# This is a configuration file for environments that use Japanese and Sudachi plug-ins.
# To use this file, copy it to the Mastodon root directory and rename the file to ".elasticsearch.yml".

version: 1

accounts:
  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:
    natural:
      tokenizer: standard
      filter:
        - lowercase
        - asciifolding
        - cjk_width
        - elision
        - english_possessive_stemmer
        - english_stop
        - english_stemmer
    sudachi_analyzer:
      type: custom
      tokenizer: sudachi_tokenizer
      filter:
        - my_posfilter
        - sudachi_normalizedform
    verbatim:
      tokenizer: standard
      filter:
        - lowercase
        - asciifolding
        - cjk_width
    edge_ngram:
      tokenizer: edge_ngram
      filter:
        - lowercase
        - asciifolding
        - cjk_width

  tokenizer:
    edge_ngram:
      type: edge_ngram
      min_gram: 1
      max_gram: 15
    sudachi_tokenizer:
      resources_path: '/etc/elasticsearch/sudachi'
      split_mode: A
      type: sudachi_tokenizer
      discard_punctuation: 'true'

public_statuses:
  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:
        - english_possessive_stemmer
        - lowercase
        - asciifolding
        - cjk_width
        - english_stop
        - english_stemmer
    hashtag:
      tokenizer: keyword
      filter:
        - word_delimiter_graph
        - lowercase
        - asciifolding
        - cjk_width
    sudachi_analyzer:
      tokenizer: sudachi_tokenizer
      type: custom
      filter:
        - 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'

statuses:
  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:
    verbatim:
      tokenizer: uax_url_email
      filter:
        - lowercase
    content:
      tokenizer: uax_url_email
      filter:
        - english_possessive_stemmer
        - lowercase
        - asciifolding
        - cjk_width
        - english_stop
        - english_stemmer
    hashtag:
      tokenizer: keyword
      filter:
        - word_delimiter_graph
        - lowercase
        - asciifolding
        - cjk_width
    sudachi_analyzer:
      tokenizer: sudachi_tokenizer
      type: custom
      filter:
        - english_possessive_stemmer
        - lowercase
        - asciifolding
        - cjk_width
        - english_stop
        - english_stemmer
        - my_posfilter
        - sudachi_normalizedform

tags:
  analyzer:
    content:
      tokenizer: keyword
      filter:
        - word_delimiter_graph
        - lowercase
        - asciifolding
        - cjk_width
    edge_ngram:
      tokenizer: edge_ngram
      filter:
        - lowercase
        - asciifolding
        - cjk_width

  tokenizer:
    edge_ngram:
      type: edge_ngram
      min_gram: 2
      max_gram: 15

accounts_analyzers:
  display_name:
    analyzer: verbatim
    edge_ngram:
      analyzer: edge_ngram
      search_analyzer: verbatim
  username:
    analyzer: verbatim
    edge_ngram:
      analyzer: edge_ngram
      search_analyzer: verbatim
  text:
    analyzer: sudachi_analyzer
    stemmed:
      analyzer: natural

public_statuses_analyzers:
  text:
    analyzer: sudachi_analyzer
    stemmed:
      analyzer: content
  tags:
    analyzer: hashtag

statuses_analyzers:
  text:
    analyzer: sudachi_analyzer
    stemmed:
      analyzer: content
  tags:
    analyzer: hashtag

tags_analyzers:
  name:
    analyzer: content
    edge_ngram:
      analyzer: edge_ngram
      search_analyzer: content