common_indexer 0.1.1 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e1a9fecdbf7f2fc69bc396780a18870cf644fbf57d3d2cc8a58d8ce288916660
-  data.tar.gz: a0308434a2594a856fd864dabfbd45f6c3dd193f3373e37bc415b505071694c1
+  metadata.gz: 4461d22a5d4531a8ddf037db23a49802daf2809c997ad5004d1df38efd47b3ce
+  data.tar.gz: fe7b4728cb10b2241b8b45b241d08eb0fa34d5ee9904fd09d8f7665d1fc4f047
 SHA512:
-  metadata.gz: c4ef48902ee9fcda97963a124914786ccc780a848cf875be093211418330a1e7227e9f1e978239c574bdedee69fbc550a5944a4369ed0313e0c64bf2adcc6bc1
-  data.tar.gz: 755eb2c497f8a52098635f22f8c5018bcf170a35940ccbb7e71a51c9bb3d0f5647eb5cb25fc6998f128dfa2638c421f81293f62b1864af9e0dc3c75653c52489
+  metadata.gz: c402bf32e0fbb80a87a0204269f9c95a3b772f19e82342be4b35db952e3444398958527ec92b3a61082d0dd6c4c75e14285638941756f6efb02f003e68fc29c5
+  data.tar.gz: 75210f819de79a6fbcc569bcf4be859d04c1b1f4c8144bbd4783ec4a2d19cb83ddc26fc0e754f7abb8fa5807b2ec48e4278f572252b332c5a8a728804d8d436e

data/README.md

@@ -1,7 +1,7 @@
 # CommonIndexer
 [![Build Status](https://travis-ci.com/nulib/common-indexer.svg?branch=master)](https://travis-ci.com/nulib/common-indexer)
 
-Indexes metadata into a central AWS Elasticsearch instance. The gem indexes based on a hash returned by the method `#to_common_index` defined in your model using an `after_save` hook. The indexer also sanitizes the input by only allowing pre-configured hash keys.
+Indexes metadata into a central AWS Elasticsearch instance. The gem indexes based on a hash returned by the method `#to_common_index` defined in your model using an `after_save` hook.
 
 ## Installation
 
@@ -28,9 +28,6 @@ Add `common_indexer` to the appropriate Rails config file (e.g. `config/settings
 common_indexer:
   endpoint: http://localhost:9201/ # default 'http://localhost:9200'
   index_name: new-index # default 'common'
-  allowed_keys: # default ['title']
-    - title
-    - creator
 ```
 
 Add an initializer to configure the CommonIndexer gem with the app settings:
@@ -40,13 +37,12 @@ Add an initializer to configure the CommonIndexer gem with the app settings:
 ::CommonIndexer.config do |config|
   config.endpoint = Settings.common_indexer.endpoint
   config.index_name = Settings.common_indexer.index_name
-  config.allowed_keys = Settings.common_indexer.allowed_keys
 end
 ```
 
