A parser for transforming CSV strings into usable input.

This library does its best to support RFC 4180, however, many CSV inputs do not strictly follow the standard. There are two major deviations assumed in this library.

1. The \n character may be used instead of \r\n for new-lines.
2. The trailing new-line may be omitted.

RFC 4180 grammar, for reference, with notes.

The trailing newline is required, but we'll make it optional.

file = [header CRLF] record *(CRLF record) [CRLF]
header = name *(COMMA name)
record = field *(COMMA field)
name = field
field = (escaped / non-escaped)

There is no room for spaces around the quotes. The specification is that

escaped = DQUOTE *(TEXTDATA / COMMA / CR / LF / 2DQUOTE) DQUOTE

In this specification, fields that don't have quotes surrounding them cannot have a quote inside them because it is excluded from TEXTDATA.

non-escaped = *TEXTDATA
COMMA = %x2C
CR = %x0D ;as per section 6.1 of RFC 2234 [2]
DQUOTE =  %x22 ;as per section 6.1 of RFC 2234 [2]
LF = %x0A ;as per section 6.1 of RFC 2234 [2]

The spec requires that new lines be CR + LF but we'll let them get away with just LF if they want..

CRLF = CR LF ;as per section 6.1 of RFC 2234 [2]

All the printable characters minus the double-quote and comma, this is important above.

TEXTDATA =  %x20-21 / %x23-2B / %x2D-7E

Types

Functions