# Hello world! with limited character repetition

In a language of your choice, write a program that exactly outputs the characters Hello world! followed by a newline. The code:

• should not use any character more than twice (alphanumeric, symbol, whitespace...anything)
• should not use any external resources
• should not use any user input
• should not output anything else

An example of a valid python program:

print("He%so world!"%(2*'l'))


An example of an invalid python program (the character 'r' is used three times):

print("Hel"+chr(108)+'o world!')


Winner is whoever has the most votes after 14 days.

EDIT: A winner has been chosen! Thanks all for your work on this question!

• What about the repeated letter o in the "valid" python example. There also these:- " ' ( ) Jan 17, 2014 at 16:35
• @AdamSpeight what do you mean? The letter is only used twice, as per the specs.
– Josh
Jan 17, 2014 at 16:37
• @AdamSpeight ... which is literally the same. Jan 17, 2014 at 16:42
• I think you mean a character should not appear more than twice, that is, it should not be repeated more than once. Jan 17, 2014 at 18:15
• I guess brainfuck is out of the question... Jan 17, 2014 at 18:26

# Ruby (1.9+)

Since this is a popularity contest let's try to not use ANY of the characters from 'Hello world!' while still using other characters only a maximum of two times:

puts("S\107VsbG8gV29ybGQhCg".unpack(?m))


It's 40 chars btw.

# Bash

And this one uses unicode magic.

Notes:

• While the orignal characters appear elsewhere (unlike the ruby example), the printed string contains only non-ascii characters.
• Two from the three spaces are actually tabs, so there are no utf-8 characters that appear more than 2 times
• As binary some of the octets do appear more than 2 times, hopefully that's not against the rules. I'm trying to resolve them though.

Code:

echo '𝓗𝐞𝑙𝒍𝓸 𝓦𝗈𝖗𝖑𝘥¡'|iconv -t  asCIi//TRANSLIT


For those who don't have a proper font installed it looks like this:

Here is the hexdump:

00000000  65 63 68 6f 20 27 f0 9d  93 97 f0 9d 90 9e f0 9d  |echo '..........|
00000010  91 99 f0 9d 92 8d f0 9d  93 b8 e2 80 8a f0 9d 93  |................|
00000020  a6 f0 9d 97 88 f0 9d 96  97 f0 9d 96 91 f0 9d 98  |................|
00000030  a5 c2 a1 27 7c 69 63 6f  6e 76 09 2d 74 09 61 73  |...'|iconv.-t.as|
00000040  43 49 69 2f 2f 54 52 41  4e 53 4c 49 54 0a        |CIi//TRANSLIT.|
0000004e


You have to run it on a machine where the default charset is utf-8. I tried on an OSX10.8 using iTerm2 with the following environment:

# PHP 5.4

This uses zLib: (unfortunately it does uses the characters e and o)

<?=gzuncompress('x▒▒H▒▒▒W(▒/▒IQ▒!qh');


Hexdump:

00000000  3c 3f 3d 67 7a 75 6e 63  6f 6d 70 72 65 73 73 28  |<?=gzuncompress(|
00000010  27 78 9c f3 48 cd c9 c9  57 28 cf 2f ca 49 51 e4  |'x..H...W(./.IQ.|
00000020  02 00 21 71 04 68 27 29  3b                       |..!q.h');|
00000029


# +1

Here is the ruby 2.0 code I used to test for duplicates:

d=ARGF.read
p [d.split(//),d.unpack('C*')].map{|x|x.inject(Hash.new(0)){|i,s|i[s]+=1;i}.select{|k,v|v>2}}

• While I always appreciate Unicode cleverness, your bash attempt doesn't seem to satisfy the question's specification. "write a program that exactly outputs the characters Hello world! followed by a newline" Jan 17, 2014 at 21:07
• @FireFly: it does satisfy it. iconv with the target encoding ascii//translit will transliterate the unicode characters to basic ascii. And of course echo will add the newline, so this one fits the spec (if we don't consider the similarity of the octets) Jan 17, 2014 at 21:10
• On my system I get Hello World? instead of Hello World! Jan 17, 2014 at 21:20
• @SztupY oh, you're right, my bad. Very clever, I like it! Jan 17, 2014 at 21:21
• How many bonus points would you get if your duplicate-checker also conformed to the repeating spec? :-) Jan 17, 2014 at 22:10

You have to use more expressive languages.

