Add opensearch support

[nicoloboschi]

Changes:

• New datasource "opensearch" that support OpenSearch v2 and AWS OpenSearch Serverless
• New asset "opensearch-index" that creates an index on opensearch. mappings and settings are fully configurable
• "query-vector-db" can now query on a opensearch datasource. The query object is fully configurable-
• "vector-db-sink" can now writes to a opensearch datasource. Only bulk is supported with batch-size and flush-interval. All the bulk options are supported (e.g. refresh, timeout..)
• Added example in the examples directory

FULL Rag example

topics:
  - name: "insert-topic"
    creation-mode: create-if-not-exists
assets:
  - name: "os-index"
    asset-type: "opensearch-index"
    creation-mode: create-if-not-exists
    config:
        index-name: "my-index-1"
        datasource: "OSDatasource"
        settings: |
            {
                "index": {
                      "knn": true,
                      "knn.algo_param.ef_search": 100
                }
            }
        mappings: |
            {
                "properties": {
                      "content": {
                            "type": "text"
                      },
                      "embeddings": {
                            "type": "knn_vector",
                            "dimension": 3
                      }
                }
            }
pipeline:
  - id: write
    name: "Write"
    type: "vector-db-sink"
    input: "insert-topic"
    configuration:
      datasource: "OSDatasource"
      index-name: "my-index-1"
      bulk-parameters:
        refresh: "true"
      id: "key"
      fields:
        - name: "content"
          expression: "value.content"
        - name: "embeddings"
          expression: "value.embeddings"
---
topics:
  - name: "input-topic"
    creation-mode: create-if-not-exists
  - name: "result-topic"
    creation-mode: create-if-not-exists
pipeline:
  - id: read
    name: "read"
    type: "query-vector-db"
    input: "input-topic"
    output: "result-topic"
    configuration:
      datasource: "OSDatasource"
      query: |
        {
          "query": {
            "knn": {
              "embeddings": {
                "vector": ?,
                "k": 1
              }
            }
          }
        }
      fields:
        - "value.embeddings"
      output-field: "value.query-result"

[eolivelli]

Over LGTM
Please add a example application, like the one for solr

[eolivelli]

Nit: indexName