generators/acts_as_xapian/USAGE generators/acts_as_xapian/acts_as_xapian_generator.rb generators/acts_as_xapian/templates/migration.rb @@ -88,7 +88,7 @@ # e.g. :texts => [ :title, :body ] # :values, things which have a range of values for indexing, or for collapsing. # Specify an array quadruple of [ field, identifier, prefix, type ] where -# - number is an arbitary numeric identifier for use in the Xapian database +# - identifier is an arbitary numeric identifier for use in the Xapian database # - prefix is the part to use in search queries that goes before the : # - type can be any of :string, :number or :date # e.g. :values => [ [ :created_at, 0, "created_at" ], [ :size, 1, "size"] ] @@ -111,8 +111,10 @@ # :if, either an attribute or a function which if returns false means the # object isn't indexed # -# 2. Make and run the migration to create the ActsAsXapianJob model, code below -# (search for ActsAsXapianJob). +# 2. Make and run the migration to create the ActsAsXapianJob model: +# script/generate acts_as_xapian +# +# (Or see code below in under ActsAsXapianJob). # # 3. Call 'rake xapian::rebuild_index models="ModelName1 ModelName2"' to build the index # the first time (you must specify all your indexed models). It's put in a @@ -249,40 +251,44 @@ module ActsAsXapian # and error check them - i.e. check for consistency between models @@query_parser.add_boolean_prefix("model", "M") @@query_parser.add_boolean_prefix("modelid", "I") - for term in options[:terms] - raise "Use a single capital letter for term code" if not term[1].match(/^[A-Z]$/) - raise "M and I are reserved for use as the model/id term" if term[1] == "M" or term[1] == "I" - raise "model and modelid are reserved for use as the model/id prefixes" if term[2] == "model" or term[2] == "modelid" - raise "Z is reserved for stemming terms" if term[1] == "Z" - raise "Already have code '" + term[1] + "' in another model but with different prefix '" + @@terms_by_capital[term[1]] + "'" if @@terms_by_capital.include?(term[1]) && @@terms_by_capital[term[1]] != term[2] - @@terms_by_capital[term[1]] = term[2] - @@query_parser.add_boolean_prefix(term[2], term[1]) + if options[:terms] + for term in options[:terms] + raise "Use a single capital letter for term code" if not term[1].match(/^[A-Z]$/) + raise "M and I are reserved for use as the model/id term" if term[1] == "M" or term[1] == "I" + raise "model and modelid are reserved for use as the model/id prefixes" if term[2] == "model" or term[2] == "modelid" + raise "Z is reserved for stemming terms" if term[1] == "Z" + raise "Already have code '" + term[1] + "' in another model but with different prefix '" + @@terms_by_capital[term[1]] + "'" if @@terms_by_capital.include?(term[1]) && @@terms_by_capital[term[1]] != term[2] + @@terms_by_capital[term[1]] = term[2] + @@query_parser.add_boolean_prefix(term[2], term[1]) + end end - for value in options[:values] - raise "Value index '"+value[1].to_s+"' must be an integer, is " + value[1].class.to_s if value[1].class != 1.class - raise "Already have value index '" + value[1].to_s + "' in another model but with different prefix '" + @@values_by_number[value[1]].to_s + "'" if @@values_by_number.include?(value[1]) && @@values_by_number[value[1]] != value[2] - - # date types are special, mark them so the first model they're seen for - if !@@values_by_number.include?(value[1]) - if value[3] == :date - value_range = Xapian::DateValueRangeProcessor.new(value[1]) - elsif value[3] == :string - value_range = Xapian::StringValueRangeProcessor.new(value[1]) - elsif value[3] == :number - value_range = Xapian::NumberValueRangeProcessor.new(value[1]) - else - raise "Unknown value type '" + value[3].to_s + "'" - end - - @@query_parser.add_valuerangeprocessor(value_range) - - # stop it being garbage collected, as - # add_valuerangeprocessor ref is outside Ruby's GC - @@value_ranges_store.push(value_range) - end - - @@values_by_number[value[1]] = value[2] - @@values_by_prefix[value[2]] = value[1] + if options[:values] + for value in options[:values] + raise "Value index '"+value[1].to_s+"' must be an integer, is " + value[1].class.to_s if value[1].class != 1.class + raise "Already have value index '" + value[1].to_s + "' in another model but with different prefix '" + @@values_by_number[value[1]].to_s + "'" if @@values_by_number.include?