lyber-core 0.9.6.2

data/lib/lyber_core.rb

@@ -0,0 +1,14 @@
+require 'dlss_service'
+require 'dor/suri_service'
+require 'dor/workflow_service'
+require 'dor/base'
+require 'lyber_core/connection'
+require 'lyber_core/destroyer'
+require 'lyber_core/log'
+require 'lyber_core/robots/robot'
+require 'lyber_core/robots/workflow'
+require 'lyber_core/robots/workspace'
+require 'lyber_core/robots/work_queue'
+require 'lyber_core/robots/work_item'
+require 'lyber_core/exceptions/empty_queue'
+

data/lib/lyber_core/connection.rb

@@ -0,0 +1,97 @@
+require 'net/https'
+require 'uri'
+require 'cgi'
+
+module LyberCore
+  class Connection
+    def Connection.get_https_connection(url)
+      LyberCore::Log.debug("Establishing connection to #{url.host} on port #{url.port}")
+      https = Net::HTTP.new(url.host, url.port)
+      if(url.scheme == 'https')
+        https.use_ssl = true
+        LyberCore::Log.debug("Using SSL")
+        https.cert = OpenSSL::X509::Certificate.new( File.read(LyberCore::CERT_FILE) )
+        LyberCore::Log.debug("Using cert file #{LyberCore::CERT_FILE}")
+        https.key = OpenSSL::PKey::RSA.new( File.read(LyberCore::KEY_FILE), LyberCore::KEY_PASS )
+        LyberCore::Log.debug("Using key file #{LyberCore::KEY_FILE} with pass #{LyberCore::KEY_PASS}")
+        https.verify_mode = OpenSSL::SSL::VERIFY_NONE
+        LyberCore::Log.debug("https.verify_mode = #{https.verify_mode} (should eql #{OpenSSL::SSL::VERIFY_NONE})")
+      end
+      https
+    end
+
+    # Returns body of the HTTP response, or passes the response to the block if it's passed in
+    #
+    # == Required Parameters
+    # - <b>full_url</b> - A string containing the full url to the resource you're trying to connect to
+    # - <b>method</b> - Recognizes the following symbols which correspond to an HTTP verb.  The convenience methods take care of this
+    #  :get for HTTP GET
+    #  :post for HTTP POST
+    #  :put for HTTP PUT
+    # - <b>body</b> The body of your request.  Can be nil if you don't have one.
+    #
+    # == Options
+    # - <b>:auth_user</b> for basic HTTP authentication.  :auth_user and :auth_password must both be set if using authentication
+    # - <b>:auth_password</b> for basic HTTP authentication. :auth_user and :auth_password must both be set if using authentication
+    # - <b>:content_type</b> if not passed in as an option, then it is set to 'application/xml'
+    #
+    # == Block
+    # By default, this method returns the body of the response, Net::HTTPResponse.body .  If you want to work with the Net::HTTPResponse
+    # object, you can pass in a block, and the response will be passed to it.
+    # 
+    # == Exceptions
+    # Any exceptions thrown while trying to connect should be handled by the caller
+    def Connection.connect(full_url, method, body, options = {}, &block)
+      url = URI.parse(full_url)
+      case method
+      when :get
+        req = Net::HTTP::Get.new(url.request_uri)
+      when :post
+        req = Net::HTTP::Post.new(url.request_uri)
+      when :put
+        req = Net::HTTP::Put.new(url.request_uri)
+      end
+      req.body = body unless(body.nil?)
+      if(options.include?(:content_type))
+        req.content_type = options[:content_type]
+      else
+        req.content_type = 'application/xml'
+      end
+
+      if(options.include?(:auth_user))
+        req.basic_auth options[:auth_user], options[:auth_password]
+      end
+
+      res = Connection.get_https_connection(url).start {|http| http.request(req) }
+      case res
+      when Net::HTTPSuccess
+        if(block_given?)
+          block.call(res)
+        else
+          return res.body
+        end
+      else
+        raise res.error!
+      end
+
+    end
+  end
+  
+  
+  # Convenience method for performing an HTTP GET using Connection.connect
+  def Connection.get(full_url, options = {}, &b)
+    Connection.connect(full_url, :get, nil, options, &b)
+  end
+
+  # Convenience method for performing an HTTP POST using Connection.connect
+  def Connection.post(full_url, body, options = {}, &b)
+    Connection.connect(full_url, :post, body, options, &b)
+  end
+
+  # Convenience method for performing an HTTP PUT using Connection.connect
+  def Connection.put(full_url, body, options = {}, &b)
+    Connection.connect(full_url, :put, body, options, &b)
+  end
+
+
+end

