Why another programming language?

• It's statically-typed and type-infered.
• It also infers memory consumption and guarantees O(n) memory use.
• It is designed for concise one-liner computations right in the shell prompt.
• It features both a mathematics library and a set of data slicing and aggregation primitives.
• It is faster than all other interpreted languages with a similar scope. (Perl, Python, awk, ...)
• It is not Turing-complete. (But can compute virtually anything nonetheless.)
• It is self-contained: distributed as a single statically linked binary and nothing else.
• It has no platform dependencies.

Download

Documentation

A concise cookbook of examples.

A primer on embedding into your own program.

History

2023-08-25

Version 9.2 released. Changes:

• Comments (with the # symbol) are now finally supported.
• Build system modernization.

2022-03-22

Version 9.1 released. Changes:

• Added the -p command line option for setting the prelude.

2022-02-16

Version 9.0 released. Changes:

• Added string interpolation with `${...}` syntax.
• Defined the $ function to be index.@ by default. (For easier indexing of arrays and tuples.)
• Added the unflatten function.
• Expanded debug output when the -vv command-line option is used.
• Added another variant for the glue function.
• Fixed parsing precedence for $.

2020-12-31

Version 8.0 released. Changes:

• Added the implicit function call syntax using $. $a is equivalent to $(@, a); the function $ must be first defined.
• Added a two-argument version of if.
• Added syntactic sugar for filtering sequences; [/ a] is equivalent to [try if(a, @)].

2019-10-04

Version 7.2 released. Changes:

• Added the product, mul and add functions.
• Added the until function.
• Improved error output when an unknown function is called.

2019-06-20

Verson 7.1 released. Changes:

• Changed the sematics for array, iarray and sort functions when several arguments are passed; each argument is treated like an individual array member. This is a more 'natural' behaviour, but potentially code-breaking.
  (That is, sort(4, 1, 3, 2) is now equivalent to sort.seq(4, 1, 3, 2).)
• Added the sorted function to mimic the old behaviour of sort; with multiple arguments an array of one element (a tuple of input arguments) is returned.
  The function sorted is meant for use as an aggregator in a map.
• Added the combo function.

2019-04-19

Version 7.0 released. Changes:

• Added the pipe operator: ... The code a .. b is equivalent to @=a, b.
• Changed precendence for function calls with f.x syntax. From now on, f.x == b is parsed as (f.x) == b. Potentially code-breaking change.
• Fixed errors and crashes when an expression has no value. E.g., [@ : a=0].
• The sample function will now accept arbitrary objects, not just arrays of atomic values.

2018-09-24

Version 6.3 released: support for user-defined functions with no arguments.

2018-08-20

Version 6.2.4 released: support for arrays in has.

2017-08-04

Version 6.2.3 released: support for negative indeces in cut.

2017-06-09

Version 6.2.2 released: fix bug in recut.

2017-05-16

Version 6.2.1 released: fix crash in url_getparam.

2017-03-27

Version 6.2 released. Changes:

• Added new functions: find, findif, url_getparam.
• Added the && and || operators.
• Added split and resplit as synonyms for cut and recut.

2017-01-23

Version 6.1 released. Changes:

• Fixed bugs in optimizer.

2016-09-27

Version 6.0 released. Changes:

• Added the def [a, b, c] construct for structured access to arrays and tuples.
• Performance enhancements.

2016-09-08

Version 5.0 released. Changes:

• Added simple error handling functionality.
• Added versions of cut and recut that accept and return sequences.
• Added new functions: and, or.
• String to integer conversions will now autodetect number base.

2016-05-22

Version 4.1 released. Changes:

• Support for hexadecimal unsigned int literals.
• Added new functions: hex, uniques, uniques_estimate.
• Made hash generic, it now works on values of all types.

2016-04-24

Version 4.0 released. Changes:

• Added multithreading support! See documentation.
• Added new functions: lines, iarray, merge.
• Added a get function for arrays.
• Fixed bugs in tuple and bucket.

2016-03-04

Version 3.1 released. Changes:

• Added new functions: box, peek, a two-argument version of take.

2016-02-18

Version 3.0 released. Changes:

• Added support for computing simple tail-recursive functions. (Using the << ... >> syntax, see documentation.)
• Added a single-argument version of seq.
• Fixed bugs in count(), ngrams(), pairs(), triples().

2016-01-30

Version 2.2 released. Changes:

• Fixed build errors on Mac OS X
• Added a new function: eq
• Added documentation for embedded use.

2016-01-21

Version 2.1 released. Changes:

• Added built-in documentation and functions reference. (See the '-h' command-line option.)
• Fixed a bug in case()

2016-01-17

Version 2.0 released. Changes:

• Made maps and arrays into implicit aggregators. (See documentation.)
• Added new functions: bucket, explode, first, flip, glue, map, ngrams, pairs, second, stripe, take, triplets, while.
• Added a command-line option ('-s') to use sorted map keys. (Default is unsorted.)
• Added a loop-style version of count.
• Added a list of examples to the documentation.
• Changed hist to use lower bound instead.
• Fixed a bug in sample.

2015.12.27

Version 1.0 released.

2014.11.21

First commit.