuristics.disambiguate_ecl(fixture("ECL/sample.ecl"), languages)
+    results = Heuristics.disambiguate_ecl(fixture("ECL/sample.ecl"))
     assert_equal Language["ECL"], results.first
   end
 
+  # Candidate languages = ["IDL", "Prolog"]
   def test_pro_prolog_by_heuristics
-    languages = ["IDL", "Prolog"]
-    results = Heuristics.disambiguate_pro(fixture("Prolog/logic-problem.pro"), languages)
+    results = Heuristics.disambiguate_pro(fixture("Prolog/logic-problem.pro"))
     assert_equal Language["Prolog"], results.first
   end
 
+  # Candidate languages = ["IDL", "Prolog"]
   def test_pro_idl_by_heuristics
-    languages = ["IDL", "Prolog"]
-    results = Heuristics.disambiguate_pro(fixture("IDL/mg_acosh.pro"), languages)
+    results = Heuristics.disambiguate_pro(fixture("IDL/mg_acosh.pro"))
     assert_equal Language["IDL"], results.first
   end
 
+  # Candidate languages = ["AGS Script", "AsciiDoc"]
   def test_asc_asciidoc_by_heuristics
-    languages = ["AGS Script", "AsciiDoc"]
-    results = Heuristics.disambiguate_asc(fixture("AsciiDoc/list.asc"), languages)
+    results = Heuristics.disambiguate_asc(fixture("AsciiDoc/list.asc"))
     assert_equal Language["AsciiDoc"], results.first
   end
 
+  # Candidate languages = ["TypeScript", "XML"]
   def test_ts_typescript_by_heuristics
-    languages = ["TypeScript", "XML"]
-    results = Heuristics.disambiguate_ts(fixture("TypeScript/classes.ts"), languages)
+    results = Heuristics.disambiguate_ts(fixture("TypeScript/classes.ts"))
     assert_equal Language["TypeScript"], results.first
   end
 
+  # Candidate languages = ["TypeScript", "XML"]
   def test_ts_xml_by_heuristics
-    languages = ["TypeScript", "XML"]
-    results = Heuristics.disambiguate_ts(fixture("XML/pt_BR.xml"), languages)
+    results = Heuristics.disambiguate_ts(fixture("XML/pt_BR.xml"))
     assert_equal Language["XML"], results.first
   end
 
@@ -99,27 +99,27 @@ class TestHeuristcs < Test::Unit::TestCase
     languages = ["Common Lisp", "OpenCL"]
     languages.each do |language|
       all_fixtures(language).each do |fixture|
-        results = Heuristics.disambiguate_cl(fixture("#{language}/#{File.basename(fixture)}"), languages)
+        results = Heuristics.disambiguate_cl(fixture("#{language}/#{File.basename(fixture)}"))
         assert_equal Language[language], results.first
       end
     end
   end
 
+  # Candidate languages = ["Hack", "PHP"]
   def test_hack_by_heuristics
-    languages = ["Hack", "PHP"]
-    results = Heuristics.disambiguate_hack(fixture("Hack/funs.php"), languages)
+    results = Heuristics.disambiguate_hack(fixture("Hack/funs.php"))
     assert_equal Language["Hack"], results.first
   end
 
+  # Candidate languages = ["Scala", "SuperCollider"]
   def test_sc_supercollider_by_heuristics
-    languages = ["Scala", "SuperCollider"]
-    results = Heuristics.disambiguate_sc(fixture("SuperCollider/WarpPreset.sc"), languages)
+    results = Heuristics.disambiguate_sc(fixture("SuperCollider/WarpPreset.sc"))
     assert_equal Language["SuperCollider"], results.first
   end
 
+  # Candidate languages = ["Scala", "SuperCollider"]
   def test_sc_scala_by_heuristics
-    languages = ["Scala", "SuperCollider"]
-    results = Heuristics.disambiguate_sc(fixture("Scala/node11.sc"), languages)
+    results = Heuristics.disambiguate_sc(fixture("Scala/node11.sc"))
     assert_equal Language["Scala"], results.first
   end
 end