Merge branch 'master' into memory_blob

lib/linguist/generated.rb

@@ -241,22 +241,26 @@ 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")

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)

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
@@ -1367,6 +1368,7 @@ Haml:
 Handlebars:
   type: markup
   color: "#01a9d6"
+  group: HTML
   aliases:
   - hbs
   - htmlbars
@@ -1575,6 +1577,14 @@ 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
@@ -1628,7 +1638,6 @@ JavaScript:
   - .jsfl
   - .jsm
   - .jss
-  - .jsx
   - .njs
   - .pac
   - .sjs
@@ -2448,6 +2457,7 @@ PLSQL:
   color: "#dad8d8"
   extensions:
   - .pls
+  - .pck
   - .pkb
   - .pks
   - .plb
@@ -3433,6 +3443,7 @@ TypeScript:
   - ts
   extensions:
   - .ts
+  - .tsx
   tm_scope: source.ts
   ace_mode: typescript
 
@@ -3649,6 +3660,7 @@ XML:
   - .tmSnippet
   - .tmTheme
   - .ts
+  - .tsx
   - .ui
   - .urdf
   - .vbproj

lib/linguist/lazy_blob.rb

@@ -79,6 +79,10 @@ module Linguist
       @size
     end
 
+    def cleanup!
+      @data.clear if @data
+    end
+
     protected
 
     # Returns true if the attribute is present and not the string "false".

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/version.rb

@@ -1,3 +1,3 @@
 module Linguist
-  VERSION = "4.5.15"
+  VERSION = "4.6.4"
 end