Merge pull request #850 from github/cpp_heuristic

A C++ shibboleth?

lib/linguist/heuristics.rb

@@ -20,12 +20,14 @@ module Linguist
     end
 
     # .h extensions are ambigious between C, C++, and Objective-C.
-    # We want to shortcut look for Objective-C.
+    # We want to shortcut look for Objective-C _and_ now C++ too!
     #
     # Returns an array of Languages or []
+    # TODO rename this method as we're not strictly disambiguating between .h files here.
     def self.disambiguate_h(data, languages)
       matches = []
       matches << Language["Objective-C"] if data.include?("@interface")
+      matches << Language["C++"] if data.include?("#include <cstdint>")
       matches
     end
 

lib/linguist/samples.json

@@ -511,8 +511,8 @@
       ".gemrc"
     ]
   },
-  "tokens_total": 436395,
-  "languages_total": 507,
+  "tokens_total": 436401,
+  "languages_total": 508,
   "tokens": {
     "ABAP": {
       "*/**": 1,
@@ -8562,7 +8562,7 @@
     "C++": {
       "class": 40,
       "Bar": 2,
-      "{": 580,
+      "{": 581,
       "protected": 4,
       "char": 127,
       "*name": 6,
@@ -8572,7 +8572,7 @@
       "hello": 2,
       "(": 2729,
       ")": 2731,
-      "}": 580,
+      "}": 581,
       "//": 278,
       "///": 843,
       "mainpage": 1,
@@ -9245,7 +9245,7 @@
       "#ifndef": 27,
       "BCM2835_H": 3,
       "#define": 341,
-      "#include": 120,
+      "#include": 121,
       "<stdint.h>": 2,
       "defgroup": 7,
       "constants": 1,
@@ -10120,7 +10120,7 @@
       "<string>": 4,
       "<vector>": 4,
       "<fstream>": 2,
-      "namespace": 30,
+      "namespace": 31,
       "std": 52,
       "DEFAULT_DELIMITER": 1,
       "CsvStreamer": 5,
@@ -11276,6 +11276,8 @@
       "wrapMode": 2,
       "wmode.": 1,
       "wmode": 1,
+      "<cstdint>": 2,
+      "Gui": 1,
       "rpc_init": 1,
       "rpc_server_loop": 1,
       "v8": 9,
@@ -11573,7 +11575,6 @@
       "desc": 2,
       "look": 1,
       "ahead": 1,
-      "<cstdint>": 1,
       "smallPrime_t": 1,
       "UTILS_H": 3,
       "<QtGlobal>": 1,
@@ -46342,7 +46343,7 @@
     "Bluespec": 1298,
     "Brightscript": 579,
     "C": 58858,
-    "C++": 31175,
+    "C++": 31181,
     "Ceylon": 50,
     "Clojure": 510,
     "COBOL": 90,
@@ -46478,7 +46479,7 @@
     "Bluespec": 2,
     "Brightscript": 1,
     "C": 26,
-    "C++": 26,
+    "C++": 27,
     "Ceylon": 1,
     "Clojure": 7,
     "COBOL": 4,
@@ -46600,5 +46601,5 @@
     "Xtend": 2,
     "YAML": 1
   },
-  "md5": "7ab5683c610f7e81d6ea5fb470111bbe"
+  "md5": "4bd97c3aa060155c82b903fea0fecc57"
 }

samples/C++/render_adapter.cpp

@@ -0,0 +1,6 @@
+#include <cstdint>
+
+namespace Gui
+{
+
+}

test/test_heuristics.rb

@@ -28,6 +28,12 @@ class TestHeuristcs < Test::Unit::TestCase
     end
   end
    
+  def test_cpp_by_heuristics
+    languages = ["C++", "Objective-C"]
+    results = Heuristics.disambiguate_h(fixture("C++/render_adapter.cpp"), languages)
+    assert_equal Language["C++"], results.first
+  end
+
   def test_detect_still_works_if_nothing_matches
     match = Language.detect("Hello.m", fixture("Objective-C/hello.m"))
     assert_equal Language["Objective-C"], match