Elasticsearch-rails 补全建议器、地图数据、创建索引、建议方法答案

【问题标题】：Elasticsearch-rails completion suggester, map data, create indexes, suggest methodElasticsearch-rails 补全建议器、地图数据、创建索引、建议方法
【发布时间】：2015-06-23 10:00:41
【问题描述】：

我正在使用 elasticsearch-model 和 elasticsearch-rails 宝石。

在我的个人资料模型中，我有

settings index: { number_of_shards: 3 } do
    mappings dynamic: 'false'  do
      indexes :profile_type, analyzer: 'keyword'
      indexes :profile do |p|
        indexes :name, type: 'multi_field' do
          indexes :name, analyzer: 'snowball'
          indexes :tokenized, analyzer: 'keyword'
        end
      end
    end
 end

我尝试添加suggest 字段

settings index: { number_of_shards: 3 } do
    mappings dynamic: 'false'  do
      indexes :profile_type, analyzer: 'keyword'
      indexes :suggest, type: 'suggest' # not sure about this
      # do not really know how to index this field
      indexes :profile do |p|
        indexes :name, type: 'multi_field' do
          indexes :name, analyzer: 'snowball'
          indexes :tokenized, analyzer: 'keyword'
        end
      end
    end
 end

我正在尝试对 profile.name 提出建议，我已添加 as_indexed_json

def as_indexed_json(options={})
    suggest= {
     suggest: {
      :input => profile.name,
      :output => profile.name,
      :payload => {
        name: profile.name
      }
     }
    }
    self.as_json(
      :only => [:profile_type, :slug],
      :include => {
        :profile => {}
      }
    ).merge(suggest)
end

任何帮助，任何建议都将不胜感激。

【问题讨论】：

标签： ruby-on-rails ruby elasticsearch elasticsearch-rails

【解决方案1】：

我想出了如何使用elasticsearch-rails gem 映射数据

settings index: { number_of_shards: 3 } do
    mappings dynamic: 'false'  do
      indexes :profile_type, analyzer: 'keyword'
      indexes :suggest, type: 'completion', index_analyzer: 'simple', search_analyzer: 'simple', payloads: true
      # will do the trick
      indexes :profile do |p|
        indexes :name, type: 'multi_field' do
          indexes :name, analyzer: 'snowball'
          indexes :tokenized, analyzer: 'keyword'
        end
      end
    end
 end

然后创建索引

def as_indexed_json(options={})
    suggest= {
     suggest: {
      :input => profile.name,
      :output => profile.name,
      :payload => {
        name: profile.name
      }
     }
    }
    self.as_json(
      :only => [:profile_type, :slug],
      :include => {
        :profile => {}
      }
    ).merge(suggest)
end

那么不要使用search方法，而是使用suggest方法

  def self.suggest query
    __elasticsearch__.client.suggest(:index => index_name, :body => {
      :suggestions => {
        :text => query,
        :completion => {
          :field => 'suggest',
          :size => 20
        }
      }
    })
  end

这样你就可以得到complete suggestion

【讨论】：