gerbil is a small programming language that tends to result in very compact code. It is a Polish (prefix) notation language where the need for parentheses and brackets, as they are normally used has been eliminated. Parentheses and brackets do appear in the language but they are not used for grouping.

gerbil offers exact arithmetic, including complex arithmetic. Whenever integers (either real or complex) are divided, they are treated as rationals and not floating point. Any arithmetic among rationals and integers (whether real or complex) is always exact. You can force a conversion to floating point by adding 0.0 to the number. Also, integers can be arbitrarily large in absolute value.

Here are some examples of what the language looks like.

."Hello World!"

=:fact->1;/.`*..1%1$.@fact5

Here's the same example with spaces and comments added. Admittedly, this is not the complete definition because it's missing the fact (by definition) that 0! = 1.

=: fact        ## assign the function to variable 'fact'
  -> 1;        ## define lambda function that takes 1 argument and no scope arguments
    /.         ## reduce a sequence to a single value by applying an operator to each element
      `*       ## the multiplication operator as an object (section) to be used by reduce (/.)
      ..1 %1   ## sequence of integers from 1 to whatever the argument is
    $          ## end function

. @fact 5      ## print the result of applying 'fact' to the number 5

This program prints 120 to standard out.

gerbil programming revolves around the idea of an expression. An expression is a stretch of code that produces a value. The code being executed is made up of operators which generally do not need to be separated by spaces, as the above factorial function example shows. An operator can be thought of as a function that takes arguments and produces a result. Numbers and variables are also considered special operators that don't take any arguments.

A number of conventions that are used in the operator table below are explained in the following table.

       | a blank in the 'input' column just means no input, but is not the same as N/A because the symbol is considered to be a normal operator

<inputs> | inputs, if any, to a function or operator that's being called indirectly <output> | output to a function or operator that's being called indirectly <symbol> | string of symbol (non-alphanumeric and non-space) characters

The following table explains the built-in operator symbols. Some of the operators have multiple meanings depending of the type of arguments they are presented with, so the same operator can appear more than once with different meanings. In the 'example' and 'result' columns, assume variable a is equal to 5.