multi-language index mappings via templates POC

template_test.sh

#!/bin/sh

# This tries to use index templates and dynamic_templates in order to define a 
# 'DRY' multi language index-mapping (index per language). 
# The 'general' template defines all field mappings (mostly via dynamic_templates) and 
# their analyzers (which are underspecified on this level).
# The language templates ('de') do instantiate the the language specific analyzers 
# and match the language suffix.

if [ -z "$ES" ]; then
  ES="http://localhost:9200/"
  echo "\$ES variable not set - using default value: '$ES'\n"
else
  echo "You have set the ES environment variable to: '$ES'\n"
fi


echo "\nDelete and create the 'general' template"
curl -XDELETE $ES/_template/general

curl -XPUT $ES/_template/general -d '{
  "template": "myindex*",
    "order": 0,
    "settings": {
      "number_of_shards": 1,
      "index" : {
        "analysis" : {
          "analyzer" : {
            "ana_index_nolang" : {
              "filter" : [ "trim", "lowercase", "asciifolding"],
              "tokenizer" : "standard"
            }
          }
        }
      }
    },
    "mappings": {
      "document": {
        "_source": {
          "enabled": true
        },
        "dynamic_templates": [
        {
          "string_ngram_lang": {
            "match": "*_lang",
            "mapping": {
              "type": "string",
              "index_analyzer": "ana_index_ngram",
              "search_analyzer": "ana_search",
              "fields": {
                "simple": {
                  "type": "string",
                  "analyzer": "ana_index_nolang"
                }
              }
            }
          }
        }],
        "properties" : {
          "boost" : {
            "type" : "double"
          }
        }
      }
    }
}'



echo "\nDelete and create the 'de' template"
curl -XDELETE $ES/_template/de

curl -XPUT $ES/_template/de -d '{
  "template": "*_de",
    "order": 1,
    "settings": {
      "index" : {
        "analysis" : {
          "analyzer" : {
            "ana_index_ngram" : {
              "filter" : [ "trim", "lowercase", "stopwords", "stemmer", "asciifolding", "ngram1_30" ],
              "tokenizer" : "standard"
            },
            "ana_search": {
              "filter": [ "trim", "lowercase", "stopwords", "stemmer", "asciifolding" ],
              "tokenizer": "standard"
            }
          },
          "filter": {
            "stopwords": {
              "type": "stop",
              "stopwords": ["mit", "auf", "..."]
            },
            "stemmer": {
              "type": "stemmer",
              "name": "german"
            },
            "ngram1_30": {
              "type": "nGram",
              "min_gram": 1,
              "max_gram": 30
            }
          }
        }
      }
    }
}'


echo "\nDelete index 'myindex_de'"
curl -XDELETE $ES/myindex_de

echo "\nCreate index by putting a 'document'"
curl -XPUT $ES/myindex_de/document/1 -d '{
  "myfield_lang": "Der Habicht, das Lamm, Hurz!",
  "boost": 9
}'

echo