data/lib/lyber_core/destroyer.rb

@@ -0,0 +1,74 @@
+module LyberCore
+    
+  class LyberCore::Destroyer
+    
+  require 'rubygems'
+  require 'active-fedora'
+  require 'open-uri'
+  
+  attr_reader :repository
+  attr_reader :workflow
+  attr_reader :registration_robot
+  attr_reader :druid_list
+  attr_reader :current_druid
+  
+  # Given a repository, a workflow and the name of a robot that registers objects in fedora
+  # we can generate a list of all the druids that were created by that robot
+  def initialize(repository, workflow, registration_robot)
+    @repository = repository
+    @workflow = workflow
+    @registration_robot = registration_robot
+    @druid_list = self.get_druid_list
+  end
+  
+  def get_druid_list
+    begin
+      druid_list = []
+      url_string = "#{WORKFLOW_URI}/workflow_queue?repository=#{@repository}&workflow=#{@workflow}&completed=#{@registration_robot}"
+      LyberCore::Log.info("Fetching druids from #{url_string}")
+      doc = Nokogiri::XML(open(url_string))
+      doc.xpath("//objects/object/@id").each do |id|
+        druid_list << id.to_s
+      end
+      return druid_list
+    rescue Exception => e
+      raise e, "Couldn't fetch druid list from #{url_string}"
+    end
+  end
+    
+    # Iterate through the druids in @druid_list and delete each of them from FEDORA
+    def delete_druids
+      begin
+        connect_to_fedora
+        @druid_list.each do |druid|
+          @current_druid = druid
+          LyberCore::Log.info("Deleting #{@current_druid}")
+          begin
+            obj = ActiveFedora::Base.load_instance(@current_druid)
+            obj.delete
+          rescue ActiveFedora::ObjectNotFoundError
+            LyberCore::Log.info("#{@current_druid} does not exist in this repository")
+          end
+        end
+      rescue Exception => e
+        raise e
+      end
+    end
+    
+    def connect_to_fedora
+      begin
+        if @repository == "dor"
+          repo_url = FEDORA_URI
+        elsif @repository == "sdr"
+          repo_url = SEDORA_URI
+        else
+          raise "I don't know how to connect to repository #{@repository}"
+        end
+        LyberCore::Log.info("connecting to #{repo_url}")
+        Fedora::Repository.register(repo_url)
+      rescue Errno::ECONNREFUSED => e
+       raise e, "Can't connect to Fedora at url #{repo_url}"
+      end
+    end
+  end
+end

data/lib/lyber_core/exceptions/empty_queue.rb

@@ -0,0 +1,9 @@
+require 'dor_service'
+
+module LyberCore
+  module Exceptions
+    class EmptyQueue < RuntimeError
+      
+    end
+  end
+end

