Skip to content

Commit 63b570c

Browse files
estolfoestolfo
committed
[API] Add mlt back
1 parent 3e0e029 commit 63b570c

File tree

2 files changed

+260
-0
lines changed
  • elasticsearch-api

2 files changed

+260
-0
lines changed

‎elasticsearch-api/lib/elasticsearch/api/actions/mlt.rb

+130
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,130 @@
1+
module Elasticsearch
2+
module API
3+
module Actions
4+
5+
# Return documents similar to the specified one.
6+
#
7+
# Performs a `more_like_this` query with the specified document as the input.
8+
#
9+
# @example Search for similar documents using the `title` property of document `myindex/mytype/1`
10+
#
11+
# # First, let's setup a synonym-aware analyzer ("quick" <=> "fast")
12+
# client.indices.create index: 'myindex', body: {
13+
# settings: {
14+
# analysis: {
15+
# filter: {
16+
# synonyms: {
17+
# type: 'synonym',
18+
# synonyms: [ "quick,fast" ]
19+
# }
20+
# },
21+
# analyzer: {
22+
# title_synonym: {
23+
# type: 'custom',
24+
# tokenizer: 'whitespace',
25+
# filter: ['lowercase', 'stop', 'snowball', 'synonyms']
26+
# }
27+
# }
28+
# }
29+
# },
30+
# mappings: {
31+
# mytype: {
32+
# properties: {
33+
# title: {
34+
# type: 'string',
35+
# analyzer: 'title_synonym'
36+
# }
37+
# }
38+
# }
39+
# }
40+
# }
41+
#
42+
# # Index three documents
43+
# client.index index: 'myindex', type: 'mytype', id: 1, body: { title: 'Quick Brown Fox' }
44+
# client.index index: 'myindex', type: 'mytype', id: 2, body: { title: 'Slow Black Dog' }
45+
# client.index index: 'myindex', type: 'mytype', id: 3, body: { title: 'Fast White Rabbit' }
46+
# client.indices.refresh index: 'myindex'
47+
#
48+
# client.mlt index: 'myindex', type: 'mytype', id: 1, mlt_fields: 'title', min_doc_freq: 1, min_term_freq: 1
49+
# # => { ... {"title"=>"Fast White Rabbit"}}]}}
50+
#
51+
# @option arguments [String] :id The document ID (*Required*)
52+
# @option arguments [String] :index The name of the index (*Required*)
53+
# @option arguments [String] :type The type of the document (use `_all` to fetch
54+
# the first document matching the ID across all types) (*Required*)
55+
# @option arguments [Hash] :body A specific search request definition
56+
# @option arguments [Number] :boost_terms The boost factor
57+
# @option arguments [Number] :max_doc_freq The word occurrence frequency as count: words with higher occurrence
58+
# in the corpus will be ignored
59+
# @option arguments [Number] :max_query_terms The maximum query terms to be included in the generated query
60+
# @option arguments [Number] :max_word_len The minimum length of the word: longer words will be ignored
61+
# @option arguments [Number] :min_doc_freq The word occurrence frequency as count: words with lower occurrence
62+
# in the corpus will be ignored
63+
# @option arguments [Number] :min_term_freq The term frequency as percent: terms with lower occurence
64+
