Elasticsearch does not seem to be working as expected - elasticsearch

Question

I am using elasticsearch and globalize gems for full text searching and what I expect is that I can search for supplier name, localised description using czech/english analyzer.
Example:
Supplier Name: "Bonami.cz"
Supplier Description_CZ: "Test description in czech."
It works when I search for "Bonami.cz", but it does not work (0 results) when I search for:
"Bonami" (part of the word)
"test" (description)
Based on documentation, the below methods should work, but apparently I have missed something. I verified the indexes and data is in ElasticSearch.
Also do I need to install somehow the czech/english analyzer before using it in the model?
require 'elasticsearch/model'
require 'activerecord-import'
class Supplier < ActiveRecord::Base
after_commit lambda { __elasticsearch__.index_document }, on: :create
after_commit lambda { __elasticsearch__.update_document }, on: :update
translates :description, :fallbacks_for_empty_translations => true
accepts_nested_attributes_for :translations
include Elasticsearch::Model
include Elasticsearch::Model::Callbacks
include Elasticsearch::Model::Globalize::MultipleFields
mapping do
indexes :id, type: 'integer'
indexes :name, analyzer: 'czech'
indexes :description_ma, analyzer: 'czech'
indexes :description_cs, analyzer: 'czech'
indexes :description_en, analyzer: 'english'
end
def as_indexed_json(options={})
{ id: id,
name: name,
description_ma: description_ma,
description_cs: description_cs,
description_en: description_en
}
end
def self.search(query)
__elasticsearch__.search(
{
query: {
multi_match: {
query: query,
fields: ['name^10', 'description_ma', 'description_cs', 'description_en']
}
}
})
end
end
Any idea, what is wrong?
Thanks, Miroslav
UPDATE 1
I inspired with the solution from Rails 4, elasticsearch-rails, but when I try to search now, for any word I always get zero results.
settings index: {
number_of_shards: 1,
analysis: {
filter: {
trigrams_filter: {
type: 'ngram',
min_gram: 2,
max_gram: 10
},
content_filter: {
type: 'ngram',
min_gram: 4,
max_gram: 20
}
},
analyzer: {
index_trigrams_analyzer: {
type: 'custom',
tokenizer: 'standard',
filter: ['lowercase', 'trigrams_filter']
},
search_trigrams_analyzer: {
type: 'custom',
tokenizer: 'whitespace',
filter: ['lowercase']
},
english: {
tokenizer: 'standard',
filter: ['standard', 'lowercase', 'content_filter']
},
czech: {
tokenizer: 'standard',
filter: ['standard','lowercase','content_filter']
}
}
}
} do
mappings dynamic: 'false' do
indexes :name, index_analyzer: 'index_trigrams_analyzer', search_analyzer: 'search_trigrams_analyzer'
indexes :description_en, index_analyzer: 'english', search_analyzer: 'english'
indexes :description_ma, index_analyzer: 'czech', search_analyzer: 'czech'
indexes :description_cs, index_analyzer: 'czech', search_analyzer: 'czech'
end
end
def as_indexed_json(options={})
{ id: id,
name: name,
description_ma: description_ma,
description_cs: description_cs,
description_en: description_en
}
end
def self.search(query)
__elasticsearch__.search(
{
query: {
multi_match: {
query: query,
fields: ['name^10', 'description_ma', 'description_cs', 'description_en']
}
},
highlight: {
pre_tags: ['<em>'],
post_tags: ['</em>'],
fields: {
name: {},
description_ma: {},
description_cs: {},
description_en: {}
}
}
}
)
end
This is what I see when I open elastic search URL for the given model:
{"suppliers":{"aliases":{},"mappings":{"supplier":
{"dynamic":"false","properties":{"description_cs":
{"type":"string","analyzer":"czech"},"description_en":
{"type":"string","analyzer":"english"},"description_ma":
{"type":"string","analyzer":"czech"},"name":
{"type":"string","index_analyzer":"index_trigrams_analyzer","search_analyzer":"search_trigrams_analyzer"}}}},"settings":{"index":
{"creation_date":"1445797508427","analysis":{"filter":
{"trigrams_filter":
{"type":"ngram","min_gram":"2","max_gram":"10"},"content_filter":
{"type":"ngram","min_gram":"4","max_gram":"20"}},"analyzer":{"english":
{"filter":["standard","lowercase","content_filter"],"tokenizer":"standard"},"index_trigrams_analyzer":{"type":"custom","filter":["lowercase","trigrams_filter"],"tokenizer":"standard"},"search_trigrams_analyzer":{"type":"custom","filter":["lowercase"],"tokenizer":"whitespace"},"czech":{"filter":["standard","lowercase","content_filter"],"tokenizer":"standard"}}},"number_of_shards":"1","number_of_replicas":"1","version":
{"created":"1060099"},"uuid":"wX9kf3OQSva24Iwk7sZ8AQ"}},"warmers":{}}}
UPDATE 2
Two steps were missing to have it working as expected =>
1. Re-import model data
2. Typo in names of description fields (instead of description_ma/en/cs, I had to use ma/cs/en_description.
settings index: {
number_of_shards: 1,
analysis: {
filter: {
trigrams_filter: {
type: 'ngram',
min_gram: 2,
max_gram: 10
},
content_filter: {
type: 'ngram',
min_gram: 4,
max_gram: 20
}
},
analyzer: {
index_trigrams_analyzer: {
type: 'custom',
tokenizer: 'standard',
filter: ['lowercase', 'trigrams_filter']
},
search_trigrams_analyzer: {
type: 'custom',
tokenizer: 'whitespace',
filter: ['lowercase']
},
english: {
tokenizer: 'standard',
filter: ['standard', 'lowercase', 'content_filter']
},
czech: {
tokenizer: 'standard',
filter: ['standard','lowercase','content_filter' ]
}
}
}
} do
mappings dynamic: 'false' do
indexes :name, index_analyzer: 'index_trigrams_analyzer', search_analyzer: 'search_trigrams_analyzer'
indexes :en_description, index_analyzer: 'english', search_analyzer: 'english'
indexes :ma_description, index_analyzer: 'czech', search_analyzer: 'czech'
indexes :cs_description, index_analyzer: 'czech', search_analyzer: 'czech'
end
end
def as_indexed_json(options={})
{ id: id,
name: name,
ma_description: ma_description,
cs_description: cs_description,
en_description: en_description
}
end
def self.search(query)
__elasticsearch__.search(
{
query: {
multi_match: {
query: query,
fields: ['name^10', 'ma_description', 'cs_description', 'en_description']
}
},
highlight: {
pre_tags: ['<em>'],
post_tags: ['</em>'],
fields: {
name: {},
ma_description: {},
cs_description: {},
en_description: {}
}
}
}
)
end

Answer 1

In order to be able to perform the search you are trying to do you'll to use the Ngram analyzer. (as discussed in the comments)