data/lib/lyber_core/log.rb

@@ -0,0 +1,105 @@
+
+module LyberCore
+  
+  class LyberCore::Log
+    require 'logger'
+        
+    # Default values
+    DEFAULT_LOGFILE = "/tmp/lybercore_log.log"
+    DEFAULT_LOG_LEVEL = Logger::INFO
+    DEFAULT_FORMATTER = proc{|s,t,p,m|"%5s [%s] (%s) %s :: %s\n" % [s, 
+                       t.strftime("%Y-%m-%d %H:%M:%S"), $$, p, m]}
+        
+    # Initial state
+    @@logfile = DEFAULT_LOGFILE
+    @@log ||= Logger.new(@@logfile)
+    # $stdout.reopen(@@logfile)
+    # $stderr.reopen(@@logfile)
+    @@log.level = DEFAULT_LOG_LEVEL
+    @@log.formatter = DEFAULT_FORMATTER
+    
+    # Restore LyberCore::Log to its default state
+    def Log.restore_defaults
+      @@log.level = DEFAULT_LOG_LEVEL
+      Log.set_logfile(DEFAULT_LOGFILE)
+      @@log.formatter = DEFAULT_FORMATTER
+    end
+      
+    # The current location of the logfile
+    def Log.logfile
+      return @@logfile
+    end
+    
+
+    
+    # Accepts a filename as an argument, and checks to see whether that file can be 
+    # opened for writing. If it can be opened, it closes the existing Logger object
+    # and re-opens it with the new logfile location. It raises an exception if it
+    # cannot write to the specified logfile. 
+    def Log.set_logfile(new_logfile)
+      begin
+        current_log_level = @@log.level
+        current_formatter = @@log.formatter
+        @@log = Logger.new(new_logfile)
+        @@logfile = new_logfile
+        @@log.level = current_log_level
+        @@log.formatter = current_formatter
+      rescue Exception => e
+        raise e, "Couldn't initialize logfile #{new_logfile}: #{e.backtrace}"
+      end
+      
+    end
+    
+    # Set the log level. 
+    # See http://ruby-doc.org/core/classes/Logger.html for more info. 
+    # Possible values are: 
+    #   Logger::FATAL (4):  an unhandleable error that results in a program crash
+    #   Logger::ERROR (3):  a handleable error condition
+    #   Logger::WARN (2): a warning
+    #   Logger::INFO (1): generic (useful) information about system operation
+    #   Logger::DEBUG (0):  low-level information for developers
+    def Log.set_level(loglevel)
+     begin
+        if [0,1,2,3,4].contains? loglevel
+          @@log.level = loglevel
+          @@log.debug "Setting LyberCore::Log.level to #{loglevel}"
+        else
+          @@log.warn "I received an invalid option for log level. I expected a number between 0 and 4 but I got #{loglevel}"
+          @@log.warn "I'm setting the loglevel to 0 (debug) because you seem to be having trouble."
+          @@log.level = 0
+        end
+      rescue Exception => e
+        raise e, "Couldn't set log level: #{e.backtrace}"
+      end
+    end
+    
+    # Return the current log level
+    def Log.level
+      @@log.level
+    end
+    
+    def Log.fatal(msg)
+      @@log.add(Logger::FATAL) { msg }
+    end
+    
+    def Log.error(msg)
+      @@log.add(Logger::ERROR) { msg }
+    end
+    
+    def Log.warn(msg)
+      @@log.add(Logger::WARN) { msg }
+    end
+    
+    def Log.info(msg)
+      @@log.add(Logger::INFO) { msg }
+    end
+    
+    def Log.debug(msg)
+      @@log.add(Logger::DEBUG) { msg }
+    end
+    
+    
+  end
+  
+  
+end

