A string or an array of strings identifying one or more files. Uses glob so /some/dir/*.csv style patterns are supported, as is directory recursion via /some/dir/**/*.csv
options
object
An object configuring how output should be produced. See Options for more information.
callback
function
To be of the form function(err, manifest). Manifest contains a summary of the output files produced.
Options
Property
Type
Description
outputDirRootPath
string
An absolute directory path where to write output to. All missing directories will be created.
parser
object
An parser object for configuring how input CSV files should be parsed.
dirSplits
[object]
An array of of dirSplit objects
fileSplits
object
A fileSplit object
parser object
Configures how to parse incoming CSV lines. Uses csv-streamify under the bonnet.
Property
Type
Description
skipFirstLine
boolean
Should the first line of each file be ignored? Set to true if files include a header line, for example.
delimiter
string
Comma, semicolon, whatever - defaults to comma.
newline
string
Newline character (use \r\n for CRLF files).
quote
string
What's considered a quote.
empty
string
Empty fields are replaced by this value.
dirSplit object
Smithereens can break CSV files across a nested set of directories based on values defined in each line.
Property
Type
Description
columnIndex
integer
Each line of each CSV file will be parsed into an array of strings. This value identifies which value to split on.
valueToDirMap
object
A simple mapping of an expected string value (as identified by columnIndex) and the directory name that this line should be routed to.
fileSplit object
In a similar way, Smithereens can route lines to different files, based on the contents of a parsed CSV column.
Property
Type
Description
columnIndex
integer
Identifies which of the parsed string values from each CSV line should be used to determine a filename that a row should be routed to.
valueToFileMap
object
A key/value map where key is a string value that is expected via columnIndex and value is a file object.
file object
Defines which filename a CSV row should be routed to, along with some output-formatting configuration.
Property
Type
Description
filename
string
The filename which a row should be routed to. All output files will be in CSV format. Note that the .csv extension is added automatically, so don't include it here.
outputColumns
[object]
An array of outputColumn objects - each defining a column that should appear in the output file.
outputColumn object
Defines the values for each column in the output.
Property
Type
Description
name
string
The name of the column, used in the first line of the CSV output
columnIndex
integer
Identifies the column in the corresponding row of the parsed incoming CSV array to copy to the output. If columnIndex is given any type parameter is ignored.
type
string
Describes other output - hash, uuid, constant. hash generates a hash value from the contents of the corresponding input row, uuid outputs a unique id, constant output a fix value, given by the value property.
value
number or string
A fixed value to put output when the type property is constant.