1
0
Fork 0
forked from gitea/nas

Change: #532 ElasticSearch設定の外出し (#650)

* Change: #532 ElasticSearch設定の外出し

* バージョンチェック

* 起動時にエラー
This commit is contained in:
KMY(雪あすか) 2024-03-12 12:11:13 +09:00 committed by GitHub
parent 8e7c66522e
commit a8fbcb3fb6
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
9 changed files with 508 additions and 269 deletions

View file

@ -3,81 +3,9 @@
class PublicStatusesIndex < Chewy::Index
include DatetimeClampingConcern
settings index: index_preset(refresh_interval: '30s', number_of_shards: 5), analysis: {
filter: {
english_stop: {
type: 'stop',
stopwords: '_english_',
},
english_stemmer: {
type: 'stemmer',
language: 'english',
},
english_possessive_stemmer: {
type: 'stemmer',
language: 'possessive_english',
},
my_posfilter: {
type: 'sudachi_part_of_speech',
stoptags: [
'助詞',
'助動詞',
'補助記号,句点',
'補助記号,読点',
],
},
},
analyzer: {
content: {
tokenizer: 'uax_url_email',
filter: %w(
english_possessive_stemmer
lowercase
asciifolding
cjk_width
english_stop
english_stemmer
),
},
hashtag: {
tokenizer: 'keyword',
filter: %w(
word_delimiter_graph
lowercase
asciifolding
cjk_width
),
},
sudachi_analyzer: {
tokenizer: 'sudachi_tokenizer',
type: 'custom',
filter: %w(
english_possessive_stemmer
lowercase
asciifolding
cjk_width
english_stop
english_stemmer
my_posfilter
sudachi_normalizedform
),
},
},
tokenizer: {
sudachi_tokenizer: {
resources_path: '/etc/elasticsearch/sudachi',
split_mode: 'A',
type: 'sudachi_tokenizer',
discard_punctuation: 'true',
},
},
}
# ElasticSearch config is moved to "/config/elasticsearch.default.yml".
# Edit it when original Mastodon changed ElasticSearch config.
settings index: index_preset(refresh_interval: '30s', number_of_shards: 5), analysis: ChewyConfig.instance.public_statuses
index_scope ::Status.unscoped
.kept
@ -87,8 +15,8 @@ class PublicStatusesIndex < Chewy::Index
root date_detection: false do
field(:id, type: 'long')
field(:account_id, type: 'long')
field(:text, type: 'text', analyzer: 'sudachi_analyzer', value: ->(status) { status.searchable_text }) { field(:stemmed, type: 'text', analyzer: 'content') }
field(:tags, type: 'text', analyzer: 'hashtag', value: ->(status) { status.tags.map(&:display_name) })
field(:text, type: 'text', analyzer: ChewyConfig.instance.public_statuses_analyzers.dig('text', 'analyzer'), value: ->(status) { status.searchable_text }) { field(:stemmed, type: 'text', analyzer: ChewyConfig.instance.public_statuses_analyzers.dig('text', 'stemmed', 'analyzer')) }
field(:tags, type: 'text', analyzer: ChewyConfig.instance.public_statuses_analyzers.dig('tags', 'analyzer'), value: ->(status) { status.tags.map(&:display_name) })
field(:language, type: 'keyword')
field(:domain, type: 'keyword', value: ->(status) { status.account.domain || '' })
field(:properties, type: 'keyword', value: ->(status) { status.searchable_properties })