data/lib/lyber_core/rake/dlss_release.rb

@@ -0,0 +1,126 @@
+require 'rake/tasklib'
+require 'pony'
+
+module LyberCore
+
+  # A rake task that will tag, build and publish your gem to the DLSS gemserver
+  # 
+  # == Usage
+  # Include the following two lines in your <tt>Rakefile</tt>:
+  #   require 'lyber_core/rake/dlss_release'
+  #   LyberCore::DlssRelease.new
+  # 
+  # To build and release the gem, run the following AFTER pushing your code to the git repository on AFS:
+  #   rake dlss_release
+  #
+  # == Requirements
+  # - You need <b>ONE</b> <tt>your-gemname.gemspec</tt> file in the root of your project. See the example in <tt>lyber-core.gemspec</tt> or look at http://yehudakatz.com/2010/04/02/using-gemspecs-as-intended/
+  # - Inside the <tt>Gem::Specification</tt> block of the <tt>your-gemname.gemspec</tt> file, you must set the version number.  For example
+  #     Gem::Specification.new do |s|
+  #       s.name        = "your-gemname"
+  #       ...
+  #       s.version     = "0.9.6"
+  #       ...
+  #     end
+  # - You need to have access to the account on the DLSS gemserver, usually by adding your sunetid to the .k5login.  You also need git installed.
+  class DlssRelease < Rake::TaskLib
+    # Name of the rake task. Defaults to :dlss_release
+    attr_accessor :name
+    
+    # Name of the gem
+    attr_accessor :gemname
+    
+    # Version of the gem being released
+    attr_accessor :version
+    
+    def initialize(name=:dlss_release)
+      @name = name
+      @gemname = determine_gemname
+      yield self if block_given?
+      raise "Gemname must be set" if @gemname.nil?
+      define
+    end
+    
+    # Tries to parse the gemname from the prefix of the <tt>your-gemname.gemspec</tt> file.
+    def determine_gemname
+      files = Dir.glob("*.gemspec")
+      raise "Unable to find project gemspec file.\nEither it doesn't exist or there are too many files ending in .gemspec" if(files.size != 1)
+      
+      files.first =~ /(.*)\.gemspec/
+      $1
+    end
+    
+    def send_release_announcement
+      Pony.mail(:to => 'dlss-developers@lists.stanford.edu', 
+                :from => 'dlss-developers@lists.stanford.edu', 
+                :subject => "Release #{@version} of the #{@gemname} gem", 
+                :body => "The #{@gemname}-#{@version}.gem has been released to the DLSS gemserver", 
+                :via => :smtp, 
+                :via_options => {
+                  :address        => 'smtp.stanford.edu',
+                  :port           => '25',
+                  :enable_starttls_auto => true,
+                  :authentication => :plain, # :plain, :login, :cram_md5, no auth by default
+                  :domain         => "localhost.localdomain" # the HELO domain provided by the client to the server
+                }
+      )
+    end
+    
+    def define
+      desc "Tag, build, and release DLSS specified gem"
+      task @name do
+        IO.foreach(File.join(Rake.original_dir, "#{@gemname}.gemspec")) do |line|
+          if(line =~ /\.version.*=.*"(.*)"/)
+            @version = $1 
+            break
+          end
+        end
+
+        if(@version.nil?)
+          raise "Unable to find version number in #{@gemname}.gemspec"
+        end
+        created_gem = "#{@gemname}-#{@version}.gem"
+
+        puts "Make sure:"
+        puts "  1) Version #{@version} of #{@gemname} has not been tagged and released previously"
+        puts "  2) All of the tests pass"
+        puts "  3) You've pushed the code to the git repository on AFS (usually 'git push origin master')"
+        puts "Type 'yes' to continue if all of these statements are true"
+
+        resp = STDIN.gets.chomp
+        unless(resp =~ /yes/ )
+          raise "Please complete the prerequisites"
+        end
+
+        puts "Releasing version #{@version} of the #{@gemname} gem"
+        
+        begin
+          tags = `git tag`.split("\n")
+          raise "tag v#{@version} already exists. Change the version number in the #{@gemname}.gemspec file" if tags.include?("v#{@version}")
+          
+          puts "...Tagging release"
+          sh "git tag -a v#{@version} -m 'Gem version #{@version}'"
+          sh "git push origin --tags"
+
+          puts "...Building gem"
+          sh "gem build #{@gemname}.gemspec"
+
+          puts "...Publishing gem to sulair-rails-dev DLSS gemserver" 
+          sh "scp #{created_gem} webteam@sulair-rails-dev.stanford.edu:/var/www/html/gems"
+          sh "ssh webteam@sulair-rails-dev.stanford.edu gem generate_index -d /var/www/html"
+
+          puts "Done!!!!!  A local copy of the gem is in the pkg directory"
+          FileUtils.mkdir("pkg") unless File.exists?("pkg")
+          FileUtils.mv("#{created_gem}", "pkg")
+        rescue Exception => e
+          FileUtils.rm("#{created_gem}") if(File.exists?("#{created_gem}"))
+          raise e
+        end
+      
+        puts "Sending release announcement to DLSS developers list"
+        send_release_announcement
+        
+      end
+    end
+  end
+end