(value[1]) && @@values_by_number[value[1]] != value[2] + + # date types are special, mark them so the first model they're seen for + if !@@values_by_number.include?(value[1]) + if value[3] == :date + value_range = Xapian::DateValueRangeProcessor.new(value[1]) + elsif value[3] == :string + value_range = Xapian::StringValueRangeProcessor.new(value[1]) + elsif value[3] == :number + value_range = Xapian::NumberValueRangeProcessor.new(value[1]) + else + raise "Unknown value type '" + value[3].to_s + "'" + end + + @@query_parser.add_valuerangeprocessor(value_range) + + # stop it being garbage collected, as + # add_valuerangeprocessor ref is outside Ruby's GC + @@value_ranges_store.push(value_range) + end + + @@values_by_number[value[1]] = value[2] + @@values_by_prefix[value[2]] = value[1] + end end end end @@ -433,6 +439,7 @@ module ActsAsXapian # Index # Offline indexing job queue model, create with this migration: + # (or use script/generate acts_as_xapian) # class ActsAsXapianMigration < ActiveRecord::Migration # def self.up # create_table :acts_as_xapian_jobs do |t| @@ -445,7 +452,7 @@ module ActsAsXapian # end # # def self.down - # remove_table :acts_as_xapian_jobs + # drop_table :acts_as_xapian_jobs # end # end class ActsAsXapianJob < ActiveRecord::Base @@ -564,15 +571,21 @@ module ActsAsXapian doc.add_term("M" + self.class.to_s) doc.add_term("I" + doc.data) - for term in self.xapian_options[:terms] - doc.add_term(term[1] + xapian_value(term[0])) + if self.xapian_options[:terms] + for term in self.xapian_options[:terms] + doc.add_term(term[1] + xapian_value(term[0])) + end end - for value in self.xapian_options[:values] - doc.add_value(value[1], xapian_value(value[0], value[3])) + if self.xapian_options[:values] + for value in self.xapian_options[:values] + doc.add_value(value[1], xapian_value(value[0], value[3])) + end end - for text in self.xapian_options[:texts] - ActsAsXapian.term_generator.increase_termpos # stop phrases spanning different text fields - ActsAsXapian.term_generator.index_text(xapian_value(text)) + if self.xapian_options[:texts] + for text in self.xapian_options[:texts] + ActsAsXapian.term_generator.increase_termpos # stop phrases spanning different text fields + ActsAsXapian.term_generator.index_text(xapian_value(text)) + end end ActsAsXapian.writable_db.replace_document("I" + doc.data, doc) lib/acts_as_xapian.rb @@ -1,6 +1,8 @@ require 'rubygems' require 'rake' require 'rake/testtask' +require 'activerecord' +require File.dirname(__FILE__) + '/../lib/acts_as_xapian.rb' namespace :xapian do # Parameters - specify "flush=true" to save changes to the Xapian database @@ -16,7 +18,7 @@ namespace :xapian do # web server afterwards to make sure it gets the changes, rather than # still pointing to the old deleted database. desc 'Completely rebuilds Xapian search index (must specify all models)' - task :rebuild_index do + task (:rebuild_index => :environment) do raise "specify ALL your models with models=\"ModelName1 ModelName2\" as parameter" if ENV['models'].nil? ActsAsXapian.rebuild_index(ENV['models'].split(" ").map{|m| m.constantize}) end @@ -24,7 +26,7 @@ namespace :xapian do # Parameters - are models, query, offset, limit, sort_by_prefix, # collapse_by_prefix desc 'Run a query, return YAML of results' - task :query do + task (:query => :environment) do raise "specify models=\"ModelName1 ModelName2\" as parameter" if ENV['models'].nil? raise "specify query=\"your terms\" as parameter" if ENV['query'].nil? s = ActsAsXapian::Search.new(ENV['models'].split(" ").map{|m| m.constantize}, tasks/xapian.rake f29f23658c4eb2c33583d14db59afedc2890639d a23c6d1df44c4074dd957af6369c88b9d18b1660 Francis Irving francis@cat.(none) http://github.com/frabcus/acts_as_xapian/commit/9070c81ea69e67b20062d34c7d5ac9e6b44f9942 9070c81ea69e67b20062d34c7d5ac9e6b44f9942 2008-05-16T07:27:10-07:00 2008-05-16T07:27:10-07:00 Merge commit 'boone/master' Conflicts: lib/acts_as_xapian.rb fc6ee2ae319a23f5422cce3ad1a2af4b238d5df4 Francis Irving francis@cat.(none)