# in the source document will be ignored
65+
# @option arguments [Number] :min_word_len The minimum length of the word: shorter words will be ignored
66+
# @option arguments [List] :mlt_fields Specific fields to perform the query against
67+
# @option arguments [Number] :percent_terms_to_match How many terms have to match in order to consider
68+
# the document a match (default: 0.3)
69+
# @option arguments [String] :routing Specific routing value
70+
# @option arguments [Number] :search_from The offset from which to return results
71+
# @option arguments [List] :search_indices A comma-separated list of indices to perform the query against
72+
# (default: the index containing the document)
73+
# @option arguments [String] :search_query_hint The search query hint
74+
# @option arguments [String] :search_scroll A scroll search request definition
75+
# @option arguments [Number] :search_size The number of documents to return (default: 10)
76+
# @option arguments [String] :search_source A specific search request definition (instead of using the request body)
77+
# @option arguments [String] :search_type Specific search type (eg. `dfs_then_fetch`, `count`, etc)
78+
# @option arguments [List] :search_types A comma-separated list of types to perform the query against
79+
# (default: the same type as the document)
80+
# @option arguments [List] :stop_words A list of stop words to be ignored
81+
#
82+
# @see http://elasticsearch.org/guide/reference/api/more-like-this/
83+
#
84+
def mlt(arguments={})
85+
raise ArgumentError, "Required argument 'index' missing" unless arguments[:index]
86+
raise ArgumentError, "Required argument 'type' missing" unless arguments[:type]
87+
raise ArgumentError, "Required argument 'id' missing" unless arguments[:id]
88+
method = HTTP_GET
89+
path = Utils.__pathify Utils.__escape(arguments[:index]),
90+
Utils.__escape(arguments[:type]),
91+
Utils.__escape(arguments[:id]),
92+
'_mlt'
93+
94+
params = Utils.__validate_and_extract_params arguments, ParamsRegistry.get(__method__)
95+
96+
[:mlt_fields, :search_indices, :search_types, :stop_words].each do |name|
97+
params[name] = Utils.__listify(params[name]) if params[name]
98+
end
99+
100+
body = arguments[:body]
101+
102+
perform_request(method, path, params, body).body
103+
end
104+
105+
# Register this action with its valid params when the module is loaded.
106+
#
107+
# @since 6.2.0
108+
ParamsRegistry.register(:mlt, [
109+
:boost_terms,
110+
:max_doc_freq,
111+
:max_query_terms,
112+
:max_word_len,
113+
:min_doc_freq,
114+
:min_term_freq,
115+
:min_word_len,
116+
:mlt_fields,
117+
:percent_terms_to_match,
118+
:routing,
119+
:search_from,
120+
:search_indices,
121+
:search_query_hint,
122+
:search_scroll,
123+
:search_size,
124+
:search_source,
125+
:search_type,
126+
:search_types,
127+
:stop_words ].freeze)
128+
end
129+
end
130+
end

