Merge pull request #2654 from github/vmg/git-linguist-fixes

Small fixes for git-linguist

bin/git-linguist

@@ -8,20 +8,19 @@ require 'tmpdir'
 require 'zlib'
 
 class GitLinguist
-  attr_reader :repo_path
-  attr_reader :commit_oid
-  attr_reader :incremental
-
   def initialize(path, commit_oid, incremental = true)
     @repo_path = path
-    @commit_oid = commit_oid || rugged.head.target_id
+    @commit_oid = commit_oid
     @incremental = incremental
   end
 
   def linguist
-    repo = Linguist::Repository.new(rugged, commit_oid)
+    if @commit_oid.nil?
+      raise "git-linguist must be called with a specific commit OID to perform language computation"
+    end
+    repo = Linguist::Repository.new(rugged, @commit_oid)
 
-    if incremental && stats = load_language_stats
+    if @incremental && stats = load_language_stats
       old_commit_oid, old_stats = stats
 
       # A cache with NULL oid means that we want to froze
@@ -33,19 +32,19 @@ class GitLinguist
 
     result = yield repo
 
-    save_language_stats(commit_oid, repo.cache)
+    save_language_stats(@commit_oid, repo.cache)
     result
   end
 
   def load_language_stats
-    version, commit_oid, stats = load_cache
-    if version == LANGUAGE_STATS_CACHE_VERSION && commit_oid && stats
-      [commit_oid, stats]
+    version, oid, stats = load_cache
+    if version == LANGUAGE_STATS_CACHE_VERSION && oid && stats
+      [oid, stats]
     end
   end
 
-  def save_language_stats(commit_oid, stats)
-    cache = [LANGUAGE_STATS_CACHE_VERSION, commit_oid, stats]
+  def save_language_stats(oid, stats)
+    cache = [LANGUAGE_STATS_CACHE_VERSION, oid, stats]
     write_cache(cache)
   end
 
@@ -64,25 +63,28 @@ class GitLinguist
   LANGUAGE_STATS_CACHE_VERSION = "v3:#{Linguist::VERSION}"
 
   def rugged
-    @rugged ||= Rugged::Repository.bare(repo_path)
+    @rugged ||= Rugged::Repository.bare(@repo_path)
   end
 
   def cache_file
-    File.join(repo_path, LANGUAGE_STATS_CACHE)
+    File.join(@repo_path, LANGUAGE_STATS_CACHE)
   end
 
   def write_cache(object)
-    tmp_path = Dir::Tmpname.make_tmpname(cache_file, nil)
+    return unless File.directory? @repo_path
 
+    begin
+      tmp_path = Dir::Tmpname.make_tmpname(cache_file, nil)
       File.open(tmp_path, "wb") do |f|
         marshal = Marshal.dump(object)
         f.write(Zlib::Deflate.deflate(marshal))
       end
 
       File.rename(tmp_path, cache_file)
-    tmp_path = nil
-  ensure
-    (File.unlink(tmp_path) rescue nil) if tmp_path
+    rescue => e
+      (File.unlink(tmp_path) rescue nil)
+      raise e
+    end
   end
 
   def load_cache
@@ -97,24 +99,18 @@ end
 def git_linguist(args)
   incremental = true
   commit = nil
-  git_dir = nil
 
   parser = OptionParser.new do |opts|
     opts.banner = "Usage: git-linguist [OPTIONS] stats|breakdown|dump-cache|clear|disable"
 
     opts.on("-f", "--force", "Force a full rescan") { incremental = false }
-    opts.on("--git-dir=DIR", "Path to the git repository") { |v| git_dir = v }
     opts.on("--commit=COMMIT", "Commit to index") { |v| commit = v}
   end
 
   parser.parse!(args)
 
-  git_dir ||= begin
-    pwd = Dir.pwd
-    dotgit = File.join(pwd, ".git")
-    File.directory?(dotgit) ? dotgit : pwd
-  end
-
+  git_dir = `git rev-parse --git-dir`.strip
+  raise "git-linguist must be ran in a Git repository" unless $?.success?
   wrapper = GitLinguist.new(git_dir, commit, incremental)
 
   case args.pop

github-linguist.gemspec

@@ -10,7 +10,7 @@ Gem::Specification.new do |s|
   s.homepage = "https://github.com/github/linguist"
   s.license  = "MIT"
 
-  s.files = Dir['lib/**/*'] - ['lib/linguist/grammars.rb']
+  s.files = Dir['lib/**/*'] - ['lib/linguist/grammars.rb'] + ['LICENSE']
   s.executables = ['linguist', 'git-linguist']
 
   s.add_dependency 'charlock_holmes', '~> 0.7.3'

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

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

lib/linguist/version.rb

@@ -1,3 +1,3 @@
 module Linguist
-  VERSION = "4.6.0"
+  VERSION = "4.6.3"
 end

test/test_pedantic.rb

@@ -9,7 +9,7 @@ class TestPedantic < Minitest::Test
     assert_sorted LANGUAGES.keys
   end
 
-  def test_extensions_are_sorted
+  def test_nonprimary_extensions_are_sorted
     LANGUAGES.each do |name, language|
       extensions = language['extensions']
       assert_sorted extensions[1..-1].map(&:downcase) if extensions && extensions.size > 1