csv2json

Turns a CSV into a JSON file

Installation:

$ cargo install csv2json

Usage:

$ csv2json --in <csv file> > <json file>

CSV Delimiter

By default, the csv is split by commas. If your csv is delimited in a different way, you can specify the character using the --delimiter or -d option

Eg:

colon:delimited
one:two

Without specifying:

[
  {
    "colon:delimited": "one:two"
  }
]

Using -d :

[
  {
    "colon": "one",
    "delimited": "two"
  }
]

Dimensional Seperator

If your CSV contains multidimensional data, you can add use the dimensional separator argument -d

Eg:

name.first,name.last,age
Daniel,Mason,not telling

Without using the separator:

[
  {
    "age": "not telling",
    "name.first": "Daniel",
    "name.last": "Mason"
  }
]

Setting the separator -d .:

[
  {
    "name": {
      "first": "Daniel",
      "last": "Mason"
    },
    "age": "not telling"
  }
]

Arrays

You can use --arrays (or -a) with -d to break items into arrays

name,pets.1,pets.2
Daniel Mason,Yuki,Tinky

Without using arrays:

[
  {
    "name": "Daniel Mason",
    "pets.1": "Yuki",
    "pets.2": "Tinky"
  }
]

With arrays (-d . -a):

[
  {
    "name": "Daniel Mason",
    "pets": [
        "Yuki",
        "Tinky"
    ]
  }
]

Note: The number of the key is irrelevant, it only need be a number for example:

name,pets.45,pets.22
Daniel Mason,,Tinky

Will produce:

[
  {
    "name": "Daniel Mason",
    "pets": [
        "",
        "Tinky"
    ]
  }
]

Remove Empty Strings

You can remove empty strings from objects and arrays with the --remove-empty-strings flag.

Note: this happens for both objects and arrays, which may have undesirable affects.

name.first,name.last,age,pets.1,pets.2
daniel,,34,,

$ csv2json --in test.csv -d . -a --remove-empty-strings

[
  {
    "age": "34",
    "name": {
      "first": "daniel"
    },
    "pets": []
  }
]

Remove Empty Objects

You can remove empty objects from objects and arrays with the --remove-empty-objects flag.

Note: this happens for both objects and arrays, which may have undesirable affects.

name.first,name.last,pets.1.name,pets.1.type,pets.2.name,pets.2.type
james,smith,,,,
daniel,mason,yuki,cat,tinky,cat

$ csv2json --in test.csv -d . -a --remove-empty-strings --remove-empty-objects

[
  {
    "name": {
      "first": "james",
      "last": "smith"
    },
    "pets": []
  },
  {
    "name": {
      "first": "daniel",
      "last": "mason"
    },
    "pets": [
      {
        "name": "yuki",
        "type": "cat"
      },
      {
        "name": "tinky",
        "type": "cat"
      }
    ]
  }
]

Output to directory

Using the --out-dir <dir> to write the .json file to the output dir. It will use the name of the original file so --in /some/dir/my-data.csv --out-dir /some/other/dir will produce the file /some/other/dir/my-data.json.

Output to files based on names

Using the --out-name <template> with --out-dir <dir> to write multiple files of json using the template to generate their name from the original data. For example

Given test.csv

name.first,name.last,pets.1.name,pets.1.type,pets.2.name,pets.2.type
james,smith,suki,cat,,
daniel,mason,yuki,cat,tinky,cat

Running csv2json with the following naming template

$ csv2json --in test.csv --out-dir . --out-name "{name.first}-{name.last}" -d . -a --remove-empty-strings --remove-empty-objects

Will produce the following files

james-smith.json

{
  "name": {
    "first": "james",
    "last": "smith"
  },
  "pets": [
    {
      "name": "suki",
      "type": "cat"
    }
  ]
}

daniel-mason.json

{
  "name": {
    "first": "daniel",
    "last": "mason"
  },
  "pets": [
    {
      "name": "yuki",
      "type": "cat"
    },
    {
      "name": "tinky",
      "type": "cat"
    }
  ]
}

Types

Booleans

You can specify a column contains a boolean value by using the --boolean option

type,option.a,option.b,option.c,option.d
true,1,true,anything,TRUE
false,0,false,,FALSE

$ csv2json --in test.csv -d . --boolean option.a --boolean option.b --boolean option.c --boolean option.d
[
  {
    "option": {
      "a": true,
      "b": true,
      "c": true,
      "d": true
    },
    "type": "true"
  },
  {
    "option": {
      "a": false,
      "b": false,
      "c": false,
      "d": false
    },
    "type": "false"
  }
]

Numerics

You can specify a column contains a numeric value by using the --numeric option

number
0
1
-1
1.0

$ csv2json --in test.csv --numeric number
[
  {
    "number": 0
  },
  {
    "number": 1
  },
  {
    "number": -1
  },
  {
    "number": 1.0
  }
]