ebeigarts-thinking-sphinx 1.1.21

data/lib/thinking_sphinx/collection.rb

@@ -0,0 +1,148 @@
+module ThinkingSphinx
+  class Collection < ::Array
+    attr_reader :total_entries, :total_pages, :current_page, :per_page
+    attr_accessor :results
+
+    # Compatibility with older versions of will_paginate
+    alias_method :page_count, :total_pages
+    
+    def initialize(page, per_page, entries, total_entries)
+      @current_page, @per_page, @total_entries = page, per_page, total_entries
+      
+      @total_pages = (entries / @per_page.to_f).ceil
+    end
+    
+    def self.ids_from_results(results, page, limit, options)
+      collection = self.new(page, limit,
+        results[:total] || 0, results[:total_found] || 0
+      )
+      collection.results = results
+      collection.replace results[:matches].collect { |match|
+        match[:attributes]["sphinx_internal_id"]
+      }
+      return collection
+    end
+    
+    def self.create_from_results(results, page, limit, options)
+      collection = self.new(page, limit,
+        results[:total] || 0, results[:total_found] || 0
+      )
+      collection.results = results
+      collection.replace instances_from_matches(results[:matches], options)
+      return collection
+    end
+    
+    def self.instances_from_matches(matches, options = {})
+      if klass = options[:class]
+        instances_from_class klass, matches, options
+      else
+        instances_from_classes matches, options
+      end
+    end
+    
+    def self.instances_from_class(klass, matches, options = {})
+      index_options = klass.sphinx_index_options
+
+      ids = matches.collect { |match| match[:attributes]["sphinx_internal_id"] }
+      instances = ids.length > 0 ? klass.find(
+        :all,
+        :joins      => options[:joins],
+        :conditions => {klass.primary_key.to_sym => ids},
+        :include    => (options[:include] || index_options[:include]),
+        :select     => (options[:select]  || index_options[:select]),
+        :order      => (options[:sql_order] || index_options[:sql_order])
+      ) : []
+
+      # Raise an exception if we find records in Sphinx but not in the DB, so
+      # the search method can retry without them. See 
+      # ThinkingSphinx::Search.retry_search_on_stale_index.
+      if options[:raise_on_stale] && instances.length < ids.length
+        stale_ids = ids - instances.map {|i| i.id }
+        raise StaleIdsException, stale_ids
+      end
+
+      # if the user has specified an SQL order, return the collection
+      # without rearranging it into the Sphinx order
+      return instances if options[:sql_order]
+
+      ids.collect { |obj_id|
+        instances.detect { |obj| obj.id == obj_id }
+      }
+    end
+    
+    # Group results by class and call #find(:all) once for each group to reduce
+    # the number of #find's in multi-model searches.
+    # 
+    def self.instances_from_classes(matches, options = {})
+      groups = matches.group_by { |match| match[:attributes]["class_crc"] }
+      groups.each do |crc, group|
+        group.replace(
+          instances_from_class(class_from_crc(crc), group, options)
+        )
+      end
+      
+      matches.collect do |match|
+        groups.detect { |crc, group|
+          crc == match[:attributes]["class_crc"]
+        }[1].detect { |obj|
+          obj.id == match[:attributes]["sphinx_internal_id"]
+        }
+      end
+    end
+    
+    def self.class_from_crc(crc)
+      @@models_by_crc ||= ThinkingSphinx.indexed_models.inject({}) do |hash, model|
+        hash[model.constantize.to_crc32] = model
+        model.constantize.subclasses.each { |subclass|
+          hash[subclass.to_crc32] = subclass.name
+        }
+        hash
+      end
+      @@models_by_crc[crc].constantize
+    end
+    
+    def previous_page
+      current_page > 1 ? (current_page - 1) : nil
+    end
+    
+    def next_page
+      current_page < total_pages ? (current_page + 1): nil
+    end
+    
+    def offset
+      (current_page - 1) * @per_page
+    end
+    
+    def method_missing(method, *args, &block)
+      super unless method.to_s[/^each_with_.*/]
+      
+      each_with_attribute method.to_s.gsub(/^each_with_/, ''), &block
+    end
+    
+    def each_with_groupby_and_count(&block)
+      results[:matches].each_with_index do |match, index|
+        yield self[index], match[:attributes]["@groupby"], match[:attributes]["@count"]
+      end
+    end
+    
+    def each_with_attribute(attribute, &block)
+      results[:matches].each_with_index do |match, index|
+        yield self[index], (match[:attributes][attribute] || match[:attributes]["@#{attribute}"])
+      end
+    end
+    
+    def each_with_weighting(&block)
+      results[:matches].each_with_index do |match, index|
+        yield self[index], match[:weight]
+      end
+    end
+    
+    def inject_with_groupby_and_count(initial = nil, &block)
+      index = -1
+      results[:matches].inject(initial) do |memo, match|
+        index += 1
+        yield memo, self[index], match[:attributes]["@groupby"], match[:attributes]["@count"]
+      end
+    end
+  end
+end

