Using JsonPath to read Web Service Response

I made a demo project on GitHub:

This project was developed to propose a solution to a discussion “Get json value with skipping 1st and 2nd level of dynamic value” in the Katalon Forum.

Problem to solve

The questioner has a JSON text as follows:

{
    "-1": [
        {
            "-1": {
                "build": "",
                "createdDate": "",
                "description": "",
                "endDate": "",
                "ended": "",
                "environment": "",
                "executionSummaries": {
                    "executionSummary": [

                    ]
                },
                "expand": "executionSummaries",
                "isExecutionWorkflowEnabledForProject": true,
                "isTimeTrackingEnabled": true,
                "modifiedBy": "",
                "name": "Ad hoc",
                "projectId": 14900,
                "projectKey": "TUT",
                "startDate": "",
                "started": "",
                "totalCycleExecutions": 2,
                "totalExecuted": 0,
                "totalExecutions": 0,
                "versionId": -1,
                "versionName": "Unscheduled"
            },
            "13": {
                "build": "",
                "createdBy": "sysadmin",
                "createdByDisplay": "System Administrator",
                "createdDate": "2019-01-30 17:41:12.403",
                "description": "Create cycle from Katalon",
                "endDate": "",
                "ended": "",
                "environment": "",
                "executionSummaries": {
                    "executionSummary": [

                    ]
                },
                "expand": "executionSummaries",
                "isExecutionWorkflowEnabledForProject": true,
                "isTimeTrackingEnabled": true,
                "modifiedBy": "sysadmin",
                "name": "ST Cycle 1",
                "projectId": 14900,
                "projectKey": "TUT",
                "sprintId": 1,
                "startDate": "",
                "started": "",
                "totalCycleExecutions": 1,
                "totalDefects": 0,
                "totalExecuted": 0,
                "totalExecutions": 0,
                "totalFolders": 0,
                "versionId": -1,
                "versionName": "Unscheduled"
            },
            "recordsCount": 2
        }
    ]
}

He wants to look up “13” out of this JSON. He wants to do it with the "name": "ST Cycle 1" portion as key for look up.

Evan, the questioner, tried to parse the JSON using JsonSlurper, tried to consume it using basic Groovy language constructs (Iterator, for loop, etc). He found it difficult.

Solution

Previously, Vinh_Nguyen suggested JsonPath to analyze and selectively extract data out of JSON document. I think JsonPath would be a powerful tool for anybody who wants to process Web Service Responses in JSON. However nobody in Katalon Forum posted sample which shows how to use JsonPath.

Then, I will do it.

Description

How to run the demo

  1. Download the zip of demo from Releases page
  2. unzip it and open it with your Katalon Studio
  3. select Test Cases/TC1 and run it.
  4. in the Console tab, you will find the output

How the demo is constructed

You can find and read the source code of TC1 located here.

The Drivers/ directory contains the jar files of JsonPath downloaded from Releases page.

The input file is located at Include/resources/fixture/response.json

JsonPath examples and outcomes

Here I show the code snippet and corresponding outputs. I would not describe the meanings of JsonPath expressions here. Please read the original Jayway JsonPath document for detail.

Case0 : show the JSON as is

code:

Path projectDir = Paths.get(RunConfiguration.getProjectDir())
Path responseFile = projectDir.resolve('Include/resources/fixture/response.json')
def json = responseFile.toFile().getText('UTF-8')
Object document = Configuration.defaultConfiguration().jsonProvider().parse(json);
def fragment0 = JsonPath.read(document, '$')
println "\n---------- Case 0 ----------\n" + prettyPrint(fragment0)

output:

{
    "-1": [
        {
            "-1": {
                "build": "",
                "createdDate": "",
                "description": "",
                "endDate": "",
                "ended": "",
                "environment": "",
                "executionSummaries": {
                    "executionSummary": [

