A smart reader for CSV files.
- Smart statistical heuristics to guess pretty much anything about your csv file, from delimiter to quotes and whether a header is present
- Handles for you the boring but dangerous stuff, like encoding detection and bom skipping if present, but also embedded new lines and quote escaping
- Uses Prismatic Schema to coerce the numerical values that have been recognised. The types and coercions can be extended by the user to dates, phone numbers, etc. and the inputs can be validated
- Designed to be both very easy to use in an exploratory way to get a quick feel for the data, and then be put into production with almost the same code
- Special tools to use the parsing features of ultra-csv inside environments where you access the data line by line as Strings, like Hadoop
ultra-csv is available as a Maven artifact from
Clojars:
In your project.clj dependencies for leiningen:
(def csv-seq (ultra-csv/read-csv "/my/path/to.csv"))More documentation forthcoming, for now you can use the quite exhaustive docstrings in the API Documentation.
Copyright © 2014-2019 Nils Grunwald
Distributed under the Eclipse Public License, the same as Clojure.