GlobalSecondaryIndexes with projection creates the field with double quotation marks

[rbodoque]

Hello
I have a json definition like that

{
    "AttributeDefinitions": [{
        "AttributeName": "UserName",
        "AttributeType": "S"
    },
    {
        "AttributeName": "Email",
        "AttributeType": "S"
    },
    {
        "AttributeName": "LastName",
        "AttributeType": "S"
    }],
    "TableName": "CustomerService_User",
    "KeySchema": [{
        "AttributeName": "UserName",
        "KeyType": "HASH"
    }],
    "GlobalSecondaryIndexes": [{
        "IndexName": "EmailLastNameIndex",
        "KeySchema": [{
            "AttributeName": "Email",
            "KeyType": "HASH"
        },
        {
            "AttributeName": "LastName",
            "KeyType": "RANGE"
        }],
        "Projection": {
            "NonKeyAttributes": ["FirstName"],
            "ProjectionType": "INCLUDE"
        },
        "ProvisionedThroughput": {
            "ReadCapacityUnits": "1",
            "WriteCapacityUnits": "1"
        }
    }],
    "ProvisionedThroughput": {
        "ReadCapacityUnits": "1",
        "WriteCapacityUnits": "1"
    }

}

As you can see in the Projection of the GlobalSecondaryIndexes there is one field "FirstName" that needs to be projected to the index.

When I create the table by maven and then query Dynamo for describe the table it returns:

{
    Table: {
        AttributeDefinitions: [{
            AttributeName: UserName,
            AttributeType: S
        },
        {
            AttributeName: Email,
            AttributeType: S
        },
        {
            AttributeName: LastName,
            AttributeType: S
        }],
        TableName: CustomerService_User,
        KeySchema: [{
            AttributeName: UserName,
            KeyType: HASH
        }],
        TableStatus: ACTIVE,
        CreationDateTime: WedOct0505: 26: 04PDT2016,
        ProvisionedThroughput: {
            LastIncreaseDateTime: WedDec3116: 00: 00PST1969,
            LastDecreaseDateTime: WedDec3116: 00: 00PST1969,
            NumberOfDecreasesToday: 0,
            ReadCapacityUnits: 1,
            WriteCapacityUnits: 1
        },
        TableSizeBytes: 218,
        ItemCount: 1,
        TableArn: arn: aws: dynamodb: ddblocal: 000000000000: table/CustomerService_User,
        GlobalSecondaryIndexes: [{
            IndexName: EmailLastNameIndex,
            KeySchema: [{
                AttributeName: Email,
                KeyType: HASH
            },
            {
                AttributeName: LastName,
                KeyType: RANGE
            }],
            Projection: {
                ProjectionType: INCLUDE,
                NonKeyAttributes: ["FirstName"]
            },
            IndexStatus: ACTIVE,
            ProvisionedThroughput: {
                ReadCapacityUnits: 1,
                WriteCapacityUnits: 1
            },
            IndexSizeBytes: 218,
            ItemCount: 1,
            IndexArn: arn: aws: dynamodb: ddblocal: 000000000000: table/CustomerService_User/index/EmailLastNameIndex
        }],

    }
}

As you can see the projection is created with double quotation marks

Projection: {ProjectionType: INCLUDE,NonKeyAttributes: ["FirstName"]}

I think the issue is in com.jcabi.dynamodb.core.Tables in the method

private Projection projection(final JsonObject json)

in the line 290 is doing

nonkeyattrs.add(nonkey.toString());

And this causes put the double quotation.

Maybe a solution could be change to something like

JsonString nonkeyStr = (JsonString)nonkey;
nonkeyattrs.add(nonkeyStr.getString());

thanks
Rafa

[dmarkov]

@yegor256 please dispatch this issue

[yegor256]

@rbodoque I'm not sure I understand the problem. What do you expect to see there instead of "FirstName"? Also, this code:

x = (JsonString) y;
print(y.toString());

is equivalent to:

print(x.toString());

[rbodoque]

Hello @yegor256 sorry for the late response, Time ago that I'm not working with this.

I think assumption is not correct I mean =>
x = (JsonString) y;
print(y.toString());
is equivalent to:

print(x.toString());

for show you I've created this simple test https://github.com/rbodoque/-jacabitest.git

the test is the method

private static void projection(final JsonObject json) {
        final JsonObject projn = json.getJsonObject("Projection");
//        final Projection projection = new Projection()
//            .withProjectionType(projn.getString("ProjectionType"));
        final Collection<String> nonkeyattrs = new LinkedList<String>();
        if (projn.containsKey("NonKeyAttributes")) {
            for (final JsonValue nonkey
                : projn.getJsonArray("NonKeyAttributes")) {
                //nonkeyattrs.add(nonkey.toString());
                System.out.println(nonkey.toString());
                JsonString nonkeyStr = (JsonString)nonkey;
                //nonkeyattrs.add(nonkeyStr.getString());
                System.out.println(nonkeyStr.getString());
                
                
            }
           // projection.setNonKeyAttributes(nonkeyattrs);
        }
        //return projection;
    }

If both were the same it will print FirstName twice but what is printing is

"FirstName"
FirstName

I think this is because the current implementation is executing this code
https://java.net/projects/jsonp/sources/git/content/impl/src/main/java/org/glassfish/json/JsonStringImpl.java => line 93. As you can see it first add the quotation marks

What the issue is causing it creates a projection but with the "FirstName" column instead of the FirstName column