#!/usr/bin/env ruby # encoding: UTF-8 # (c) Copyright 2014 Hewlett-Packard Development Company, L.P. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # rubocop: disable Metrics/AbcSize # Module Lorj which contains several classes. # # Those classes describes : # - processes (BaseProcess) : How to create/delete/edit/query object. # - controler (BaseControler) : If a provider is defined, define how will do # object creation/etc... # - definition(BaseDefinition): Functions to declare objects, query/data mapping # and setup # this task to make it to work. module Lorj # This is the main lorj class. # It interfaces your main code with the full lorj system as shown in the # concept document. # It gives you access to the lorj model object designed by your process. # # When you start using it, your main must be as simple as you can, as you will # need to move most of your application logic to the process. # Your application can have several lorj objects running in your code, # depending of your needs. # # The main things is that you can move most of your process management, # usually in your code/modules to be part of the lorj process, make it # controller independant, and gains in implementing several controllers to # change the way to implement but not the process, you used to build your # application! # # Then, your application contributors can build their own controller and # extend your solution! # # Here an example of creating a CloudServer, using Lorj_cloud module # See http://github.com/forj-oss/lorj-cloud # # require 'lorj_cloud' # # process = { :process_module => 'cloud', # :controller_name => 'openstack' } # cloud = Lorj::Core.new(nil, [process]) # cloud.setup(:server) # Ask end user about required data # cloud.create(:server, :server_name => 'myservername') # # Another basic example. # # require 'lorj' # # process = { :process_name => 'myprocess', # :process_path => '', # :controller_name => :mock } # # core = Lorj::Core.new(nil, [process]) # core.create(:myobject, :parameter1 => "value") # # You can learn on lorj by example, see examples/ # # See BaseProcess to check how you can write a process and what kind of # functions are available for your process to be kept controller independant. # # See BaseController to see how you can write a controller and what kind of # functions are available to deal with the implementation API you need to use. class Core # Public access to a config object. # A config object can be any kind of class which should provide at least # following functions: # # - get(*key, default=nil) and [*key] : function to get a value from a key. # default is a value to get if not found. # # - set(*key, value) or [*key, value]= : function to set a value to a key. # Ex: From processes, you can set a runtime data with: # # config.set(key, value) # # OR # # config[key] = value # # - exist?(*key) : function which return false if not # found, or any other value if found. # Ex: From processes, you can get a data (runtime/account/config.yaml or # defaults.yaml) with: # # config.get(key) # # OR # # config[key] # # For each functions, *key is a list of value, which becomes an array in the # function. # It should accept to manage the key tree (hash of hashes) # # Currently lorj comes with Lorj::Config or Lorj::Account. # Thoses classes defines at least those 5 functions. And more. attr_reader :config # a wrapper to Create call. Use this function for code readibility. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def connect(oCloudObj, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_create(oCloudObj, hConfig) end # Execute the creation process to create the object `oCloudObj`. # The creation process can add any kind of complexity to # get the a memory representation of the object manipulated during creation # process. # This means that a creation process can be (non exhaustive list of # possibilities) # - a connection initialization # - an internal memory data structure, like hash, array, ruby object... # - a get or create logic # - ... # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def create(oCloudObj, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_create(oCloudObj, hConfig) end # a wrapper to Create call. Use this function for code readibility. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def get_or_create(oCloudObj, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_create(oCloudObj, hConfig) end # Execution of the delete process for the `oCloudObj` object. # It requires the object to be loaded in lorj Lorj::Data objects cache. # You can use `Create` or `Get` functions to load this object. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def delete(oCloudObj, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_delete(oCloudObj, hConfig) end # Execution of the Query process for the `oCloudObj` object. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +sQuery+ : Hash representing the query filter. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def query(oCloudObj, sQuery, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_query(oCloudObj, sQuery, hConfig) end # Execution of the Get process for the `oCloudObj` object. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +sId+ : data representing the ID (attribute :id) of a Lorj::Data # object. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def get(oCloudObj, sId, hConfig = nil) return nil if !oCloudObj || !@core_object || sId.nil? @core_object.process_get(oCloudObj, sId, hConfig) end # Execution of the Update process for the `oCloudObj` object. # Usually, the Controller object data is updated by the process # (BaseController::set_attr) # then it should call a controller_update to really update the data in the # controller. