Parsing arbitrary CSV is one thing, but using text files with field- and record-delimiters is another. When used as an intermediate data format, you can often get by with naive splitting on characters (especially if you use a less common delimiter).