data/lib/lyber_core/robots/robot.rb

@@ -0,0 +1,214 @@
+# == Usage 
+#   ruby_cl_skeleton [options] source_file
+#
+#   For help use: ruby_cl_skeleton -h
+
+module LyberCore
+  module Robots
+    require 'optparse'
+    require 'ostruct'
+
+    # ===== Usage
+    # User defined robots should derive from this class and override the #process_item method
+    class Robot
+      attr_accessor :workflow_name
+      attr_accessor :workflow_step
+      
+      # A LyberCore::Robots::Workflow object
+      attr_accessor :workflow 
+      attr_accessor :collection_name
+      attr_accessor :workspace
+      attr_accessor :args
+      attr_accessor :options
+
+
+      # Available options
+      # - :collection_name - The collection this workflow should work with.  
+      #   Defined as a subdirectory within ROBOT_ROOT/config/workflows/your_workflow/your_collection
+      # - :workspace - Full path of where to find content for a particular workflow
+      # - :logfile - Where to write log messages
+      # - :loglevel - Level of logging from 0 - 4 where 0 = DEBUG and 4 = FATAL
+      def initialize(workflow_name, workflow_step, args = {})
+        @workflow_name = workflow_name
+        @workflow_step = workflow_step
+        @collection_name = args[:collection_name]      
+        @opts = args
+
+        if args[:logfile]
+          LyberCore::Log.set_logfile(args[:logfile])
+        else
+          FileUtils.mkdir(File.join(ROBOT_ROOT, 'log')) unless(File.exists?(File.join(ROBOT_ROOT, 'log')))
+          robot_logfile = File.join(ROBOT_ROOT,'log',workflow_step+'.log')
+          LyberCore::Log.set_logfile(robot_logfile)
+        end
+
+        LyberCore::Log.set_level(args[:loglevel]) if args[:loglevel] 
+      
+        # Set defaults
+        @options = OpenStruct.new
+        self.parse_options
+        self.create_workflow
+        self.set_workspace
+      end
+      
+      # Some workflows require a directory where their content lives
+      # If a robot is invoked with a :workspace => true option, its @workspace
+      # should be set from the value in 
+      def set_workspace
+        if(@opts[:workspace])
+          @workspace = LyberCore::Robots::Workspace.new(@workflow_name, @collection_name)
+          LyberCore::Log.debug("workspace = #{workspace.inspect}")
+        end
+      end
+      
+      # Create the workflow at instantiation, not when we start running the robot.
+      # That way we can do better error checking and ensure that everything is going
+      # to run okay before we actually start things.
+      def create_workflow
+        
+        unless defined?(WORKFLOW_URI)
+          LyberCore::Log.fatal "FATAL: WORKFLOW_URI is not defined"
+          LyberCore::Log.fatal "Usually this is a value like 'http://lyberservices-dev.stanford.edu/workflow'"
+          LyberCore::Log.fatal "Usually you load it by setting ROBOT_ENVIRONMENT when you invoke your robot"
+          raise "WORKFLOW_URI is not set! Do you need to set your ROBOT_ENVIRONMENT value?"
+        end
+        LyberCore::Log.debug("About to instatiate a Workflow object
+            -- LyberCore::Robots::Workflow.new(#{@workflow_name},#{collection_name}")
+        @workflow = LyberCore::Robots::Workflow.new(@workflow_name, {:logger => @logger, :collection_name => @collection_name})
+        
+      end
+    
+      # == Create a new workflow 
+      def start()
+        
+        begin
+          LyberCore::Log.debug("Starting robot...")
+
+          queue = @workflow.queue(@workflow_step)
+      
+          # If we have arguments, parse out the parts that indicate druids
+          if(@options.file or @options.druid)
+            queue.enqueue_druids(get_druid_list)
+          else
+            queue.enqueue_workstep_waiting()
+          end
