Six String Operations in Five Languages

This is my favorite way to choose a new language for a project: Implement some code that I need to write in several. Then I ask myself, “which would I rather see and work with”?

Python

	from typing import Generator


	def clean_up(s: str) -> str:
	return next(split_into_sentences(fix_hyphenation(fix_whitespace(s))))

	def fix_whitespace(s: str) -> str:
	return s.replace("\n", " ")

	def fix_hyphenation(s: str) -> str:
	return s.replace("- ", "")

	def split_into_sentences(s: str) -> Generator[str, None, None]:
	return (ensure_ends_with_period(sentence) for sentence in s.split(". "))

	def ensure_ends_with_period(sentence: str) -> str:
	return sentence + ("" if sentence.endswith(".") else ".")

view raw string_ops.py hosted with ❤ by GitHub

Haskell

	module StringOps(cleanUp, fixHyphenation, fixWhitespace) where

	import Data.List (isSuffixOf)
	import Data.String.Utils (replace, split)

	cleanUp ∷ String → String
	cleanUp = fixWhitespace
	>>> fixHyphenation
	>>> splitIntoSentences
	>>> first

	fixWhitespace ∷ String → String
	fixWhitespace = replace "\n" " "

	fixHyphenation ∷ String → String
	fixHyphenation = replace "- " ""

	splitIntoSentences ∷ String → [String]
	splitIntoSentences = split ". " ⋙ map ensureEndsWithPeriod

	ensureEndsWithPeriod :: String -> String
	ensureEndsWithPeriod sentence =
	sentence ++ (if "." `isSuffixOf` sentence then "" else ".")

view raw StringOps.hs hosted with ❤ by GitHub

	module StringOpsSpec where

	import StringOps
	import Test.Hspec


	main :: IO ()
	main = hspec spec

	spec :: Spec
	spec = do
	describe "fixHyphenation" $ do
	it "rejoins a word" $ do
	fixHyphenation "auto- mobile" `shouldBe` "automobile"

	describe "fixWhitespace" $ do
	it "changes a newline to a space" $ do
	fixWhitespace "and\nthe story" `shouldBe` "and the story"

	describe "cleanUp" $ do
	it "handles extra text" $ do
	let input = "Relating to the state transient lodging tax; creating\nnew provisions; amending ORS 284.131 and\n320.305; prescribing an effective date; and pro-\nviding for revenue raising that requires approval\nby a three-fifths majority.\nWhereas Enrolled House Bill 2267 (chapter 818,"
	cleanUp input `shouldBe` "Relating to the state transient lodging tax; creating new provisions; amending ORS 284.131 and 320.305; prescribing an effective date; and providing for revenue raising that requires approval by a three-fifths majority."

view raw StringOpsSpec.hs hosted with ❤ by GitHub

Elixir

	import String

	defmodule StringOps do
	def clean_up(title) do
	title
	\|> fixWhitespace
	\|> fixHyphenation
	\|> splitIntoSentences
	\|> List.first
	end

	def fixWhitespace(s), do: replace(s, "\n", " ")

	def fixHyphenation(s), do: replace(s, "- ", "")

	def splitIntoSentences(a_string) do
	a_string
	\|> split(". ")
	\|> Enum.map(&ensure_ends_with_period/1)
	end

	def ensure_ends_with_period(sentence) do
	sentence <> if ends_with?(sentence, "."), do: "", else: "."
	end
	end

view raw string_ops.ex hosted with ❤ by GitHub

	import StringOps

	defmodule StringOpsTest do
	use ExUnit.Case

	test "fixHyphenation rejoins a word" do
	assert fixHyphenation("auto- mobile") == "automobile"
	end

	test "fixWhitespace changes a newline to a space" do
	assert fixWhitespace("and\nthe story") == "and the story"
	end

	test "clean_up handles extra text" do
	input = "Relating to the state transient lodging tax; creating\nnew provisions; amending ORS 284.131 and\n320.305; prescribing an effective date; and pro-\nviding for revenue raising that requires approval\nby a three-fifths majority.\nWhereas Enrolled House Bill 2267 (chapter 818,"
	assert clean_up(input) == "Relating to the state transient lodging tax; creating new provisions; amending ORS 284.131 and 320.305; prescribing an effective date; and providing for revenue raising that requires approval by a three-fifths majority."
	end
	end

view raw string_ops_test.ex hosted with ❤ by GitHub

Swift

	func cleanUp(s: String) -> String {
	return splitIntoSentences(summary: fixHyphenation(s: fixWhitespace(s: s)))[0]
	}

	func fixHyphenation(s: String) -> String {
	return s.replacingOccurrences(of: "- ", with: "")
	}

