custom_func uses properties from two arrays of objects

[tb-artomu]

1. input data
   {
   "count":7,
   "next":"xxx",
   "previous":"xxx",
   "results":[
   {
   "order_reference":"T3008659677",
   "order_date":"2023-01-06T12:37:37",
   "test_flag":true,
   "supplier":"https://XXXX",
   "currency_code":"GBP",
   "subtotal":"1.40",
   "tax":"8.40",
   "total":"9.80",
   "items":[
   {
   "url":"https://items",
   "part_number":"RESKU1",
   "supplier_sku_reference":"SUSKU1",
   "line_reference":"1",
   "quantity":3
   },
   {
   "url":"https://items2",
   "part_number":"RESKU2",
   "supplier_sku_reference":"SUSKU2",
   "line_reference":"2",
   "quantity":1
   }
   ],
   "goods":[
   {
   "g_url":"https://goods",
   "g_part_number":"RESKU1",
   "g_supplier_sku_reference":"SUSKU1",
   "g_line_reference":"1",
   "g_quantity":3
   },
   {
   "g_url":"https://goods2",
   "g_part_number":"RESKU2",
   "g_supplier_sku_reference":"SUSKU2",
   "g_line_reference":"2",
   "g_quantity":1
   }
   ]
   }
   ]}

2. expected output：
   eg: use concat func = > iterm_url_number = items.url + “_” + g_part_number

3. How to write schema？Is there a sample?
   {
   "parser_settings": {
   "version": "omni.2.1",
   "file_format_type": "json"
   },
   "transform_declarations": {
   "FINAL_OUTPUT": {
   "object": {
   "datas": {
   "array": [{
   "xpath": "/results/",
   "object": {
   "iterms_list": { "array": [ { "xpath": "items/", "object": {
   "iterm_url": {
   "xpath": "url"
   },
   "iterm_url_number": {
   "custom_func": {
   "name": "concat",
   "args": [
   {"xpath": "url"},
   {"const": "_"},
   {"xpath": "../goods//g_part_number"}
   ]
   }
   }
   }} ] },
   "goods_list": { "array": [ { "xpath": "goods/", "object": {
   "iterm_part_number": {
   "xpath": "g_part_number"
   }
   }} ] }
   }
   }]
   }
   }
   }
   }
   }

[jf-tech]

1. next time, please use ``` the triple ticks code block so your json text/schema would be more readable. See markdown on the triple tick usage (https://docs.github.com/en/get-started/writing-on-github/working-with-advanced-formatting/creating-and-highlighting-code-blocks).

2. I'm following up with underlying xpath library omniparser utilizes, trying to see if there is a way to achieve what you want. See How to "zip" two lists together? antchfx/xpath#90

[jf-tech]

Actually, i figured out a different way by using javascript_with_context (please read the linked doc for more context and understanding).

Input JSON:

{
    "count":7,
    "next":"xxx",
    "previous":"xxx",
    "results":[
        {
            "order_reference":"T3008659677",
            "order_date":"2023-01-06T12:37:37",
            "test_flag":true,
            "supplier":"https://XXXX",
            "currency_code":"GBP",
            "subtotal":"1.40",
            "tax":"8.40",
            "total":"9.80",
            "items":[
                {
                    "url":"https://items",
                    "part_number":"RESKU1",
                    "supplier_sku_reference":"SUSKU1",
                    "line_reference":"1",
                    "quantity":3
                },
                {
                    "url":"https://items2",
                    "part_number":"RESKU2",
                    "supplier_sku_reference":"SUSKU2",
                    "line_reference":"2",
                    "quantity":1
                }
            ],
            "goods":[
                {
                    "g_url":"https://goods",
                    "g_part_number":"RESKU1",
                    "g_supplier_sku_reference":"SUSKU1",
                    "g_line_reference":"1",
                    "g_quantity":3
                },
                {
                    "g_url":"https://goods2",
                    "g_part_number":"RESKU2",
                    "g_supplier_sku_reference":"SUSKU2",
                    "g_line_reference":"2",
                    "g_quantity":1
                }
            ]
        }
    ]
}

Schema:

{
    "parser_settings":{
        "version":"omni.2.1",
        "file_format_type":"json"
    },
    "transform_declarations":{
        "FINAL_OUTPUT":{
            "object":{
                "datas":{
                    "array":[
                        {
                            "xpath":"results/*",
                            "object":{
                                "iterms_list2":{
                                    "custom_func":{
                                        "name":"javascript_with_context",
                                        "args":[
                                            {
                                                "const":"var n = JSON.parse(_node); var ret = []; for (var i = 0; i < n['items'].length; i++) { var obj = new Object(); obj.iterm_url = n['items'][i]['url']; obj.iterm_url_number = obj.iterm_url + '_' + n['goods'][i]['g_part_number']; ret.push(obj); }; ret;"
                                            }
                                        ]
                                    }
                                },
                                "goods_list":{
                                    "array":[
                                        {
                                            "xpath":"goods/*",
                                            "object":{
                                                "iterm_part_number":{
                                                    "xpath":"g_part_number"
                                                }
                                            }
                                        }
                                    ]
                                }
                            }
                        }
                    ]
                }
            }
        }
    }
}

Output by cli.sh

[
	{
		"datas": [
			{
				"goods_list": [
					{
						"iterm_part_number": "RESKU1"
					},
					{
						"iterm_part_number": "RESKU2"
					}
				],
				"iterms_list2": [
					{
						"iterm_url": "https://items",
						"iterm_url_number": "https://items_RESKU1"
					},
					{
						"iterm_url": "https://items2",
						"iterm_url_number": "https://items2_RESKU2"
					}
				]
			}
		]
	}
]

The key is the use of javascript_with_context which is the ultimate custom_func that provides you with the IDR context and enables you to do any complex logic.

The javascript code (flattened in the schema) is:

const ":"
var n = JSON.parse(_node);
var ret = [];
for (var i = 0; i < n['items'].length; i++) {
    var obj = new Object();
    obj.iterm_url = n['items'][i]['url'];
    obj.iterm_url_number = obj.iterm_url + '_' + n['goods'][i]['g_part_number'];
    ret.push(obj);
};
ret;

The code is fairly self-explanatory. Let me know if you have any further questions.

[jf-tech]

@tb-artomu shameless plug here: please consider being a sponsor to the project, one time or recurring, big or small, equally appreciated! Thanks!

[tb-artomu]

OK, Thank you for your answer.