 ## Usage
 
-Include the CommonIndexer into your model that you want to index with `include ::CommonIndexer::Base`, and define a `#to_common_index` method in that model. `#to_common_index` should return a hash with metadata key/values that conform to the allowed keys defined in the local configuration or fall back to the defaults.
+Include the CommonIndexer into your model that you want to index with `include ::CommonIndexer::Base`, and define a `#to_common_index` method in that model. `#to_common_index` should return a hash with metadata key/values that conform to the common index mapping.
 
 ```ruby
 class Example < ActiveFedora::Base

data/common_indexer.gemspec

@@ -22,6 +22,8 @@ Gem::Specification.new do |spec|
 
   spec.add_dependency 'active-fedora'
   spec.add_dependency 'activesupport'
+  spec.add_dependency 'dry-configurable'
+  spec.add_dependency 'json'
   spec.add_dependency 'typhoeus'
   spec.add_dependency 'elasticsearch'
   spec.add_development_dependency 'bixby'

data/config/schema.json

@@ -0,0 +1,154 @@
+{
+  "properties": {
+    "full_text": {
+      "type": "text"
+    },
+    "abstract": {
+      "type": "text",
+      "copy_to": "full_text"
+    },
+    "admin_set": {
+      "type": "nested",
+      "properties": {
+        "id": {
+          "type": "keyword"
+        },
+        "title": {
+          "type": "text",
+          "copy_to": ["full_text", "admin_set_facet"]
+        }
+      }
+    },
+    "caption": {
+      "type": "text",
+      "copy_to": "full_text"
+    },
+    "collection": {
+      "type": "nested",
+      "properties": {
+        "id": {
+          "type": "keyword"
+        },
+        "title": {
+          "type": "text",
+          "copy_to": ["full_text", "collection_facet"]
+        }
+      }
+    },
+    "contributor": {
+      "type": "nested",
+      "properties": {
+        "role": {
+          "type": "keyword"
+        },
+        "uri": {
+          "type": "keyword"
+        },
+        "label": {
+          "type": "text",
+          "copy_to": ["full_text", "contributor_facet"]
+        }
+      }
+    },
+    "date": {
+      "type": "text"
+    },
+    "description": {
+      "type": "text"
+    },
+    "keyword": {
+      "type": "keyword",
+      "copy_to": "full_text"
+    },
+    "language": {
+      "type": "nested",
+      "properties": {
+        "uri": {
+          "type": "keyword"
+        },
+        "label": {
+          "type": "text",
+          "copy_to": ["full_text", "language_facet"]
+        }
+      }
+    },
+    "location": {
+      "type": "geo_point"
+    },
+    "permalink": {
+      "type": "keyword"
+    },
+    "provenance": {
+      "type": "text"
+    },
+    "publisher": {
+      "type": "text",
+      "copy_to": "publisher_facet"
+    },
+    "rights_holder": {
+      "type": "text"
+    },
+    "source": {
+      "type": "text"
+    },
+    "subject": {
+      "type": "nested",
+      "properties": {
+        "role": {
+          "type": "keyword"
+        },
+        "uri": {
+          "type": "keyword"
+        },
+        "label": {
+          "type": "text",
+          "copy_to": ["full_text", "subject_facet"]
+        }
+      }
+    },
+    "title": {
+      "type": "object",
+      "properties": {
+        "primary": {
+          "type": "text",
+          "copy_to": "full_text"
+        },
+        "alternate": {
+          "type": "text",
+          "copy_to": "full_text"
+        }
+      }
+    },
+    "visibility": {
+      "type": "keyword",
+      "copy_to": "visibility_facet"
+    },
+    "expanded_date": {
+      "type": "date"
+    },
+    "year": {
+      "type": "integer"
+    },
+    "extra_fields": {
+      "type": "nested"
+    }
+  },
+  "dynamic_templates": [
+    {
+      "facets": {
+        "match":   "*_facet",
+        "mapping": {
+          "type": "keyword"
+        }
+      }
+    },
+    {
+      "nested": {
+        "match": "*_nested",
+        "mapping": {
+          "type": "nested"
+        }
+      }
+    }
+  ]
+}

data/lib/common_indexer/base.rb

@@ -7,6 +7,7 @@ module CommonIndexer # :nodoc:
 
     included do
       after_save :update_common_index
+      after_destroy :delete_document
     end
 
     # Override to_common_index in the including model and return a hash of common index attributes
@@ -14,11 +15,21 @@ module CommonIndexer # :nodoc:
       raise CommonIndexer::Error, "Index Error: #{self.class.name} does not implement #to_common_index!"
     end
 
+    def delete_document
+      CommonIndexer.client.delete index: CommonIndexer.index_name,
+                                  id: id,
+                                  type: '_doc'
+    rescue Elasticsearch::Transport::Transport::Error => err
+      Rails.logger.warn("Common Index delete failure: #{err.message}")
+    end
+
     def update_common_index
       CommonIndexer.client.index index: CommonIndexer.index_name,
                                  id: id,
-                                 type: self.class.name.underscore,
-                                 body: CommonIndexer.sanitize(to_common_index)
+                                 type: '_doc',
+                                 body: to_common_index
+    rescue Elasticsearch::Transport::Transport::Error => err
+      Rails.logger.warn("Common Indexing failure: #{err.message}")
     end
   end
 end

data/lib/common_indexer/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module CommonIndexer
-  VERSION = '0.1.1'
+  VERSION = '0.2.0'
 end

data/lib/common_indexer.rb

@@ -1,30 +1,45 @@
 # frozen_string_literal: true
 
+require 'active_support/core_ext/module/delegation'
+require 'dry-configurable'
+require 'json'
 require 'typhoeus'
 require 'typhoeus/adapters/faraday'
 require 'elasticsearch'
 require 'common_indexer/version'
 require 'common_indexer/base'
-require 'common_indexer/config'
 
 module CommonIndexer # :nodoc:
   class Error < StandardError; end
 
-  def self.client
-    @client ||= Elasticsearch::Client.new(hosts: config.endpoint)
-  end
+  extend Dry::Configurable
 
-  def self.index_name
-    config.index_name
-  end
+  setting :endpoint, 'http://localhost:9200/'
+  setting :index_name, 'common'
+  setting :schema, JSON.parse(File.read(File.expand_path('../../config/schema.json', __FILE__)))
 
-  def self.sanitize(input)
-    input.select do |key, _value|
-      config.allowed_keys.include?(key.to_s)
+  class << self
+    delegate :index_name, to: :config
+
+    def client
+      @client ||= Elasticsearch::Client.new(hosts: config.endpoint)
     end
-  end
 
-  def self.config
-    (@config ||= Config.new).tap { |c| yield c if block_given? }
+    def configure_index!
+      new_index = [index_name, Time.now.utc.strftime('%Y%m%d%H%M%S%3N')].join('_')
+      client.indices.create(index: new_index)
+      client.indices.put_mapping(index: new_index, type: '_doc', body: { _doc: config.schema })
+      reindex_into(new_index) if client.indices.exists(index: index_name)
+      client.indices.put_alias(index: new_index, name: index_name)
+    end
+
+    private
+
+    def reindex_into(new_index)
+      existing_index = client.indices.get(index: index_name).keys.first
+      raise ArgumentError, "Cannot reindex #{index_name} into itself" if existing_index == index_name
+      client.reindex(body: { source: { index: index_name }, dest: { index: new_index } })
+      client.indices.delete(index: existing_index)
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: common_indexer
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.2.0
 platform: ruby
 authors:
 - Michael Klein
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-07-26 00:00:00.000000000 Z
+date: 2018-08-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: active-fedora
@@ -39,6 +39,34 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: dry-configurable
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: json
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: typhoeus
   requirement: !ruby/object:Gem::Requirement
@@ -156,9 +184,9 @@ files:
 - bin/console
 - bin/setup
 - common_indexer.gemspec
+- config/schema.json
 - lib/common_indexer.rb
 - lib/common_indexer/base.rb
-- lib/common_indexer/config.rb
 - lib/common_indexer/version.rb
 homepage: https://github.com/nulib/common-indexer
 licenses:

data/lib/common_indexer/config.rb

@@ -1,13 +0,0 @@
-# frozen_string_literal: true
-
-module CommonIndexer
-  DEFAULT_KEYS       = ['title'].freeze
-  DEFAULT_ENDPOINT   = 'http://localhost:9200/'
-  DEFAULT_INDEX_NAME = 'common'
-
-  Config = Struct.new(:endpoint, :index_name, :allowed_keys) do
-    def initialize(endpoint = DEFAULT_ENDPOINT, index_name = DEFAULT_INDEX_NAME, allowed_keys = DEFAULT_KEYS)
-      super
-    end
-  end
-end