data/lib/thinking_sphinx/configuration.rb

@@ -0,0 +1,275 @@
+require 'erb'
+require 'singleton'
+
+module ThinkingSphinx
+  # This class both keeps track of the configuration settings for Sphinx and
+  # also generates the resulting file for Sphinx to use.
+  # 
+  # Here are the default settings, relative to RAILS_ROOT where relevant:
+  #
+  # config file::           config/#{environment}.sphinx.conf
+  # searchd log file::      log/searchd.log
+  # query log file::        log/searchd.query.log
+  # pid file::              log/searchd.#{environment}.pid
+  # searchd files::         db/sphinx/#{environment}/
+  # address::               127.0.0.1
+  # port::                  3312
+  # allow star::            false
+  # min prefix length::     1
+  # min infix length::      1
+  # mem limit::             64M
+  # max matches::           1000
+  # morphology::            nil
+  # charset type::          utf-8
+  # charset table::         nil
+  # ignore chars::          nil
+  # html strip::            false
+  # html remove elements::  ''
+  # searchd_binary_name::   searchd
+  # indexer_binary_name::   indexer
+  #
+  # If you want to change these settings, create a YAML file at
+  # config/sphinx.yml with settings for each environment, in a similar
+  # fashion to database.yml - using the following keys: config_file,
+  # searchd_log_file, query_log_file, pid_file, searchd_file_path, port,
+  # allow_star, enable_star, min_prefix_len, min_infix_len, mem_limit,
+  # max_matches, morphology, charset_type, charset_table, ignore_chars,
+  # html_strip, html_remove_elements, delayed_job_priority,
+  # searchd_binary_name, indexer_binary_name.
+  #
+  # I think you've got the idea.
+  # 
+  # Each setting in the YAML file is optional - so only put in the ones you
+  # want to change.
+  #
+  # Keep in mind, if for some particular reason you're using a version of
+  # Sphinx older than 0.9.8 r871 (that's prior to the proper 0.9.8 release),
+  # don't set allow_star to true.
+  # 
+  class Configuration
+    include Singleton
+    
+    SourceOptions = %w( mysql_connect_flags sql_range_step sql_query_pre
+      sql_query_post sql_ranged_throttle sql_query_post_index )
+    
+    IndexOptions  = %w( charset_table charset_type docinfo enable_star
+      exceptions html_index_attrs html_remove_elements html_strip ignore_chars
+      min_infix_len min_prefix_len min_word_len mlock morphology ngram_chars
+      ngram_len phrase_boundary phrase_boundary_step preopen stopwords
+      wordforms )
+    
+    CustomOptions = %w( disable_range )
+        
+    attr_accessor :config_file, :searchd_log_file, :query_log_file,
+      :pid_file, :searchd_file_path, :address, :port, :allow_star,
+      :database_yml_file, :app_root, :bin_path, :model_directories,
+      :delayed_job_priority, :searchd_binary_name, :indexer_binary_name
+    
+    attr_accessor :source_options, :index_options
+    
+    attr_writer :type
+    
+    attr_reader :environment, :configuration
+    
+    # Load in the configuration settings - this will look for config/sphinx.yml
+    # and parse it according to the current environment.
+    # 
+    def initialize(app_root = Dir.pwd)
+      self.reset
+    end
+    
+    def self.configure(&block)
+      yield instance
+      instance.reset(instance.app_root)
+    end
+    
+    def reset(custom_app_root=nil)
+      if custom_app_root
+        self.app_root = custom_app_root
+      else
+        self.app_root          = RAILS_ROOT if defined?(RAILS_ROOT)
+        self.app_root          = Merb.root  if defined?(Merb)
+        self.app_root        ||= app_root
+      end
+      
+      @configuration = Riddle::Configuration.new
+      @configuration.searchd.address    = "127.0.0.1"
+      @configuration.searchd.port       = 3312
+      @configuration.searchd.pid_file   = "#{self.app_root}/log/searchd.#{environment}.pid"
+      @configuration.searchd.log        = "#{self.app_root}/log/searchd.log"
+      @configuration.searchd.query_log  = "#{self.app_root}/log/searchd.query.log"
+      
+      self.database_yml_file    = "#{self.app_root}/config/database.yml"
+      self.config_file          = "#{self.app_root}/config/#{environment}.sphinx.conf"
+      self.searchd_file_path    = "#{self.app_root}/db/sphinx/#{environment}"
+      self.allow_star           = false
+      self.bin_path             = ""
+      self.model_directories    = ["#{app_root}/app/models/"] +
+        Dir.glob("#{app_root}/vendor/plugins/*/app/models/")
+      self.delayed_job_priority = 0
+      
+      self.source_options  = {}
+      self.index_options   = {
+        :charset_type => "utf-8"
