Attention! The file “element-mapping.json” gets overwritten on Stages updates. Create a backup copy if you want to save your work.

For multi language support the following steps have to be executed:

(Delete the nodes folder in “<Stages Installation Directory>\elasticsearch\data\nodes”)

(Currently, it is not necessary to modify the other search configuration files “user-mapping.json” and “workspace-mapping.json”.)

Hints:

• The language IDs are the same as those in your file “config.xml” in the “supported.languages” attribute.
• The setting section name is the string before the colon, the values are behind the colon.
• Not every duplicated section contains values where the language ID can be simply replaced. Instead, they must be configured correctly, which will be described in the step 5.
• Currently there are more than 70 occurrences of “[*]_en”.
• Example for Korean language (i.e “ko”) with values which can be replaced trivially:

Before:

"name_analyzer_en": {
   "type": "custom",
   "tokenizer": "standard",
   "filter": [ "lowercase", "icu_folding", "possessive_stemmer_en", "stemmer_en", "edge_ngram" ],
   "char_filter" : [ "icu_normalizer" ]
}

After:

"name_analyzer_ko": {
   "type": "custom",
   "tokenizer": "standard",
   "filter": [ "lowercase", "icu_folding", "edge_ngram" ],
   "char_filter" : [ "icu_normalizer" ]
}

Help for configuring the specifics of your language (e.g. stemming, stop words, tokenizing and filtering…) which provides optimized search results can be found here:

• https://www.elastic.co/guide/en/elasticsearch/guide/current/language-intro.html
• https://www.elastic.co/de/blog/using-korean-analyzers
• https://www.elastic.co/de/blog/how-to-search-ch-jp-kr-part-1

(To verify if the index creation is completed, open “stages.log” and search for the log message “Started creation/update of search index for all workspaces.”)