# Chinese,  6  4   3 chars

喂世!


Running google translate on this produces Hello world!

(thanks @sreservoir and @Quincunx for the update)

• +1 for rule bending... "In a language of your choice..." Jan 19, 2014 at 22:54
• Love it. But how exactly is that a program? Jan 20, 2014 at 8:42
• Actually you can look at it as a program in a google translate language - for Chinese as its sub-language :) Jan 20, 2014 at 8:44
• +1. Nice idea. 世界 = Shì jiè = World. 你好 = nǐ hǎo = Hello. "World, Hello!" translates to "Hello World!". It can be golfed down some more: 你好世! Jan 20, 2014 at 21:56
• can be golfed down further to 喂世! at the expense of ... well, nothing more is really being expended. Jan 20, 2014 at 23:40

## HQ9+, 1 char

H


keeping it simple :)

• HQ9+ outputs a comma though ;)
– Josh
Jan 17, 2014 at 15:48
• I'd still keep your answer out...it is a popularity contest after all, rules are for the wind!
– Josh
Jan 17, 2014 at 15:53
• @Josh This is one of the rare events where I miss the downvote button on comments. If we throw away rules there is nothing left to justify the contest at all. Jan 17, 2014 at 16:24
• @Howard I feel dirty. I upvoted yours and Josh's comment Jan 17, 2014 at 17:03
• All print Hello World questions are actually who can submit a HQ9+ answer faster questions. Jan 18, 2014 at 5:49

# Vim command (18 keystrokes)

iHeEsc3alEscio WorRightd!Enter

Doesn't repeat any keystroke more than twice.

It kinda violates the "user input" rule since it's still the user that needs to input that sequence, but I suppose if you store it as a macro or an nnoremap beforehand it would count, since you're just running it without explicitly doing any input to make it happen.

It also requires nocompatible to be set, which may count as using external resources, so I have provided another variation below:

# Vim command (21 keystrokes)

iHeEsc3alEscio WorCtrl+[$ad!Enter This variation doesn't require nocompatible to be set, although it does work around using Esc three times by using Ctrl+[ in its place. • If you're talking about the characters in Right, Esc, and Enter, they're not literally those characters. Vim deals in keystrokes, which I am deeming equivalent to characters for the purposes of this contest (and the "don't repeat more than twice" rule). I only use the i keystroke twice, and the <Right> keystroke once. As for E, I never use it at all - but I do use <Esc> twice and <Enter> once. Jan 17, 2014 at 19:11 • Now I'm disappointed that Emacs doesn't have a M-x hello-world command... Jan 17, 2014 at 19:34 • Even if it did, that's three ls. Jan 17, 2014 at 19:35 • You should note that you have to :set nocompatible. Jan 19, 2014 at 14:15 • Amazing, but myself find using Right not very vim-y. Jan 19, 2014 at 16:11 ## C, 192 chars # # /*$$@*/ATION_[]={9.};main(BBCDDEEFFGGHJJKKLLMMPPQQRRSSUUVVWWXXYYZZabbdefgghhjjkkmpqqsstuuvvwxyyzz) {printf("He%clo \ world!%c\ ",2^7&!8.&~1|~-1?4|5?0x6C:48:6<3>2>=3<++ATION_[0],'@'^79-5);}  Since this isn't golf, I decided to have some fun and try to use every character exactly twice (while putting as few as possible in a comment, because that's boring). I realise that I didn't do terribly well, since my code contains a ton of boring "dead" code too (not in the literal sense, but in the sense of placeholder characters just used in order to fullfil the requirement). Anyway, this was surprisingly hard (but fun) with the two-character limitation, so unfortunately I couldn't come up with anything more interesting. It did give me an idea for a new problem though... (Updated per @ugoren's comment.) • Creatively impressive. I like it. – Josh Jan 17, 2014 at 22:48 • The slash \ ending lines three and four make it a line continuation, not newline. And line four is terminated by a EOF, making it legal. – Josh Jan 18, 2014 at 6:02 • I think you can easily uncomment . by using it for fractions. Also you can declare an array and use []. Jan 19, 2014 at 15:18 • @ugoren good call about .; I only thought about using it for struct access. <s>As for array declarations, I'm out of ,s due to the printf call, and I'm also out of ;s, so I'm not sure how I could declare one.</s> Duh, I could replace ATION_... Jan 19, 2014 at 15:29 # Piet-- No characters whatsoever! • I would consider the codels as "characters". Jan 17, 2014 at 21:16 • @PaŭloEbermann What would you consider to be a unique codel? EG different shapes? colors? sizes? some combination thereof? Jan 19, 2014 at 19:19 • @ValekHalfHeart the definition of a unique codel is given in the specs of the language: it s a block of codel that could be simplified as a single colored pixel. Mar 7, 2014 at 9:30 • @plannapus In that case I don't this challenge is technically possible in Piet. I may come up with something creative though. Mar 7, 2014 at 17:57 • @ValekHalfHeart oh it wasn t a criticism of this answer: i love it! This challenge was not a codegolf anyway. Mar 8, 2014 at 7:44 Perl, 29 characters This answer includes x-rated clogs! say"07-8*d<#B+>!"^xRATEDkL0GZ  Perl, 23 characters Shorter, but no porno shoes. :-( Same basic idea though... say'0@@lo World!'^"x%,"  Perl, 20 characters Boring... say"Hello Wor\x6Cd!"  • Um....kinky? +1 :) – Josh Jan 18, 2014 at 14:32 ## Powershell, 20 "He$('l'*2)o world!"


