How to search exact text without matching case in Elasticsearch - spring-boot

Question

I want to search the user name in the Elasticsearch. For this I want to match the exact username ignoring its case whether it is capital or small, I just want to find that user name. I'm using the following query for this:
QueryBuilder queryBuilder = QueryBuilders.termQuery("user_name.keyword", userName);
NativeSearchQuery build = new NativeSearchQueryBuilder().withQuery(queryBuilder).build();
List<Company> companies = elasticsearchTemplate.queryForList(build, User.class);
But it is also matching the exact word with the case. for example: if the user name is "Ram" and I search "ram" then it is not returning that name. If I search "Ram" then it is giving me the result. But I want that it only matches the word not the case of that word. Please, someone, help me to solve this problem. I searched a lot but couldn't find any solution.

Answer 1

Issue is you are using user_name.keyword and terms query. Terms query matches exact word instead of that you can use MatchQueryBuilder query :
Code :
QueryBuilder queryBuilder = QueryBuilders.matchQuery("user_name", userName);
NativeSearchQuery build = new NativeSearchQueryBuilder().withQuery(queryBuilder).build();
List<Company> companies = elasticsearchTemplate.queryForList(build, User.class);
When using .keyword field, elastic does not analyze the text but if you use your text field ElasticSerach analyzes your text using default analyzer on that field. Default Analyzer basically converts your text in lowercase and remove stopwords from it. You can read about it from here : https://www.elastic.co/guide/en/elasticsearch/reference/current/analysis-standard-analyzer.html
Since you want to do case insensitive search so you don't need to use .keyword.
Also, terms query matches exact terms but again since you want to do case insensitive search you should you match query which also by default internally converts your search text in lowercase and then search the field for that text.
And, now since both your field and search term is in lowercase you can do case insensitive search but this will not do exact match.
For doing exact case insensitive match you need to update your index and use normalizer with your keyword field which guarantees that the analysis chain produces a single token and case insensitive search. You can read more about it from here.
Index Creation:
curl -X PUT "localhost:9200/<index-name>" -H 'Content-Type: application/json' -d
{
"settings": {
"analysis": {
"normalizer": {
"case_insensitive_normalizer": {
"type": "custom",
"filter": [
"lowercase"
]
}
}
}
},
"mappings": {
"properties": {
"user_name": {
"type": "keyword",
"normalizer": "case_insensitive_normalizer"
}
}
}
}
I have indexed these documents :
Doc1 :
{
"user_name": "Ram"
}
Doc2 :
{
"user_name": "Ram Mohan"
}
Search Query :
{
"query" : {
"match" : {
"user_name" : "ram"
}
}
}
Result :
"hits": [
{
"_source": {
"user_name": "Ram"
}
}
]

Answer 2

Try to use Lowercase Token Filter in your index mapping.
https://www.elastic.co/guide/en/elasticsearch/reference/current/analysis-lowercase-tokenfilter.html
Such token filter is applied in both indexing and searching so "Ram" will be indexed as "ram" and then if you'll search for "rAm" it'll be changed to "ram" so it'll hit your document.

Answer 3

If you want to do case insensitive match on a keyword field, you can use normalizer with a lowercase filter
The normalizer property of keyword fields is similar to analyzer
except that it guarantees that the analysis chain produces a single
token.
{
"settings": {
"analysis": {
"normalizer": {
"lowercase_normalizer": {
"type": "custom",
"filter": [
"lowercase"
]
}
}
}
},
"mappings": {
"properties": {
"name": {
"type": "text",
"fields": {
"keyword": {
"type": "keyword",
"normalizer": "lowercase_normalizer"
}
}
}
}
}
}
Data
POST index41/_doc
{
"name":"Ram"
}
Query:
{
"query": {
"term": {
"name.keyword": {
"value": "ram"
}
}
}
}
Result:
"hits" : [
{
"_index" : "index41",
"_type" : "_doc",
"_id" : "IyieGHIBZsF59xnAhb47",
"_score" : 0.6931471,
"_source" : {
"name" : "Ram"
}
}
]

Answer 4

You can simply use the text field on your user-name field, text field uses by default standard analyzer which lowercase the tokens, and then match query applies the same analyzer which is used index time(in this case, standard) which will provide you case-insensitive search.
Tokens generated using the standard analyzer
POST /_analyzer
{
"text" : "ram",
"analyzer" : "standard"
}
{
"tokens": [
{
"token": "ram",
"start_offset": 0,
"end_offset": 3,
"type": "<ALPHANUM>",
"position": 0
}
]
}