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +sId+ : data representing the ID (attribute :id) of a Lorj::Data # object. # - +hConfig+ : Hash of hashes containing data required to initialize # the object. # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def update(oCloudObj, hConfig = nil) return nil if !oCloudObj || !@core_object @core_object.process_update(oCloudObj, hConfig) end # Function used to ask users about setting up his account. # # * *Args* : # - +oCloudObj+ : Name of the object to initialize. # - +sAccountName+ : Obsolete. You have to load the account data before # # If you use this variable, any other runtime config defined # by the Data model will be cleaned before # # * *Returns* : # - +Lorj::Data+ : Represents the Object initialized. # # * *Raises* : # No exceptions def setup(oCloudObj, _sAccountName = nil) return nil if !oCloudObj || !@core_object @core_object.process_setup(oCloudObj) end # Function to add an object to Lorj cache. # # This function is typically used when a previous query has been # executed and you want to keep the object in lorj cache. # Lorj cache is used by create/delete/etc... To avoid requerying or # creating those objects (+dependencies) if missing. # # * *Args* : # - +oObject+ : object to register. # - +sObjectType+ : Object type to register. # By default, the object type is determined by the Object itself # if this object is an ObjectData. Otherwise, you need to define the # object type. # # * *Raises* : # No exceptions def register(oObject, sObjectType = nil) #:doc: return nil if !oObject || !@core_object @core_object.register(oObject, sObjectType) end # Core parameters are: # the_config : Optional. An instance of a configuration system which *HAVE* # to provide get/set/exist?/[]/[]= # # * *Args*: # - +Processes+: Array of processes with controller # This array, contains a list of process to load and optionnaly a # controller. # # You can define your own process or a process module. # The array is structured as follow: # - each element contains a Hash with: # If you are using a process module, set the following: # - :process_module : Name of the process module to load # # If you are not using a Process module, you need to set the following: # - :process_path : Path to a local process code. # This path must contains at least 'process' subdir. And if needed # a 'controllers' path # - :process_name : Name of the local process # # Optionnally, you can set a controller name to use with the process. # - :controller_name: Name of the controller to use. # - :controller_path: Path to the controller file. # def initialize(the_config = nil, processes = nil, controller_class = nil) # Loading ProcessClass # Create Process derived from respectively BaseProcess PrcLib.core_level = 0 if PrcLib.core_level.nil? init_core_config(the_config) PrcLib.model.config = @config model = initialize_model # Compatibility with old 'new syntax' # `processes` will get an Array of string/symbol or simply a string/symbol # `controller_class` is used to define the controller to load. # string/symbol processes = adapt_core_parameters(processes, controller_class) # Load Application processes init_processes(model, processes) PrcLib.runtime_fail 'Lorj::Core: No valid process loaded. '\ 'Aborting.' if model[:process_class].nil? # Create Core object with the application model loaded # (processes & controller) initialize_core_object(model) PrcLib.model.clear_heap end private # This function is used to keep compatibility with old way to load # processes and controllers # If processes is a Array of Hash => new way # otherwise we need to create it. def adapt_core_parameters(processes, controller) return [] unless processes.is_a?(Array) return [] if processes.length == 0 return processes if processes[0].is_a?(Hash) PrcLib.warning('lorj initialization with Process & controller parameters'\ ' is obsolete. Read Lorj::Core.new to update it and '\ 'eliminate this warning') # Declare processes processes = processes_as_array(processes) processes_built = [] processes.each do |process| process = process.to_s if process.is_a?(Symbol) a_process = {} if process.include?('/') a_process[:process_path] = process a_process[:process_name] = File.basename(process) else a_process[:process_module] = process end processes_built << a_process end _adapt_with_controller(processes_built, controller) processes_built end def _adapt_with_controller(processes_built, controller) return if controller.nil? if controller.include?('/') processes_built[-1][:controller_path] = controller processes_built[-1][:controller_name] = File.basename(controller) else processes_built[-1][:controller_name] = controller end end end module_function # Any Lorj process module will need to declare itself to Lorj # with this function. # # * *args* : # - +process_name+: name of the process declared to Lorj. This name must be # unique. Otherwise the declaration won't happen. # # - +path+ : Path where process dir structure are located. # at least, it expects to find the process/.rb # Each controllers found will be added as well. # It must be controllers//.rb # You can change 'controllers' by any name, with :controllers_dir # # - +properties : Optional. # - :controllers_dir : Name of the controllers directory. # By default 'controllers' # # The process will be added in Lorj.processes Hash # def declare_process(process_name, path, properties = {}) process_data = Lorj::ProcessResource.new(process_name, path, properties) return nil if process_data.nil? @processes = {} if @processes.nil? return nil if process_data.process.nil? process_name = process_data.name @processes[process_name] = process_data unless @processes.key?(process_name) process_data end # Define module data for lorj library configuration class << self attr_reader :processes end end