CSV [beta]

Linux:

Windows:

A package for working with CSV and other delimited files.

Types/functions:

• CSV.Source/CSV.Sink: Data.Source and Data.Sink types for the DataStreams.jl interface
• CSV.Options: a type that collects various parsing configurations that can be passed to CSV.Source or CSV.csv
• CSV.csv: convenience method that supports all the same options/inputs as CSV.Source; it creates a CSV.Source and then calls Data.stream!(source, Data.Table)
• Data.stream!(::CSV.Source,::CSV.Sink): method for streaming data from a CSV source to a CSV sink (csv file to csv file)
• Data.stream!(::CSV.Source,::Data.Table): method for streaming data from a CSV source to a Data.Table (Julia structure)
• Data.stream!(::Data.Table,::CSV.Sink): method for streaming data from a Data.Table to a CSV sink
• CSV.getfield{T}(io::IOBuffer, ::Type{T}, opt::CSV.Options=CSV.Options(), row=0, col=0): various custom parsing methods for types T
• CSV.readline(f::IO,q::UInt8,e::UInt8,buf::IOBuffer=IOBuffer()): custom readline implementation that accounts for potentially quoted newlines
• CSV.readsplitline(f::IO,d::UInt8,q::UInt8,e::UInt8,buf::IOBuffer=IOBuffer()): similar to readline, but also splits each field into a separate element in a Vector{UTF8String}
• CSV.countlines(f::IO,q::UInt8,e::UInt8): custom countlines implementation that accounts for potential quoted newlines

See the help documentation for any of the above for additional details (e.g. ?CSV.Options, ?CSV.read, etc.)

The package is currently in "beta", which means it's been tested, used by various and sundry whipper-snappers, but most likely has some corner cases yet to be fleshed out. Please let us know about your experience, any bugs, or feature requests here!