Sha256: 4a52a47103896643276e3440f6315016af653d326194b3a2ca777a52f047263b
Contents?: true
Size: 1.5 KB
Versions: 1
Compression:
Stored size: 1.5 KB
Contents
# to test in 1.8.x, make sure to use ruby specname.rb require File.dirname(__FILE__) + "/../lib/faster_rubygems" require 'sane' require 'spec' # rspec 1 require 'spec/autorun' require 'fileutils' raise if ENV['RUBYOPT'] describe Gem do before do ENV['GEM_PATH'] = 'test_dir' @gem_path = 'test_dir/gems/' # wow. FileUtils.rm_rf @gem_path FileUtils.mkdir_p @gem_path end context "speeding Gem.bin_path by inferring" do it "should fake guess the right path instead of loading full rubygems for now" do assert Gem.bin_path('after', 'after', ">= 0") =~ /after.*bin.*after/ # should not have loaded full gems... # if this line fails then make sure you are running the spec file like ruby xxx not spec xxx assert !defined?(Gem::Dependency) end end it "should pass test files" do Dir['files/*'].each{|f| raise f unless system(OS.ruby_bin + " #{f}") } end context "gem xxx" do it "should allow you to load 'older' gem versions somehow, like maybe cacheing" # lower prio end it "in 1.9 if you pass it --disable-gems, it should load its full gem_prelude right there...I guess...though that's not going to fix any bottlenecks though..." it "should make a directory cacheing list (yes!) in the root of each gem and load that, and use that to activate. Oh baby. Maybe, though that might fall more under the faster_require aspect of it...hmm...maybe it's fast enough for now, except rails, which needs faster_require no matter what anyway..." end
Version data entries
1 entries across 1 versions & 1 rubygems
Version | Path |
---|---|
faster_rubygems-0.9.2 | spec/spec.faster_rubygems.rb |