Parse JSON Schema - Required properties are missing from object.

[oappdev]

Hello Power Automate Community,

So I have the following fields:

First NameRow1,Last NameRow1
First NameRow1_2,LastName Row1_2
First NameRow1_3,LastName Row1_3
First NameRow1_4,LastName Row1_4
First NameRow1_5,LastName Row1_5

Still learning JSON,  What would be an sample JSON to Fill in the rest of the pending fields. 

INPUT Content:

[
    {
        "First NameRow1": "John",
        "Last NameRow1": "Doe",
    },
    {
        
        "First NameRow1": "John",
        "Last NameRow1": "Doe1",
        
    }
]

Schema:

[
    {
        "First NameRow1": "",
        "Last NameRow1": "",
    },
    {
        "First NameRow1_2": "",
        "Last NameRow1_2": "",        
    },
    {
        "First NameRow1_3": "",
        "Last NameRow1_3": "",        
    },
    {
        "First NameRow1_4": "",
        "Last NameRow1_4": "",        
    },
    {
        "First NameRow1_5": "",
        "Last NameRow1_5": ""
    },
"required": [
      "First NameRow1",
      "Last NameRow1",
      "First NameRow1_2",
      "Last NameRow1_2",
      "First NameRow1_3",
      "Last NameRow1_3",
      "First NameRow1_4",
      "Last NameRow1_4",
      "First NameRow1_4",
      "Last NameRow1_4",
      "First NameRow1_5",
      "Last NameRow1_5"

    ]
 }
]

OUTPUT Errors:

[
  {
    "message": "Required properties are missing from object: First NameRow1_2, Last NameRow1_2,First NameRow1_3,Last NameRow1_3,First NameRow1_4,Last NameRow1_4,First NameRow1_5,Last NameRow1_5",
    "lineNumber": 0,
    "linePosition": 0,
    "path": "[0]",
    "value": [
      "First NameRow1_2",
      "Last NameRow1_2",
      "First NameRow1_3",
      "Last NameRow1_3",
      "First NameRow1_4",
      "Last NameRow1_4",
      "First NameRow1_5",
      "Last NameRow1_5"
    ],
    "schemaId": "#/items",
    "errorType": "required",
    "childErrors": []
  }
]

As always, Thank you so much in advance.

[Pstork1 · Accepted solution]

A JSON array is denoted by the use of square brackets [].  Inside the brackets are comma delimited records denoted by curly braces {}.  Each record consists of properties that have a "Name": and "value".  Multiple properties are separated by commas.  So your input is almost a valid JSON array.  The commas after the last properties should be removed.  The following is a valid JSON array.

[
    {
        "First NameRow1": "John",
        "Last NameRow1": "Doe"
    },
    {
        
        "First NameRow1": "Paul",
        "Last NameRow1": "Stork"
        
    }
]

 Note that the property names are repeated for each record.  You don't want different property names.  The schema for the JSON array above is

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "First NameRow1": {
                "type": "string"
            },
            "Last NameRow1": {
                "type": "string"
            }
        },
        "required": [
            "First NameRow1",
            "Last NameRow1"
        ]
    }
}

You'll notice the section at the end called "required" this is the list of properties that each object in the array MUST have.  In your example you changed the property name for each object and listed all the variations as required.  But each object only had a few of those properties.  That's why you are getting an error.

[oappdev]

Thank you so much! @Pstork1 Will give it a try. 

[oappdev]

@Pstork1 What l'm i missing or doing wrong?

Error:

Additional text encountered after finished reading JSON content: ,. Path '', line 4, position 1.

Sorry, been under a lot of pressure.

[Pstork1]

From what I can see of the input, you've got two JSON records, but no Array.  There should be square brackets around the JSON for it to be an array.

Yours

{
 record
},
{
 record2
}

should be 

[
{
 record
},
{
 record2
}
]

[shreyans-s-jain]

I have a table with 3 properties out of which only two properties have data in it. I have using SQL connector to get the data and then passing it into a "Parse JSON" module. 

Following is the output of my SQL Connector:

{
  "@odata.context": "https://logic-apis-southeastasia.azure-apim.net/apim/sql/somethings/items",
  "value": [
    {
      "@odata.etag": "",
      "ItemInternalId": "Some alpha-numeric id",
      "userEmail": "somename@blabla.com",
      "someCode": "someCode"
    },
    {
      ...
    },
    ...
  ]
}

As you ( @Pstork1 ) explained in your solution,  For the above output, my JSON Schema is as follows for "Values" tag:

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "@@odata.etag": {
                "type": "string"
            },
            "ItemInternalId": {
                "type": "string"
            },
            "userEmail": {
                "type": "string"
            },
            "someCode": {
                "type": "string"
            }
        },
        "required": [
            "@@odata.etag",
            "ItemInternalId",
            "userEmail",
            "someCode"
        ]
    }
}

 
 When I run my Logic App, I get the following error in the "Parse JSON" connector:

[
  {
    "message": "Required properties are missing from object: someCode.",
    "lineNumber": 0,
    "linePosition": 0,
    "path": "[318]",
    "value": [],
    "schemaId": "#/items",
    "errorType": "required",
    "childErrors": []
  }
]

Initially I had 4-5 data pointers in my table and the Parse JSON connector was working fine. I inserted few more records and now I am getting this error. I can't understand what could be the problem here.. 

Please help.

[Pstork1]

Since some of the fields are null for some records you can't list them as required properties in the schema.  You need to do two things.  Edit the schema to change the type designator for those properties to allow eiter a string or a null.

            "someCode": {
                "type": ["string", "null"]
            }

and remove them from the list of required properties.

        "required": [
            "@@odata.etag",
            "ItemInternalId",
            "userEmail",
            "someCode" <- Delete this line and the comma at the end of the previous line
        ]

[Raghu3636]

if we put this condition and remove the Columns in "required 

we cannot custom select columns after parse json step. resulting with table[COLUMN1] as column header In the attachment. is there a way we can get only columns instaed of table[column names] 

[Pstork1]

Are you saying that you aren't seeing the columns as dynamic content? If so please post this as a new question because that should not be happening if you entered the values correctly as described.

[Raghu3636]

I am trying to create scheduled power automate flow using Run a query against a dataset to send an email with data in csv format. I copied the dax code from Optimize--> performance analyzer(Power BI desktop) and pasted in query text of  run a query against dataset step.

I followed below scenarios 

1) I used parse json and I tried to specify the columns in schema with syntax "required":["table[column]","table[column]"]. When I tried testing the flow I got below error 

"Message": "Required properties are missing from object: " Because my data has null values for few records. 

I also tried to change the syntax "type": ["string","null"] for each column but no luck-same error. 

2) I removed columns from "required":[ ] and created a csv file and sent an email, but in email teh column headers are Table[column1],table[column2],...........

Note: I could not use Custom select of columns while create csv table or select columns to define names in select operation  

I want only column names instead of table[column1] in the file as column header  

[Pstork1]

Please post this as a new question.  It's really a different question than the original post. You'll get more responses and others will be able to find the answer more easily later. If you send me a link to the post I'll try to help there. The key is to start with all the columns and then use a data select to keep only the columns you want.  Simply removing them from the schema won't keep them from being included in the download. It will simply keep them from causing an error if they aren't included.