csvfmt

csvfmt

A nicer command line csv tool than cut -d , -f 1,2,3

Install

This is a command line tool, so install it globally so it's available on your path:

npm install --global csvfmt

Specifying files

Provide files on the command line to parse. csvparse gives you a JSON representation of your CSV file (Useful with jq)

$ csvfmt samples/people.csv
["name","sex","age"]
["morris","m","12"]
["jenna","f","13"]
["yarris","c","200"]

csvfmt reads CSV files from its standard input stream if no files are specified.

Formatting

--headers parses headers, and returns objects instead of arrays:

$ csvfmt --headers samples/people.csv
{"name":"morris","sex":"m","age":"12"}
{"name":"jenna","sex":"f","age":"13"}
{"name":"yarris","sex":"c","age":"200"}

...which makes it quite handy with jq:

$ csvfmt samples/people.csv --headers | jq 'select(.age | tonumber  < 100)'
{
  "name": "morris",
  "sex": "m",
  "age": "12"
}
{
  "name": "jenna",
  "sex": "f",
  "age": "13"
}

--format lets you specify sprintf style formatting:

$ csvfmt samples/people.csv --headers --format '%(age)4d %(sex)s %(name)s'
  12 m morris
  13 f jenna
200 c yarris

You can also use positional indexes:

$ csvfmt samples/people.csv --format '%2$s %1$s'
sex name
m morris
f jenna
c yarris