% -*- mode: latex; coding: utf-8; TeX-master: ../thesis -*-
% !TEX TS-program = pdflatexmk
% !TEX encoding = UTF-8 Unicode
% !TEX root = ../thesis.tex

In the last decade, concepts from functional programming have grown in
importance within the wider, non-functional programming community.
Often it is recommended to learn a purely functional programming language
such as Haskell to become familiar with these concepts.
However, many programmers struggle with the double duty
of learning a new paradigm and a new syntax at the same time.
This paper proposes that by learning functional programming with a
multi-paradigm programming language and a familiar syntax it is possible
to lower this effort.

To achieve this goal, the programming language Go has been chosen due to
its syntactical simplicity and familiarity.
However, a downside of Go is the lack of a built-in list type, as lists take a
central role in functional programming. Although this is remediated by Go's slices,
they are not accompanied by any higher-order list processing functions --- `map', `filter', and `fold' to name a
few --- that are present in every functional programming language (and many
other languages too).
Due to the absence of polymorphism, in order to provide these higher-order functions in
a user-friendly way it is necessary to build these functions into the compiler.

Furthermore, this paper adopts a definition of pure functional programming and
introduces `funcheck', a static code analysis tool that is designed to
report constructs that are non-functional.

In conclusion, I demonstrate that with the help of the newly built-in
functions `fmap', `filter', `foldr', `foldl' and
`prepend', as well as `funcheck' to lint code, Go proves itself to be a
suitable language for getting started with functional programming.
The primary factor for this is reflected in the Go idiom `clear is better than clever'.
While functional Go code is more verbose when compared to functional languages, it
is also more obvious about its inner workings.
At the same time, it also illustrates why there is no way around learning a
language such as Haskell if fluency with functional programming concepts
is desired. The main reasons are that, although it may be unusual at first, Haskell's
syntax is extremely concise, and that the language's design --- the type system,
pattern matching, the purity guarantees and more --- provides a very effective toolset
for purely functional programming.