fields override flatten columns

1. json2csv: "^5.0.5"
2. node --version v12.18.2
3. Include the command or code you used. const { Parser, transforms: { flatten } } = require('json2csv')

  const json2csvParser = new Parser({
    fields,
    transforms: [
      flatten({
        separator: '_',
        objects: true,
        arrays: true,
      }),
    ],
  })

const csv = json2csvParser.parse(mydata)

1. Include a sample dataset that we can test against.

[
{
    new_cases: 1,
    total_cases: 118,
    province: 'New Brunswick',
    info: [
      {
        Mon: '0',
        Mon1: '0',
        Mon2: '0',
        Mon3: '0',
        Mon4: '0',
      },
      { Tue: '1' },
      { Wed: '1' },
      { Thu: '3' },
      { Fri: '2' },
    ],
  },
  {
    new_cases: 1,
    total_cases: 26,
    province: 'PEI',
    info:{
      Mon: '0',
      Tue: '1',
      Wed: '1',
      Thu: '3',
      Fri: '2',
    },
  },
  {
    new_cases: 1,
    total_cases: 5,
    province: 'Northwest Territories',
    info: { 0: 29, 1: 18, 2: 11, 3: 7, 4: 12, 5: 23, 6: 58, 7: 72, 8: 78, 9: 81, 10: 80, 11: 65, 12: 66, 13: 61, 14: 57, 15: 52, 16: 44, 17: 49, 18: 40, 19: 50, 20: 59, 21: 68, 22: 55, 23: 45 },
  },
]

1. Include your output/error. not really an error, more like a behavior question: if my fields are specified, the flatten fields are not included. If I say my fields =[ new_cases, total_cases, province, info ] (so all of them), even though I included the info field, because it is “destructured” into other fields by flatten, those transformed fields won’t be included in my csv because I didn’t specify them in the fields portion. So if I remove one column from my fields, still not being the one with the nested json like fields =[ total_cases, province, info ] the same happens, I get my info field blank and I don’t get the “destructured” columns from the nested json generated by flatten

Is there something I’m doing wrong or an option I should be using? I’d like to be able to choose the columns I wanna include in my csv but if the chosen column has a nested json and I have passed the flatten option, I’d like the transformed fields to be included in my csv.

[tlowande]

oh my 🤦‍♀️ , you’re ARE right! I read that so many times but I couldn’t combine it with the flatten concept in my mind, since it generates those columns on the go and my JSON values don’t all have the same structure. Thanks for an awesome and prompt reply

[juanjoDiaz]

Yup. It should work just the same as the sync parser.

I just run your code with your code sample and I got the flattened CSV as expected.