+      }
+      
+      self.searchd_binary_name = "searchd"
+      self.indexer_binary_name = "indexer"
+            
+      parse_config
+      
+      self
+    end
+    
+    def self.environment
+      @@environment ||= (
+        defined?(Merb) ? Merb.environment : ENV['RAILS_ENV']
+      ) || "development"
+    end
+    
+    def environment
+      self.class.environment
+    end
+    
+    def controller
+      @controller ||= Riddle::Controller.new(@configuration, self.config_file)
+    end
+    
+    # Generate the config file for Sphinx by using all the settings defined and
+    # looping through all the models with indexes to build the relevant
+    # indexer and searchd configuration, and sources and indexes details.
+    #
+    def build(file_path=nil)
+      load_models
+      file_path ||= "#{self.config_file}"
+      
+      @configuration.indexes.clear
+      
+      ThinkingSphinx.indexed_models.each_with_index do |model, model_index|
+        @configuration.indexes.concat model.constantize.to_riddle(model_index)
+      end
+      
+      open(file_path, "w") do |file|
+        file.write @configuration.render
+      end
+    end
+    
+    # Make sure all models are loaded - without reloading any that
+    # ActiveRecord::Base is already aware of (otherwise we start to hit some
+    # messy dependencies issues).
+    # 
+    def load_models
+      return if defined?(Rails) && Rails.configuration.cache_classes
+      
+      self.model_directories.each do |base|
+        Dir["#{base}**/*.rb"].each do |file|
+          model_name = file.gsub(/^#{base}([\w_\/\\]+)\.rb/, '\1')
+        
+          next if model_name.nil?
+          next if ::ActiveRecord::Base.send(:subclasses).detect { |model|
+            model.name == model_name
+          }
+        
+          begin
+            model_name.camelize.constantize
+          rescue LoadError
+            model_name.gsub!(/.*[\/\\]/, '').nil? ? next : retry
+          rescue NameError
+            next
+          rescue StandardError
+            puts "Warning: Error loading #{file}"
+          end
+        end
+      end
+    end
+    
+    def address
+      @configuration.searchd.address
+    end
+    
+    def address=(address)
+      @configuration.searchd.address = address
+    end
+    
+    def port
+      @configuration.searchd.port
+    end
+    
+    def port=(port)
+      @configuration.searchd.port = port
+    end
+    
+    def pid_file
+      @configuration.searchd.pid_file
+    end
+    
+    def pid_file=(pid_file)
+      @configuration.searchd.pid_file = pid_file
+    end
+    
+    def searchd_log_file
+      @configuration.searchd.log
+    end
+    
+    def searchd_log_file=(file)
+      @configuration.searchd.log = file
+    end
+    
+    def query_log_file
+      @configuration.searchd.query_log
+    end
+    
+    def query_log_file=(file)
+      @configuration.searchd.query_log = file
+    end
+    
+    def type
+      return @type if defined?(@type)
+      case ::ActiveRecord::Base.sphinx_database_adapter
+      when SQLite3Adapter, OracleAdapter
+        @type = "xml"
+      else
+        @type = "sql"
+      end
+    end
+    
+    private
+    
+    # Parse the config/sphinx.yml file - if it exists - then use the attribute
+    # accessors to set the appropriate values. Nothing too clever.
+    # 
+    def parse_config
+      path = "#{app_root}/config/sphinx.yml"
+      return unless File.exists?(path)
+      
+      conf = YAML::load(ERB.new(IO.read(path)).result)[environment]
+      
+      conf.each do |key,value|
+        self.send("#{key}=", value) if self.respond_to?("#{key}=")
+        next if key == "type"
+        
+        set_sphinx_setting self.source_options, key, value, SourceOptions
+        set_sphinx_setting self.index_options,  key, value, IndexOptions
+        set_sphinx_setting self.index_options,  key, value, CustomOptions
+        set_sphinx_setting @configuration.searchd, key, value
+        set_sphinx_setting @configuration.indexer, key, value
+      end unless conf.nil?
+      
+      self.bin_path += '/' unless self.bin_path.blank?
+      
+      if self.allow_star
+        self.index_options[:enable_star]    = true
+        self.index_options[:min_prefix_len] = 1
+      end
+    end
+    
+    def set_sphinx_setting(object, key, value, allowed = {})
+      if object.is_a?(Hash)
+        object[key.to_sym] = value if allowed.include?(key.to_s)
+      else
+        object.send("#{key}=", value) if object.respond_to?("#{key}")
+        send("#{key}=", value) if self.respond_to?("#{key}")
+      end
+    end
+  end
+end