# Sclipting, 11 characters

丟낆녬닆묬긅덯댦롤긐뤊


I saw this beautiful HelloWorld program on esolang's Hello World program list.

• Whoa! Congratulations, you are the first person ever to post anything about any of my esolangs outside of esolangs.org :) (or at least the first I found out about) Jan 20, 2014 at 12:30

# Scala: 34 29 characters

I'm proud of myself for this one:

printf("He%c%co world!\n",108,108)


Had a really hard time overcoming duplicate 'l's, 'r's, quotation marks and brackets. Then I discovered the old Java printf function, which will happily convert numbers to letters when given the %c format specifier.

### Update

MrWonderful did a wonderful thing by pointing out that a whole bunch of characters can be saved by using up my second 'l' manually in the string!

printf("Hel%co world!\n",108)

• @KCaloux, Since you are allowed up to 2 'l's, wouldn't printf("Hel%co world\n",108) at 28 be even better? Jan 17, 2014 at 19:11
• @MrWonderful I think you're absolutely correct! (Also I just realized that I forgot to include the '!') Jan 17, 2014 at 19:17
• From what I understand, this isn't a valid entry, though a good attempt at it. printf contains a r as does world. Same goes for the letter o which is used more than once. This is based on my interpretation of the following statement from the OP "An example of an invalid python program (the character 'r' is used three times): print("Hel"+chr(108)+'o world!')" Jan 18, 2014 at 9:58
• @JAnderton I had a ruby program parse out my script to make sure there were no characters included more than twice. Read it again. There are 2 rs, not 3. One in "printf" and one in "world". The reason the python one is invalid is because it includes chr Jan 18, 2014 at 16:36

## Python 3 [38 bytes]

exec('import '+chr(95)*2+"h\x65llo__")


I wouldn't consider import __hello__ as an external resource.

# Perl: 34 characters

$_="He12o wor3d! ";s{\d}{l}g;print  Sample run: bash-4.1# perl -e '$_="He12o wor3d!
> ";s{\d}{l}g;print'
Hello world!


(Not a big deal. Posted just to use at least once in my life s/// with those fancy delimiters.)

# PHP, 33 chars

I just love how much PHP is forgiving and understanding!

<?=Hel.str_rot13("yb jbe").'ld!';


Before it was deleted (or if it's still there, I'm totally blind), I saw a comment saying "No brainf*ck? :D". Well, it is pretty much impossible to write a solution in BrainF*ck, as you know. But I managed to code this, just for the lulz.

++++++++++[>++++>+++++>++++++>+++++++>++++++++++>+++++++++++>+++++++++<<<<<<<-]
>>>.+++.--.>++.>+.+++++++.<<<<++++++.>>>>>+++++.+.--.>+++++.<.---.+++++.<<<<-.+
+.<------.------.>>>>>+++++.<----------.<<<<--.>>>>++++++++.--------.+++.
<<<<++.+++++++.+++++.-------.>>>>+++++++.--------.<<<<------.++++++.>++++++++.