	func fixWhitespace(s: String) -> String {
	return s.replacingOccurrences(of: "\n", with: " ")
	}

	func splitIntoSentences(summary: String) -> [String] {
	return summary
	.components(separatedBy: ". ")
	.map { ensureEndsWithPeriod(sentence: $0) }
	}

	func ensureEndsWithPeriod(sentence: String) -> String {
	if let lastCharacter = sentence.characters.last {
	if lastCharacter == "." {
	return sentence
	} else {
	return sentence + "."
	}
	} else {
	return sentence
	}
	}

view raw string_ops.swift hosted with ❤ by GitHub

	import XCTest

	class Tests: XCTestCase {
	func testFixHyphenation() {
	XCTAssertEqual( fixHyphenation(s: "auto- mobile"), "automobile" )
	}

	func testFixWhitespace() {
	XCTAssertEqual( fixWhitespace(s: "and\nthe story"), "and the story" )
	}

	func testCleanUp() {
	let input = "Relating to the state transient lodging tax; creating\nnew provisions; amending ORS 284.131 and\n320.305; prescribing an effective date; and pro-\nviding for revenue raising that requires approval\nby a three-fifths majority.\nWhereas Enrolled House Bill 2267 (chapter 818,"
	let expected_output = "Relating to the state transient lodging tax; creating new provisions; amending ORS 284.131 and 320.305; prescribing an effective date; and providing for revenue raising that requires approval by a three-fifths majority."

	XCTAssertEqual( cleanUp(s: input), expected_output )
	}
	}

view raw string_ops_test.swift hosted with ❤ by GitHub

Ruby

	module StringOps
	module_function

	def clean_up(summary:)
	split_into_sentences(fix_hyphenation(fix_whitespace(summary))).first
	end

	def fix_whitespace(a_string)
	a_string.gsub "\n", " "
	end

	def fix_hyphenation(a_string)
	a_string.gsub "- ", ""
	end

	def split_into_sentences(a_string)
	a_string
	.split(". ")
	.map { \|sentence\| ensure_ends_with_period(sentence) }
	end

	def ensure_ends_with_period(sentence)
	sentence + (sentence.end_with?(".") ? "" : ".")
	end
	end

view raw string_ops.rb hosted with ❤ by GitHub

	require 'string_ops'
	include StringOps

	describe StringOps do
	describe '#fix_whitespace' do
	it 'changes newline to a blank' do
	expect( fix_whitespace "new\nline" ).to eq 'new line'
	end
	end

	describe '#fix_hyphenation' do
	it 'de-hyphenates' do
	expect( fix_hyphenation 'zip- per' ).to eq 'zipper'
	end
	end

	describe '#clean_up' do
	it "handles extra text" do
	input = "Relating to the state transient lodging tax; creating\nnew provisions; amending ORS 284.131 and\n320.305; prescribing an effective date; and pro-\nviding for revenue raising that requires approval\nby a three-fifths majority.\nWhereas Enrolled House Bill 2267 (chapter 818,"
	expect( clean_up summary: input ).to eq "Relating to the state transient lodging tax; creating new provisions; amending ORS 284.131 and 320.305; prescribing an effective date; and providing for revenue raising that requires approval by a three-fifths majority."
	end
	end
	end

view raw string_ops_spec.rb hosted with ❤ by GitHub

Racket

Thanks to Raevnos for the Racket code:

	#lang racket

	(define (clean-up s)
	(first (split-into-sentences (fix-hyphenation (fix-whitespace s)))))

	(define (fix-whitespace s)
	(string-replace s "\n" " "))

	(define (fix-hyphenation s)
	(string-replace s "- " ""))

	(define (split-into-sentences s)
	(map ensure-ends-with-period (string-split s ". ")))

	(define (ensure-ends-with-period s)
	(if (char=? (string-ref s (- (string-length s) 1)) #\.)
	s
	(string-append s ".")))

view raw string_ops.rkt hosted with ❤ by GitHub

	(module+ test
	(require rackunit)
	(check-equal? (fix-whitespace "new\nline") "new line")
	(check-equal? (fix-hyphenation "auto- mobile") "automobile")
	(check-equal? (clean-up "Relating to the state transient lodging tax; creating\nnew provisions; amending ORS 284.131 and\n320.305; prescribing an effective date; and pro-\nviding for revenue raising that requires approval\nby a three-fifths majority.\nWhereas Enrolled House Bill 2267 (chapter 818,")
	"Relating to the state transient lodging tax; creating new provisions; amending ORS 284.131 and 320.305; prescribing an effective date; and providing for revenue raising that requires approval by a three-fifths majority."))

view raw string_ops_test.rkt hosted with ❤ by GitHub