                    ]
                },
                "expand": "executionSummaries",
                "isExecutionWorkflowEnabledForProject": true,
                "isTimeTrackingEnabled": true,
                "modifiedBy": "",
                "name": "Ad hoc",
                "projectId": 14900,
                "projectKey": "TUT",
                "startDate": "",
                "started": "",
                "totalCycleExecutions": 2,
                "totalExecuted": 0,
                "totalExecutions": 0,
                "versionId": -1,
                "versionName": "Unscheduled"
            },
            "13": {
                "build": "",
                "createdBy": "sysadmin",
                "createdByDisplay": "System Administrator",
                "createdDate": "2019-01-30 17:41:12.403",
                "description": "Create cycle from Katalon",
                "endDate": "",
                "ended": "",
                "environment": "",
                "executionSummaries": {
                    "executionSummary": [

                    ]
                },
                "expand": "executionSummaries",
                "isExecutionWorkflowEnabledForProject": true,
                "isTimeTrackingEnabled": true,
                "modifiedBy": "sysadmin",
                "name": "ST Cycle 1",
                "projectId": 14900,
                "projectKey": "TUT",
                "sprintId": 1,
                "startDate": "",
                "started": "",
                "totalCycleExecutions": 1,
                "totalDefects": 0,
                "totalExecuted": 0,
                "totalExecutions": 0,
                "totalFolders": 0,
                "versionId": -1,
                "versionName": "Unscheduled"
            },
            "recordsCount": 2
        }
    ]
}

Case1 : skipping the upper levels

code:

def fragment1 = JsonPath.read(document, '$.*.*')
println "\n---------- Case 1 ----------\n" + prettyPrint(fragment1)

output:

---------- Case 1 ----------
[
    {
        "-1": {
            "build": "",
            "createdDate": "",
            "description": "",
            "endDate": "",
            "ended": "",
            "environment": "",
            "executionSummaries": {
                "executionSummary": [

                ]
            },
            "expand": "executionSummaries",
            "isExecutionWorkflowEnabledForProject": true,
            "isTimeTrackingEnabled": true,
            "modifiedBy": "",
            "name": "Ad hoc",
            "projectId": 14900,
            "projectKey": "TUT",
            "startDate": "",
            "started": "",
            "totalCycleExecutions": 2,
            "totalExecuted": 0,
            "totalExecutions": 0,
            "versionId": -1,
            "versionName": "Unscheduled"
        },
        "13": {
            "build": "",
            "createdBy": "sysadmin",
            "createdByDisplay": "System Administrator",
            "createdDate": "2019-01-30 17:41:12.403",
            "description": "Create cycle from Katalon",
            "endDate": "",
            "ended": "",
            "environment": "",
            "executionSummaries": {
                "executionSummary": [

                ]
            },
            "expand": "executionSummaries",
            "isExecutionWorkflowEnabledForProject": true,
            "isTimeTrackingEnabled": true,
            "modifiedBy": "sysadmin",
            "name": "ST Cycle 1",
            "projectId": 14900,
            "projectKey": "TUT",
            "sprintId": 1,
            "startDate": "",
            "started": "",
            "totalCycleExecutions": 1,
            "totalDefects": 0,
            "totalExecuted": 0,
            "totalExecutions": 0,
            "totalFolders": 0,
            "versionId": -1,
            "versionName": "Unscheduled"
        },
        "recordsCount": 2
    }
]

Case2 : skipping more

code:

def fragment2 = JsonPath.read(document, '$.*.*.*')
println "\n---------- Case 2 ----------\n" + prettyPrint(fragment2)

output:

[
    {
        "build": "",
        "createdDate": "",
        "description": "",
        "endDate": "",
        "ended": "",
        "environment": "",
        "executionSummaries": {
            "executionSummary": [

            ]
        },
        "expand": "executionSummaries",
        "isExecutionWorkflowEnabledForProject": true,
        "isTimeTrackingEnabled": true,
        "modifiedBy": "",
        "name": "Ad hoc",
        "projectId": 14900,
        "projectKey": "TUT",
        "startDate": "",
        "started": "",
        "totalCycleExecutions": 2,
        "totalExecuted": 0,
        "totalExecutions": 0,
        "versionId": -1,
        "versionName": "Unscheduled"
    },
    {
        "build": "",
        "createdBy": "sysadmin",
        "createdByDisplay": "System Administrator",
        "createdDate": "2019-01-30 17:41:12.403",
        "description": "Create cycle from Katalon",
        "endDate": "",
        "ended": "",
        "environment": "",
        "executionSummaries": {
            "executionSummary": [

