From fd7633518f1aff15eee5338fd6715fa2717341fe Mon Sep 17 00:00:00 2001
From: Charlie Somerville <charlie@charliesomerville.com>
Date: Wed, 25 Feb 2015 12:34:07 +1100
Subject: [PATCH] add instrumentation to detection and classification

---
 lib/linguist.rb            | 12 ++++++++++++
 lib/linguist/classifier.rb |  8 +++++---
 lib/linguist/language.rb   | 28 +++++++++++++++-------------
 3 files changed, 32 insertions(+), 16 deletions(-)

diff --git a/lib/linguist.rb b/lib/linguist.rb
index ff9fc3a2..4419ff5b 100644
--- a/lib/linguist.rb
+++ b/lib/linguist.rb
@@ -6,3 +6,15 @@ require 'linguist/repository'
 require 'linguist/samples'
 require 'linguist/shebang'
 require 'linguist/version'
+
+class << Linguist
+  attr_accessor :instrumenter
+
+  def instrument(*args, &bk)
+    if instrumenter
+      instrumenter.instrument(*args, &bk)
+    else
+      yield
+    end
+  end
+end
diff --git a/lib/linguist/classifier.rb b/lib/linguist/classifier.rb
index 89a0df2f..208467e4 100644
--- a/lib/linguist/classifier.rb
+++ b/lib/linguist/classifier.rb
@@ -16,9 +16,11 @@ module Linguist
     #
     # Returns an Array of Language objects, most probable first.
     def self.call(blob, possible_languages)
-      language_names = possible_languages.map(&:name)
-      classify(Samples.cache, blob.data, language_names).map do |name, _|
-        Language[name] # Return the actual Language objects
+      Linguist.instrument("linguist.bayesian_classification") do
+        language_names = possible_languages.map(&:name)
+        classify(Samples.cache, blob.data, language_names).map do |name, _|
+          Language[name] # Return the actual Language objects
+        end
       end
     end
 
diff --git a/lib/linguist/language.rb b/lib/linguist/language.rb
index 2490a9f6..68b4c4fc 100644
--- a/lib/linguist/language.rb
+++ b/lib/linguist/language.rb
@@ -105,19 +105,21 @@ module Linguist
       # Bail early if the blob is binary or empty.
       return nil if blob.likely_binary? || blob.binary? || blob.empty?
 
-      # Call each strategy until one candidate is returned.
-      STRATEGIES.reduce([]) do |languages, strategy|
-        candidates = strategy.call(blob, languages)
-        if candidates.size == 1
-          return candidates.first
-        elsif candidates.size > 1
-          # More than one candidate was found, pass them to the next strategy.
-          candidates
-        else
-          # No candiates were found, pass on languages from the previous strategy.
-          languages
-        end
-      end.first
+      Linguist.instrument("linguist.detection") do
+        # Call each strategy until one candidate is returned.
+        STRATEGIES.reduce([]) do |languages, strategy|
+          candidates = strategy.call(blob, languages)
+          if candidates.size == 1
+            return candidates.first
+          elsif candidates.size > 1
+            # More than one candidate was found, pass them to the next strategy.
+            candidates
+          else
+            # No candiates were found, pass on languages from the previous strategy.
+            languages
+          end
+        end.first
+      end
     end
 
     # Public: Get all Languages