purescript-csv

Simple configurable PureScript CSV parser.

Based on purescript-parsing. Parse list of rows (getting a simple list of fields) or using header row (getting a map from field name to value).

Example (copy into your REPL)

import Prelude

import Text.Parsing.CSV (defaultParsers, makeParsers)
import Text.Parsing.Parser (runParser)

import Data.List
import Data.Either

let isTrue exp = either (\_-> false) (== (exp :: List (List String)))

let testFile = "a,,c,\n,1,2,3\n\"x\",\"field,quoted\",z\n" :: String
let testResult = toList $ toList <$> [["a", "", "c", ""], ["", "1", "2", "3"], ["x", "field,quoted", "z"], [""]]

let parseResult = runParser testFile defaultParsers.file
parseResult
isTrue testResult parseResult

:t runParser testFile defaultParsers.file`

If you want to create your own parsers using different field seperators, differnt quotation chars and different line endings (instead of using the defaultParsers) you can do that by

excelParsers = makeParsers '\'' ";" "\r\n"

This will generate a value of type

type Parsers a =
{
  quoted :: (P a -> P a),
  chars :: P String,
  qchars :: P String,
  field :: P String,
  row :: P (List String),
  file :: P (List (List String)),
  fileHeaded :: P (List (M.Map String String))
} 

In this specific case using ' as a quotation character a ; as a field seperator and using windows \r\n line seperators instead of the *nix \n.

The easiest and fastest way of course is just to use the defaultParsers who are defined as defaultParsers = makeParsers '"' "," "\n".