+          process_queue(queue)
+        # TODO: Implement a FatalError class
+        # rescue LyberCore::Exceptions::FatalError => e
+        #   LyberCore::Log.fatal("e.msg")
+        #   exit
+        rescue LyberCore::Exceptions::EmptyQueue
+          LyberCore::Log.info("Empty queue -- no objects to process")
+        rescue Exception => e
+          LyberCore::Log.error(e.msg)
+          LyberCore::Log.error(e.backtrace.join("\n"))
+        end
+      end
+
+      # Generate a list of druids to process
+      def get_druid_list
+      
+        druid_list = Array.new
+        
+        # append any druids passed explicitly
+        if(@options.druid)
+          druid_list << @options.druid
+        end
+      
+        # identifier list is in a file
+         if (@options.file && File.exist?(@options.file))
+          File.open(@options.file) do |file|
+            file.each_line do |line|
+              druid = line.strip
+              if (druid.length > 0)
+                druid_list << druid
+              end
+            end
+          end
+        end
+      
+        return druid_list
+      end
+
+      def process_queue(queue)
+        while work_item = queue.next_item do
+          begin
+            #call overridden method
+            process_item(work_item)
+            work_item.set_success
+          rescue Exception => e
+            # LyberCore::Log.error("Encountered exception processing #{work_item.druid}: #{e.to_s}")
+            # LyberCore::Log.debug("Encountered exception processing #{work_item.druid}: #{e.backtrace.join("\n")}")
+            work_item.set_error(e)
+          end
+        end
+        queue.print_stats()
+      end
+    
+      # Override this method in your robot instance.  The method in this base class will throw an exception if it is not overriden.
+      def process_item(work_item)
+        #to be overridden by child classes
+        raise 'You must implement this method in your subclass'
+      end 
+      
+    # ###########################
+    # command line option parsing 
+  
+        def parse_options
+        
+          options = {}
+
+          o = OptionParser.new do |opts|
+            opts.banner = "Usage: example.rb [options]"
+            opts.separator ""
+
+            opts.on("-d DRUID", "--druid DRUID", "Pass in a druid to process") do |d|
+              @options.druid = d
+            end
+          
+            opts.on("-f", "--file FILE", "Pass in a file of druids to process") do |f|
+              @options.file = f
+            end
+          
+          end
+        
+          # Parse the command line options and ignore anything not specified above
+          begin
+            o.parse!
+          rescue OptionParser::InvalidOption => e
+            LyberCore::Log.debug("e.inspect")
+          rescue OptionParser::ParseError => e
+            LyberCore::Log.error("Couldn't parse options: #{e.backtrace}") 
+            raise e
+          end
+        
+        end
+
+        # def output_options
+        #   puts "Options:\n"
+        # 
+        #   @options.marshal_dump.each do |name, val|        
+        #     puts "  #{name} = #{val}"
+        #   end
+        # end
+        # 
+        # def output_help
+        #   output_version
+        #   RDoc::usage() #exits app
+        # end
+        # 
+        # def output_usage
+        #   RDoc::usage('usage') # gets usage from comments above
+        # end
+        # 
+        # def output_version
+        #   puts "#{File.basename(__FILE__)} version #{VERSION}"
+        # end
+  
+    # ##################################
+    # end of command line option parsing 
+    # ##################################
+  
+    end # end of class
+  end # end of Robots module
+end # end of LyberCore module