data/lib/thinking_sphinx/core/string.rb

@@ -0,0 +1,15 @@
+require 'zlib'
+
+module ThinkingSphinx
+  module Core
+    module String
+      def to_crc32
+        Zlib.crc32 self
+      end
+    end
+  end
+end
+
+class String
+  include ThinkingSphinx::Core::String
+end

data/lib/thinking_sphinx/deltas.rb

@@ -0,0 +1,30 @@
+require 'thinking_sphinx/deltas/default_delta'
+require 'thinking_sphinx/deltas/delayed_delta'
+require 'thinking_sphinx/deltas/datetime_delta'
+
+module ThinkingSphinx
+  module Deltas
+    def self.parse(index)
+      delta_option = index.local_options.delete(:delta)
+      case delta_option
+      when TrueClass, :default
+        DefaultDelta.new index, index.local_options
+      when :delayed
+        DelayedDelta.new index, index.local_options
+      when :datetime
+        DatetimeDelta.new index, index.local_options
+      when FalseClass, nil
+        nil
+      else
+        if delta_option.is_a?(String)
+          delta_option = Kernel.const_get(delta_option)
+        end
+        if delta_option.ancestors.include?(ThinkingSphinx::Deltas::DefaultDelta)
+          delta_option.new index, index.local_options
+        else
+          raise "Unknown delta type"
+        end
+      end
+    end
+  end
+end