Fix custom prompt substitute with List issue

[mingshl]

Description

fix custom prompt substitute with List issue in ml inference search response processor

using toString() method in placeholder to convert list into string, for example, ${parameters.context.toString()} in the following processors config:

PUT /_search/pipeline/my_pipeline_request_review_llm
{
  "response_processors": [
    {
      "ml_inference": {
        "tag": "ml_inference",
        "description": "This processor is going to run llm",
        "model_id": "cf46K5EBoVpekzRp8x_3",
        "function_name": "REMOTE",
        "input_map": [
          {
            "context": "review"
          }
        ],
        "output_map": [
          {
            "llm_response": "response"
          }
        ],
        "model_config": {
          "prompt":"\n\nHuman: You are a professional data analysist. You will always answer question based on the given context first. If the answer is not directly shown in the context, you will analyze the data and find the answer. If you don't know the answer, just say I don't know. Context: ${parameters.context.toString()}. \n\n Human: please summarize the documents \n\n Assistant:"
        },
        "ignore_missing": false,
        "ignore_failure": false
      }
    }
  ]
}

reverting https://github.com/opensearch-project/ml-commons/pull/2811/files as a more appropriate approach to fix the toString issue #2839

#2880

Check List

• New functionality includes testing.
• New functionality has been documented.
• API changes companion pull request created.
• Commits are signed per the DCO using --signoff.
• Public documentation issue/PR created.

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.

[ylwu-amzn]

Add prefix to method parameter list ? For connector, it will be parameters. For other usecase, it maybe different, like local model could be model_config

Suggested change

	public static Map<String, String> parseParameters(Map<String, String> parameters) {
	public static Map<String, String> parseParameters(String prefix, Map<String, String> parameters) {

[mingshl]

all parameters will be prefix with parameters in here

ml-commons/common/src/main/java/org/opensearch/ml/common/connector/HttpConnector.java

Line 327 in 59b417b

StringSubstitutor substitutor = new StringSubstitutor(parameters, "${parameters.", "}");

I added the test to test when it's parameters.model_config.context, then it would be taking the prefix as model_config.context properly, please see this UT

ml-commons/common/src/test/java/org/opensearch/ml/common/utils/StringUtilsTest.java

Lines 316 to 340 in 59b417b

	public void testParseParametersListToStringModelConfig() {
	Map<String, String> parameters = new HashMap<>();
	parameters
	.put(
	"prompt",
	"\\n\\nHuman: You are a professional data analyst. You will always answer question based on the given context first. If the answer is not directly shown in the context, you will analyze the data and find the answer. If you don't know the answer, just say I don't know. Context: ${parameters.model_config.context.toString()}. \\n\\n Human: please summarize the documents \\n\\n Assistant:"
	);
	ArrayList<String> listOfDocuments = new ArrayList<>();
	listOfDocuments.add("document1");
	parameters.put("model_config.context", toJson(listOfDocuments));
	parseParameters(parameters);
	System.out.println(parameters.get("context" + TO_STRING_FUNCTION_NAME));
	System.out.println(parameters);
	assertEquals(parameters.get("model_config.context" + TO_STRING_FUNCTION_NAME), "[\\\"document1\\\"]");
	String requestBody = "{\"prompt\": \"${parameters.prompt}\"}";
	StringSubstitutor substitutor = new StringSubstitutor(parameters, "${parameters.", "}");
	requestBody = substitutor.replace(requestBody);
	System.out.println(requestBody);
	assertEquals(
	requestBody,
	"{\"prompt\": \"\\n\\nHuman: You are a professional data analyst. You will always answer question based on the given context first. If the answer is not directly shown in the context, you will analyze the data and find the answer. If you don't know the answer, just say I don't know. Context: [\\\"document1\\\"]. \\n\\n Human: please summarize the documents \\n\\n Assistant:\"}"
	);
	}

[dhrubo-os]

Shouldn't we use model interface instead of adding toString to the parameter to identify if we need to convert that to string? cc @ylwu-amzn