Question

这纯粹是一个实验，但我想知道是否有可能通过某种元编程在运行时获得require'd gems的列表。例如，说我有：

require 'rubygems'
require 'sinatra'
require 'nokogiri'

# don't know what to do here

如何在运行时打印出以下内容？

this app needs rubygems, sinatra, nokogiri

Answer 1

你不能完全这样做，因为要求一个文件可能需要其他文件，并且Ruby无法区分你所需的文件和其他人所需的文件。

您可以查看$LOADED_FEATURES以获取所需内容的列表。但是，如果要明确指定依赖项，则应使用Bundler。

这是猜测宝石名称并列举所有内容的完全不完美的方法：

ruby-1.9.2-p180 :001 > $LOADED_FEATURES.
   select { |feature| feature.include? 'gems' }.
   map { |feature| File.dirname(feature) }.
   map { |feature| feature.split('/').last }.
   uniq.sort
 => ["1.9.1", "action_dispatch", "action_pack", "action_view", "actions", "active_model", "active_record", "active_support", "addressable", "agent", "array", "aws", "builder", "bundler", "cache_stores", "cancan", "cdn", "class", "client", "common", "compute", "connection", "control", "controllers", "core", "core_ext", "core_extensions", "css", "data_mapper", "decorators", "dependencies", "dependency_detection", "deprecation", "devise", "digest", "dns", "encodings", "encryptor", "engine", "errors", "excon", "ext", "failure", "faraday", "fields", "fog", "formatador", "geographer", "haml", "hash", "helpers", "heroku_san", "hmac", "hooks", "hoptoad_notifier", "html", "http", "i18n", "idna", "importers", "inflector", "initializers", "instrumentation", "integrations", "interpolate", "interval_skip_list", "jquery-rails", "json", "kaminari", "kernel", "lib", "mail", "metric_parser", "mime", "mixins", "model_adapters", "models", "module", "mongo_mapper", "mongoid", "multibyte", "new_relic", "node", "nokogiri", "numeric", "oauth", "object", "omniauth", "orm_adapter", "package", "parser", "parsers", "plugin", "pp", "providers", "queued", "rack", "rails", "railtie", "redis", "request", "request_proxy", "resp    ruby-1.9.2-p180 :008 >onse", "resque", "retriever_methods", "routing", "ruby_extensions", "ruby_flipper", "rubygems", "runtime", "samplers", "sass", "sax", "script", "scss", "selector", "sequel", "ses", "shell", "signature", "simple_geo", "state_machine", "stats_engine", "storage", "strategies", "string", "tar_reader", "template", "terremark", "thor", "tokens", "tree", "treetop", "twitter", "us", "util", "vendor", "version_specific", "visitors", "warden", "xml", "xml_mini", "xpath", "xslt"]

Answer 2

这是一种获取所有需要的调用的方法。创建此文件：show_requires.rb

alias :orig_require :require
def require s
  print "Requires #{s}\n" if orig_require(s)
end

然后使用

启动您的应用

ruby -r show_requires.rb myapp.rb

这会产生类似：

C:\code\test>ruby -r show_requires.rb test.rb
Requires stringio
Requires yaml/error
Requires syck
Requires yaml/ypath
Requires yaml/basenode
Requires yaml/syck
Requires yaml/tag
Requires yaml/stream
Requires yaml/constants
Requires date/format
Requires date
Requires yaml/rubytypes
Requires yaml/types
Requires yaml
Requires etc
Requires dl
Requires rbreadline
Requires readline

如果您只想要顶级要求，请添加全局以跟踪嵌套级别：

$_rq_lvl = 0
alias :orig_require :require
def require s
  $_rq_lvl+=1
  print "Requires #{s}\n" if orig_require(s) and $_rq_lvl == 1
  $_rq_lvl -=1
end

然后你得到：

C:\code\test>ruby -r require_test.rb test.rb
Requires yaml
Requires readline

Answer 3

只需稍加一点即可-考虑到为了精确地替换#require的行为，您还必须返回一个布尔值，因此这是一个更真实的覆盖：

module Kernel
  alias :orig_require :require
  def require(name)
    print "Requiring #{name}"
    is_okay = orig_require(name)
    puts " - #{is_okay ? 'Yup!' : 'Nope :('}"
    is_okay
  end
end

有趣的是，我正在做一些测试-跟踪需要模块时炸毁的东西链-然后这变得必要！

如何获取Ruby中已经“必需”的文件列表？

3 个答案: