Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Semantic CSV


A Clojure library for higher level CSV parsing/processing functionality.

The two most popular CSV parsing libraries for Clojure presently - clojure/data.csv and clojure-csv - concern themselves only with the syntax of CSV; They take CSV text, transform it into a collection of vectors of string values, and that's it. Semantic CSV takes the next step by giving you tools for addressing the semantics of your data, helping you put it in a form that better reflects what it represents.

See the introductory blog post for more information on the design philosophy.


  • Absorb header row as a vector of column names, and return remaining rows as maps of column-name -> row-val
  • Write from a collection of maps, given a header
  • Apply casting/formatting functions by column name, while reading or writing
  • Remove commented out lines (by default, those starting with #)
  • Compatible with any CSV parsing library returning/writing a sequence of row vectors
  • Transducer versions of all the main functions by requiring semantic-csv.transducers instead of semantic-csv.core
  • (UNTESTED) Cljs compatibility
  • (SOON) A "sniffer" that reads in N lines, and uses them to guess column types


Semantic CSV is structured around a number of composable processing functions for transforming data as it comes out of or goes into a CSV file. This leaves room for you to use whatever parsing/formatting tools you like, reflecting a nice decoupling of grammar and semantics. However, a couple of convenience functions are also provided which wrap these individual steps in an opinionated but customizable manner, helping you move quickly while prototyping or working at the REPL.


Assuming you're using Leiningen, just add the following to your dependencies:

Clojars Project

Note that this release is still in alpha and is subject to change. If you would like a stable release, please use 0.1.0 (and note that 0.2.0 was an accidental release and was supposed to be alpha). We're hoping to have a stable 0.2.1 release soon.


Please see for complete documentation.

Semantic CSV emphasizes a number of individual processing functions which can operate on the output of a syntactic csv parser such as or clojure-csv. This reflects a nice decoupling of grammar and semantics, in an effort to make this library as composable and interoperable as possible.

=> (require '[ :as io]
            '[clojure-csv.core :as csv]
            '[semantic-csv.core :as sc])
=> (with-open [in-file (io/reader "test/test.csv")]
       (csv/parse-csv in-file)
       (sc/cast-with {:this ->int})

({:this 1, :that "2", :more "stuff"}
 {:this 2, :that "3", :more "other yeah"})

However, some opinionated, but configurable convenience functions are also provided.

(with-open [in-file (io/reader "test/test.csv")]
    (sc/process (parse-csv in-file))))

And for the truly irreverent... (who don't need computer laziness):

(sc/slurp-csv "test/test.csv")

Writing CSV data

As with the input processing functions, the writer processing functions come in modular pieces you can use as you see fit. This time let's use clojure/data.csv:

(require '[semantic-csv.core :as sc]
         '[ :as cd-csv])

(def data [{:this 1, :that "2", :more "stuff"}
           {:this 2, :that "3", :more "other yeah"}])

(with-open [out-file (io/writer "test.csv")]
  (->> data
       (sc/cast-with {:this #(-> % float str)})
       (cd-csv/write-csv out-file)))

And again, as with the input processing functions, here we also provide a quick and dirty, opinionated convenience function for automating some of this:

(sc/spit-csv "test2.csv"
          {:cast-fns {:this #(-> % float str)}}

And there you have it.

Transducer API

There is now a transducer API available under the semantic-csv.transducers namespace. This API is now (or will soon be) in public alpha. The API is still subject to change, but we would greatly appreciate feedback.

In particular, we'd appreciate the following feedback:

  • should transducers hook up differently to io? io is sort of a "reducible" context...
  • how should the namespaces be organized? the transformer versions current live in a mirror api as the core api.

Please see for complete documentation.


Feel free to submit a pull request. If you're looking for things to help with, please take a look at the GH issues page. Contributing to the issues with comments, feedback, or requests is also greatly appreciated.


For a complete list of contributors, see the GH contributors page.

Of particular note, however:


Copyright © 2014-2017 Christopher Small

Distributed under the Eclipse Public License either version 1.0 or (at your option) any later version.


Higher level tools for working with CSV data and files








No releases published


No packages published