If you don't have a BF interpreter, the code above just prints the PHP one :P

• I didn't know even a code like this is valid! Php tag isn't closed and string Hel isn't surrounded by quotes. Also I've never heard about str_rot13 before. Jan 17, 2014 at 18:56
• @AycanYaşıt the closing php tag is not required on EOF, and if you put a string without quotes it assumes it is an undeclared constant with the same content as its name and gives a warning Jan 17, 2014 at 19:25
• You can avoid the warning (and save the first .) by putting the Hel before the <?=. Jan 17, 2014 at 21:23
• @PaŭloEbermann Hel<?=str_rot13("yb jbe")?>ld! Jan 17, 2014 at 23:47
• @PaŭloEbermann I would say that using that is like cheating, since it's not pure PHP anymore. Jan 18, 2014 at 0:43

# HTML Fiddle - 21 characters

Hel&#108;o World!<br>

• Does that include the trailing newline?
– Josh
Jan 17, 2014 at 18:32
• Doh...missed that requirement! I'll add a <br> in there. Jan 17, 2014 at 18:36
• @netinept :It asked for a newline, not a \n so I'd say <br> counts Jan 17, 2014 at 22:54
• But the character r is repeated in the solution. Did I misinterpret the requirements? Jan 18, 2014 at 9:59
• "l" and "o" too, but each character can be used twice at most.
– xem
Jan 18, 2014 at 14:04

# Ruby: 27 characters

puts [:Hel,'o wor',"d!"]*?l


Sample run:

bash-4.1# ruby <<ENDOFSCRIPT
> puts [:Hel,'o wor',"d!"]*?l
> ENDOFSCRIPT
Hello world!


# Ruby: 25 characters

(Based on Vache's comment.)

puts 'He'+?l*2+"o world!"


# Ruby: 23 characters

(Copy of Danko Durbić's Powershell answer.)

puts"He#{?l*2}o world!"

• puts 'He'+'l'*2+'o world!' is one character shorter! Jan 17, 2014 at 16:38
• But has 6 “'”'s… Jan 17, 2014 at 16:50
• haha I was so focused on letter characters that I never noticed that. never mind! Jan 17, 2014 at 16:53
• Remove the space after puts and make it 23. -- puts"He#{?l*2}o world!" Jan 17, 2014 at 17:52
• @SampritiPanda, p includes quotes in the output. I prefer to keep strictly with the required output format. But you are right, the space is not needed. Thank you. Jan 17, 2014 at 18:17

# C - 43 Characters

main(){printf("Hel%co World!%c",'k'+1,10);}


## Output

Hello World!


## Character Counts

' ' Count: 1    '!' Count: 1    '"' Count: 2    '%' Count: 2    ''' Count: 2
'(' Count: 2    ')' Count: 2    '+' Count: 1    ',' Count: 2    '0' Count: 1
'1' Count: 2    ';' Count: 1    'H' Count: 1    'W' Count: 1    'a' Count: 1
'c' Count: 2    'd' Count: 1    'e' Count: 1    'f' Count: 1    'i' Count: 2
'k' Count: 1    'l' Count: 2    'm' Count: 1    'n' Count: 2    'o' Count: 2
'p' Count: 1    'r' Count: 2    't' Count: 1    '{' Count: 1    '}' Count: 1

• Don't you need main etc? Jan 17, 2014 at 19:28
• @FireFly I guess so! It did say to write a program. Jan 17, 2014 at 19:36
• That's C but not C++. C++ does not have implicit int (and you can't spare another i). Jan 17, 2014 at 19:52
• @BenVoigt Okie dokie! Jan 17, 2014 at 19:56

## JavaScript, 66 characters

alert('Hel'+window.atob("\x62G8gd29ybGQhCg=="));//rH+in.\x689yQhC;


Inspired by FireFly, every character used by this code is used exactly twice.

• I think you are allowed to use some characters just once, and you can drop the comment. Jan 18, 2014 at 3:19
• @JanDvorak - well, sure he could have done it that way - but I think this solution is deserving of an upvote for the sheer bloodymindedness of using each and every character exactly twice. :-) Jan 19, 2014 at 1:22
• +1, but it's easy to use each character exactly twice if you just add gibberish as a comment. Jan 19, 2014 at 7:32

## JavaScript [37 bytes]

alert(atob("SGVsbG8g")+'wor\x6cd!\n')


Too primitive isn't it?

