csv/src/from.coffee

CSV parser with simple api, full of options and tested against large datasets.

fs = require 'fs'
path = require 'path'
fs.exists ?= path.exists
utils = require './utils'
Stream = require 'stream'

###

Reading data from a source
==========================

The `csv().from` property provides functions to read from an external 
source and write to a CSV instance. The source may be a string, a file, 
a buffer or a readable stream.   

You may call the `from` function or one of its sub function. For example, 
here are two identical ways to read from a file:

    csv.from('/tmp/data.csv').on('data', console.log);
    csv.from.path('/tmp/data.csv').on('data', console.log);

###
module.exports = (csv) ->
  
  ###

  `from(mixed)`
  -------------

  Read from any sort of source. It should be considered as a convenient function which 
  will discover the nature of the data source to parse.   

  If it is a string, then if check if it match an existing file path and read the file content, 
  otherwise, it treat the string as csv data. If it is an instance of stream, it consider the
  object to be an input stream. If is an array, then for each line should correspond a record.

  Here's some examples on how to use this function:

      csv()
      .from('"1","2","3","4"\n"a","b","c","d"')
      .on('end', function(){ console.log('done') })

      csv()
      .from('./path/to/file.csv')
      .on('end', function(){ console.log('done') })

      csv()
      .from(fs.createReadStream('./path/to/file.csv'))
      .on('end', function(){ console.log('done') })

      csv()
      .from(['"1","2","3","4","5"',['1','2','3','4','5']])
      .on('end', function(){ console.log('done') })

  ###
  from = (mixed, options) ->
    error = false
    switch typeof mixed
      when 'string'
        fs.exists mixed, (exists) ->
          if exists
          then from.path mixed, options
          else from.string mixed, options
      when 'object'
        if Array.isArray mixed
        then from.array mixed, options
        else
          if mixed instanceof Stream
          then from.stream mixed, options
          else error = true
      else
        error = true
    csv.error new Error "Invalid mixed argument in from" if error
    csv

  ###

  `from.options([options])`
  -------------------------

  Update and retrieve options relative to the input source. Return 
  the options as an object if no argument is provided.

  *   `delimiter`     Set the field delimiter, one character only, defaults to comma.
  *   `rowDelimiter`  String used to delimit record rows or a special value; special values are 'auto', 'unix', 'mac', 'windows', 'unicode'; defaults to 'auto' (discovered in source or 'unix' if no source is specified).
  *   `quote`         Optionnal character surrounding a field, one character only, defaults to double quotes.
  *   `escape`        Set the escape character, one character only, defaults to double quotes.
  *   `columns`       List of fields or true if autodiscovered in the first CSV line, default to null. Impact the `transform` argument and the `data` event by providing an object instead of an array, order matters, see the transform and the columns sections for more details.
  *   `flags`         Used to read a file stream, default to the r charactere.
  *   `encoding`      Encoding of the read stream, defaults to 'utf8', applied when a readable stream is created.
  *   `trim`          If true, ignore whitespace immediately around the delimiter, defaults to false.
  *   `ltrim`         If true, ignore whitespace immediately following the delimiter (i.e. left-trim all fields), defaults to false.
  *   `rtrim`         If true, ignore whitespace immediately preceding the delimiter (i.e. right-trim all fields), defaults to false.
  
  Additionnaly, in case you are working with stream, you can pass all 
  the options accepted by the `stream.pipe` function.

  ###
  from.options = (options) ->
    if options?
      utils.merge csv.options.from, options
      csv
    else
      csv.options.from
  
  ###

  `from.array(data, [options])`
  ------------------------------
  
  Read from an array. Take an array as first argument and optionally 
  some options as a second argument. Each element of the array 
  represents a csv record. Those elements may be a string, a buffer, an 
  array or an object.

  ###
  from.array = (data, options) ->
    @options options
    process.nextTick ->
      for record in data
        csv.write record
      csv.end()
    csv
  
  ###
  
  `from.string(data, [options])`
  -------------------------------
  
  Read from a string or a buffer. Take a string as first argument and 
  optionally an object of options as a second argument. The string 
  must be the complete csv data, look at the streaming alternative if your 
  CSV is large.

      csv()
      .from( '"1","2","3","4"\n"a","b","c","d"' )
      .to( function(data){} )
  
  ###
  from.string = (data, options) ->
    @options options
    process.nextTick ->
      # A string is handle exactly the same way as a single `write` call 
      # which is then closed. This is because the `write` function may receive
      # multiple and incomplete lines.
      csv.write data
      csv.end()
    csv
  
  ###
  
  `from.path(path, [options])`
  ----------------------------
  
  Read from a file path. Take a file path as first argument and optionally an object 
  of options as a second argument.
  
  ###
  from.path = (path, options) ->
    @options options
    stream = fs.createReadStream path, csv.from.options()
    csv.from.stream stream
  
  ###
  
  `from.stream(stream, [options])`
  --------------------------------
  
  Read from a stream. Take a readable stream as first argument and optionally 
  an object of options as a second argument.
  
  ###
  from.stream = (stream, options) ->
    @options options if options
    stream.setEncoding csv.from.options().encoding if stream.setEncoding
    stream.pipe csv, csv.from.options()
    csv

  from