Mapping changes which require changes to index settings (applied to existing index)

gistfile1.txt

# Original Index Structure (just took the pages field for brevity)

PUT test_index
{
    "settings": {
        "analysis": {
            "char_filter": {
                "punct_annotation": {
                    "type": "mapping",
                    "mappings": [
                        ".   => \\n_PERIOD_\\n",
                        "\\n => \\n_NEWLINE_\\n",
                        ",   => \\n_COMMA_\\n",
                        ";   => \\n_SEMI_\\n"
                    ]
                },
                "punctuation": {"type": "mapping", "mappings": [".=>"]}
            },
            "normalizer": {
                "keyword_normalizer": {"type": "custom", "char_filter": [], "filter": ["lowercase", "asciifolding"]}
            },
            "filter": {
                "english_stop": {"type": "stop", "stopwords": "_english_"},
                "english_stemmer": {"type": "stemmer", "language": "english"},
                "english_possessive_stemmer": {"type": "stemmer", "language": "possessive_english"}
            },
            "analyzer": {
                "cleanedStem": {
                    "tokenizer": "standard",
                    "filter": ["english_possessive_stemmer", "lowercase", "english_stop", "english_stemmer"],
                    "char_filter": ["punct_annotation"],
                    "type": "custom"
                },
                "exactMatch": {
                    "tokenizer": "standard",
                    "filter": ["lowercase"],
                    "char_filter": ["punct_annotation"],
                    "type": "custom"
                },
                "exactInput": {"tokenizer": "whitespace", "filter": ["lowercase"], "type": "custom"},
                "suggester_lowercase": {
                    "tokenizer": "standard",
                    "char_filter": ["punctuation"],
                    "filter": ["lowercase", "asciifolding"]
                }
            }
        }
    },
    "mappings": {
        "_source": {"enabled": true},
        "dynamic": "false",
        "properties": {
            "pages": {
                "type": "nested",
                "properties": {
                    "timestamp": {"type": "date", "store": true, "format": "strict_date_hour_minute_second_fraction"},
                    "url": {
                        "type": "keyword",
                        "store": true
                    },
                    "type": {
                        "type": "keyword",
                        "store": true
                    },
                    "text": {
                        "type": "text",
                        "store": true,
                        "index_options": "offsets",
                        "fielddata": true,
                        "fields": {
                            "stemmed": {"type": "text", "analyzer": "cleanedStem", "index_options": "offsets"},
                            "exact": {"type": "text", "analyzer": "exactMatch", "index_options": "offsets"}
                        }
                    }
                }
            }
        }
    }
}


GET test_index/_mapping/field/pages.text

PUT test_index/_mapping
{
    "properties": {
        "all_pages": {
            "type": "text",
            "similarity": "spread_similarity",
            "fields": {
                "exact": {
                    "type": "text",
                    "similarity": "spread_similarity",
                    "index_options": "offsets",
                    "analyzer": "exactMatch"
                },
                "stemmed": {
                    "type": "text",
                    "similarity": "spread_similarity",
                    "index_options": "offsets",
                    "analyzer": "cleanedStem"
                }
            }
        },
        "pages": {
            "type": "nested",
            "properties": {
                "text": {
                    "type": "text",
                    "store": true,
                    "index_options": "offsets",
                    "fielddata": true,
                    "fields": {
                        "stemmed": {"type": "text", "analyzer": "cleanedStem", "index_options": "offsets"},
                        "exact": {"type": "text", "analyzer": "exactMatch", "index_options": "offsets"}
                    },
                    "copy_to": "all_pages"
                }
            }
        }
    }
}


GET test_index/_settings

PUT test_index/_settings
{
    "settings": {
        "index": {
            "similarity": {
                "spread_similarity": {
                    "type": "BM25",
                    "b": 1,  
                    "k1": 1.75
                }
            }
        }
    }
}

# Close index before changes
POST test_index/_close

# Can writes happen?
POST test_index/_doc/1
{
  "pages": [
    {
      "text": ["Hello World!", "Is this index open?"]
    }
  ]
}

# Can reads happen
GET test_index/_search
{
  "query": {
    "nested": {
      "path": "pages",
      "query": {
        "match_phrase": {
          "pages.text.stemmed": "hello world"
        }
      }
    }
  }
}

GET test_index/_search
{
  "query": {
    "match_phrase": {
      "all_pages.stemmed": "hello world"
    }
  }
}

# Reopen index after changes
POST test_index/_open

DELETE test_index/_doc/1