‎elasticsearch-api/spec/elasticsearch/api/actions/mlt_spec.rb

+130
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,130 @@
1+
module Elasticsearch
2+
module API
3+
module Actions
4+
5+
# Return documents similar to the specified one.
6+
#
7+
# Performs a `more_like_this` query with the specified document as the input.
8+
#
9+
# @example Search for similar documents using the `title` property of document `myindex/mytype/1`
10+
#
11+
# # First, let's setup a synonym-aware analyzer ("quick" <=> "fast")
12+
# client.indices.create index: 'myindex', body: {
13+
# settings: {
14+
# analysis: {
15+
# filter: {
16+
# synonyms: {
17+
# type: 'synonym',
18+
# synonyms: [ "quick,fast" ]
19+
# }
20+
# },
21+
# analyzer: {
22+
# title_synonym: {
23+
# type: 'custom',
24+
# tokenizer: 'whitespace',
25+
# filter: ['lowercase', 'stop', 'snowball', 'synonyms']
26+
# }
27+
# }
28+
# }
29+
# },
30+
# mappings: {
31+
# mytype: {
32+
# properties: {
33+
# title: {
34+
# type: 'string',
35+
# analyzer: 'title_synonym'
36+
# }
37+
# }
38+
# }
39+
# }
40+
# }
41+
#
42+
# # Index three documents
43+
# client.index index: 'myindex', type: 'mytype', id: 1, body: { title: 'Quick Brown Fox' }
44+
# client.index index: 'myindex', type: 'mytype', id: 2, body: { title: 'Slow Black Dog' }
45+
# client.index index: 'myindex', type: 'mytype', id: 3, body: { title: 'Fast White Rabbit' }
46+
# client.indices.refresh index: 'myindex'
47+
#
48+
# client.mlt index: 'myindex', type: 'mytype', id: 1, mlt_fields: 'title', min_doc_freq: 1, min_term_freq: 1
49+
# # => { ... {"title"=>"Fast White Rabbit"}}]}}
50+
#
51+
# @option arguments [String] :id The document ID (*Required*)
52+
# @option arguments [String] :index The name of the index (*Required*)
53+
# @option arguments [String] :type The type of the document (use `_all` to fetch
54+
# the first document matching the ID across all types) (*Required*)
55+
# @option arguments [Hash] :body A specific search request definition
56+
# @option arguments [Number] :boost_terms The boost factor
57+
# @option arguments [Number] :max_doc_freq The word occurrence frequency as count: words with higher occurrence
58+
# in the corpus will be ignored
59+
# @option arguments [Number] :max_query_terms The maximum query terms to be included in the generated query
60+
# @option arguments [Number] :max_word_len The minimum length of the word: longer words will be ignored
61+
# @option arguments [Number] :min_doc_freq The word occurrence frequency as count: words with lower occurrence
62+
# in the corpus will be ignored
63+
# @option arguments [Number] :min_term_freq The term frequency as percent: terms with lower occurence
64+
# in the source document will be ignored
65+
# @option arguments [Number] :min_word_len The minimum length of the word: shorter words will be ignored
66+
# @option arguments [List] :mlt_fields Specific fields to perform the query against
67+
# @option arguments [Number] :percent_terms_to_match How many terms have to match in order to consider
68+
# the document a match (default: 0.3)
69+
# @option arguments [String] :routing Specific routing value
70+
# @option arguments [Number] :search_from The offset from which to return results
71+
# @option arguments [List] :search_indices A comma-separated list of indices to perform the query against
72+
# (default: the index containing the document)
73+
# @option arguments [String] :search_query_hint The search query hint
74+
# @option arguments [String] :search_scroll A scroll search request definition
75+
# @option arguments [Number] :search_size The number of documents to return (default: 10)
76+
# @option arguments [String] :search_source A specific search request definition (instead of using the request body)
77+
# @option arguments [String] :search_type Specific search type (eg. `dfs_then_fetch`, `count`, etc)
78+
# @option arguments [List] :search_types A comma-separated list of types to perform the query against
79+
# (default: the same type as the document)
80+
# @option arguments [List] :stop_words A list of stop words to be ignored
81+
#
82+
# @see http://elasticsearch.org/guide/reference/api/more-like-this/
83+
#
84+
def mlt(arguments={})
85+
raise ArgumentError, "Required argument 'index' missing" unless arguments[:index]
86+
raise ArgumentError, "Required argument 'type' missing" unless arguments[:type]
87+
raise ArgumentError, "Required argument 'id' missing" unless arguments[:id]
88+
method = HTTP_GET
89+
path = Utils.__pathify Utils.__escape(arguments[:index]),
90+
Utils.__escape(arguments[:type]),
91+
Utils.__escape(arguments[:id]),
92+
'_mlt'
93+
94+
params = Utils.__validate_and_extract_params arguments, ParamsRegistry.get(__method__)
95+
96+
[:mlt_fields, :search_indices, :search_types, :stop_words].each do |name|
97+
params[name] = Utils.__listify(params[name]) if params[name]
98+
end
99+
100+
body = arguments[:body]
101+
102+
perform_request(method, path, params, body).body
103+
end
104+
105+
# Register this action with its valid params when the module is loaded.
106+
#
107+
# @since 6.2.0
108+
ParamsRegistry.register(:mlt, [
109+
:boost_terms,
110+
:max_doc_freq,
111+
:max_query_terms,
112+
:max_word_len,
113+
:min_doc_freq,
114+
:min_term_freq,
115+
:min_word_len,
116+
:mlt_fields,
117+
:percent_terms_to_match,
118+
:routing,
119+
:search_from,
120+
:search_indices,
121+
:search_query_hint,
122+
:search_scroll,
123+
:search_size,
124+
:search_source,
125+
:search_type,
126+
:search_types,
127+
:stop_words ].freeze)
128+
end
129+
end
130+
end

0 commit comments

Comments
 (0)