fix(es): add java date format support

[etolbakov]

Description

add conversion from elasticsearch/opensearch date format to quickwit expected strptime

How was this PR tested?

Provide curl output below

Related Issue

#5228

The code is take from #5237

[etolbakov]

Testing results

Request 1

curl -X GET "http://localhost:7280/api/v1/_elastic/stackoverflow-schemaless/_search?pretty" -H 'Content-Type: application/json' -d'
{
  "query": {
    "range": {
      "age": {
        "gte": "20230101",
        "lte": "20231231",
        "boost": 2.0,
        "format": "yyyyMMdd"
      }
    }
  }
}
'
{
  "took": 4,
  "timed_out": false,
  "_shards": {
    "total": 1,
    "successful": 1,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": {
      "value": 0,
      "relation": "eq"
    },
    "hits": []
  }
}%

Request 2

curl -X GET "http://localhost:7280/api/v1/_elastic/stackoverflow-schemaless/_search?pretty" -H 'Content-Type: application/json' -d'
{
  "query": {
    "range": {
      "age": {
        "gte": "20230101",
        "lte": "20231231",
        "boost": 2.0,
        "format": "basic_date"
      }
    }
  }
}
'
{
  "took": 11,
  "timed_out": false,
  "_shards": {
    "total": 1,
    "successful": 1,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": {
      "value": 0,
      "relation": "eq"
    },
    "hits": []
  }
}%

Request 3 (failure case)

curl -X GET "http://localhost:7280/api/v1/_elastic/stackoverflow-schemaless/_search?pretty" -H 'Content-Type: application/json' -d'
{
  "query": {
    "range": {
      "age": {
        "gte": "2023-01-01",
        "lte": "2023-12-31",
        "boost": 2.0,
        "format": "yyyy-MM-dd'T'HH:mm:ss.SSSZ"
      }
    }
  }
}
'
{
  "status": 400,
  "error": {
    "caused_by": null,
    "reason": "failed to parse range query date format. failed to parse date format `yyyy-MM-ddTHH:mm:ss.SSSZ`. Pattern at pos 10 is not recognized.",
    "stack_trace": null,
    "type": null
  }
}%

[etolbakov]

@fulmicoton just a ping about this change.
Could you please take a look when you have time?

[fulmicoton]

Suggested change

	let format = if java_datetime_format.contains('_') {
	java_date_format_aliases()
	.get(java_datetime_format)
	.unwrap_or(&java_datetime_format)
	} else {
	java_datetime_format
	};
	// We check if the date format is present in our alias dictionary.
	let java_date_format_resolved: &str = java_date_format_aliases()
	.get(java_datetime_format)
	.unwrap_or(&java_datetime_format);

[fulmicoton]

Suggested change

regex = { workspace = true }

[fulmicoton]

Suggested change

time = { workspace = true }

[etolbakov]

This one is required by range_query.rs

use time::format_description::well_known::Rfc3339;

[fulmicoton]

we are still missing a lot of the patterns... Hours etc.

[fulmicoton]

We need rest unit tests and to handle more patterns.

[fulmicoton]

I don't think we handle this format right now.

[github-actions]

On SSD:

Average search latency is 0.875x that of the reference (lower is better).
Ref run id: 2433, ref commit: be20923
Link

On GCS:

Average search latency is 0.896x that of the reference (lower is better).
Ref run id: 2435, ref commit: be20923
Link

[kuzaxak]

@etolbakov Can I reopen it from local branch in QW repo?

[kuzaxak]

@etolbakov I think we can close this one and continue in a local branch here I refactored you tokenizer to support nested optional and uncommented test cases (they are passing on my local).