            ]
        },
        "expand": "executionSummaries",
        "isExecutionWorkflowEnabledForProject": true,
        "isTimeTrackingEnabled": true,
        "modifiedBy": "sysadmin",
        "name": "ST Cycle 1",
        "projectId": 14900,
        "projectKey": "TUT",
        "sprintId": 1,
        "startDate": "",
        "started": "",
        "totalCycleExecutions": 1,
        "totalDefects": 0,
        "totalExecuted": 0,
        "totalExecutions": 0,
        "totalFolders": 0,
        "versionId": -1,
        "versionName": "Unscheduled"
    },
    2
]

Case3 : getting the values of name

code:

def fragment3 = JsonPath.read(document, '$.*.*.*.name')
println "\n---------- Case 3 ----------\n" + prettyPrint(fragment3)

output:

[
    "Ad hoc",
    "ST Cycle 1"
]

Case4 : filtering nodes by the value of name

code:

def fragment4 = JsonPath.read(document, '$.*.*.*[?(@.name==\'ST Cycle 1\')].createdDate')
println "\n---------- Case 4 ----------\n" + prettyPrint(fragment4)

output:

---------- Case 4 ----------
[
    "2019-01-30 17:41:12.403"
]

Case 5 : extracting key value requires tricks

code:

def lookupCycleId(doc, cycleName) {
	List<Map> map2 = JsonPath.read(doc, '$.*.*')
	for (Map m : map2) {
		Set<String> keySet = m.keySet()
		for (String key : keySet) {
			Map map3 = m.get(key)
			if (map3.name == cycleName) {
				return key
			}
		}
	}
	return null
}
String cycleId = lookupCycleId(document, "ST Cycle 1")
println "\n---------- Case 4 ----------\n" +
			"Cycle Id of 'ST Cycle 1' is ${cycleId}\n"

output:

---------- Case 5 ----------
Cycle Id of 'ST Cycle 1' is 13

Notes

The Zephyr for JIRA’s original format uses Cycle ID (e.g, “13”) as the key of Json objects. Why do they do that? I can’t believe this! This makes consumer programs a lot complicated.

Pity, Evan. I feel your pain. Poorly designed JSON format is the cause of your problems.

I would rather like it to be in the format as follows:

    [
        ...
        { "cycleId":"13",
            "name":"ST Cycle 1",
            ...},
        { "cycleId":"...",
            "name":"...",
            ...},
        ...
    ]

The information contained here is no less than the original. The proposed format is far easier to look up Cycle Id value with key of name value. You can look up “13” by a simple JsonPath $.*[name=='ST Cycle 1'][0].cycleId . You do not need the lengthy def lookupCycleId() function.

3 Likes

Hey Kaz. I’m surprised you didn’t use GPath. Is your solution better?

http://docs.groovy-lang.org/latest/html/documentation/core-semantics.html#gpath_expressions

1 Like

I did not know GPath.

This is my first touch with JsonPath. I did it because Vinh suggested it.

1 Like

It’s a lot like xpath which most of you guys seem to have nailed.

Please pay attention to the Note at the end of my post.

IMHO, the real problem exists in the poor format of the JSON in question. It was hard to process it with JsonPath. Even with GPath, it would be similar

2 Likes

Highly appreciate @kazurayam spending much of his time on this! People like you makes this a high quality community :smiley:

My brain was exploding with the json format. Unfortunately, my company didn’t realise how bad their API’s are. It’s a good learning experience for someone like me with zero experience on web services.

1 Like

I added Test Cases/TC2_Collections. It does not use JsonPath. It uses the Collections.findAll(Closure) methd and Collections.collect(Closure) method instead.

For example, Case5 code in TC2 is like this:

def cycleId = document.'-1'[0].
				findAll({ it.value instanceof Map && it.value.name == 'ST Cycle 1'}).
				collect({ it.key })[0]

Quite puzzling, but it works.

2 Likes

Yep, this works for me and lesser code. thanks again @kazurayam