• Where’s the newline? Jan 17, 2014 at 20:26
• @ChristopherCreutzig Sorry, forgot it. Now it is in place. Jan 19, 2014 at 0:46

# nginx.conf

return  200 "He&#x6clo wo&#x72ld!\n";


In action:

% curl -6 http://localhost/ | lynx -dump -stdin
% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
100    21  100    21    0     0  20958      0 --:--:-- --:--:-- --:--:-- 21000

Hello world!

%

• There are 4 spaces in this solution...
– Josh
Jan 17, 2014 at 20:54
• @Josh, actually, there are 5. But couldn't you consider that two of the first four are tabs, and then the last one is a non-breakable space? :-)
– cnst
Jan 17, 2014 at 20:58
• That works for me!
– Josh
Jan 17, 2014 at 21:00
• @Josh, actually, now that we're using HTML, no more need for set. This only has 3 spaces now, any one of which could be a tab!
– cnst
Jan 17, 2014 at 21:13
• Very nicely done.
– Josh
Jan 17, 2014 at 21:15

# Emacs Command (15 keystrokes)

He<Ctrl-3>l<Left>o wor<End>d!<Enter>


If that vim answers is legal then this must be too :)

Joking aside, macro it can become too :)

More nonsense aside, I can probably squeeze some more, but this seems to be good enough for the time being (since it beats vim).

P.S., please ignore all my nonsense (I (rarely, but)use vim too!).

• I still think vim > emacs. Jan 18, 2014 at 5:07

# Befunge 98

a"!dlrow ol":'e'Hck,@


Here is a version where every character appears twice.

bka!e:dw"H@!dlrow  ol":'e'Hbk,a,@


Leaves a bunch of junk on the stack.

As a bonus, every single character has something done with it by the IP.

## Malbolge

(=<#9]~6ZY32Vw/.R,+Op(L,+k#Gh&}Cdz@aw=;zyKw%ut4Uqp0/mlejihtfrHcbaC2^W\>Z,XW)UTSL53\HGFjW

• Unfortunately, this seems to print Hello, world. instead of Hello world!\n :( Try it online! Impressive, but invalid. Feb 19, 2018 at 8:50

Actually I don't like cheating :P

Python

print("!dlrow os%eH"[::-1]%('l'*2))

• You should specify that it's for python2. In python3 the you get the reversed string as output and then a TypeError. Jan 18, 2014 at 7:09
• For use in Py3K argument of print must by enclosed in parenthesises. print("!dlrow os%eH"[::-1]%('l'*2)) work in both (Python2 and Py3K) versions.
– AMK
Jan 19, 2014 at 16:24

## GolfScript

'He

o world!'n/"l"*


Substitutes two newlines (fortunately the third one, needed for the substitution, is provided by the built-in n), using both types of string literal to avoid quadruplicate copies of a quote mark. Since l is the only character which occurs more than twice in the original string, it's not too hard.

• You forgot the "!". Jan 17, 2014 at 16:27
• @Howard, missed it. Oops. Jan 17, 2014 at 16:31
• The letter o is repeated also Jan 17, 2014 at 16:34
• @AdamSpeight: There is only twice the letter o. The rules say that it's not allowed to have a character more than twice. Jan 17, 2014 at 16:35

Hmm.

In C, given these rules, we can only have one #define (because of i and n) and at most two function calls OR definitions (( and )).

I presume there's pretty much only one way to do it (though I'm probably wrong):

main(){puts("Hello w\x6fr\154d!");}

• You still need to output the trailing newline. Doing this challenge in C is difficult...
– Josh
Jan 17, 2014 at 16:55
• @Josh I just have to use puts() instead of printf(). Jan 17, 2014 at 16:59
• But what's wrong with o in world? Jan 17, 2014 at 17:01
• @VisioN In an old draft of the same code (before I realized how hard it actually was to write such a program) I used another 'o' elsewhere. But this isn't code-golf, so I guess it doesn't have to be fixed. Jan 17, 2014 at 17:08
• In fact, I'm not sure why every answer is displaying a character count. Jan 17, 2014 at 17:10

### BASH

printf 'Hel\x6co world!\n'


Cred @manatwork

echo $'Hel\x6c\x6f world!'  • The first one has 3 “e”'s and 3 “o”'s. Jan 17, 2014 at 18:42 • @manatwork: Gag; thanx ;) Jan 17, 2014 at 18:43 • In Bash you may skip -e by using the special $'…' syntax: echo \$'Hel\x6c\x6f world!'. Jan 17, 2014 at 18:48

# C - 46 Characters

 main(){printf("He%clo wor%cd!\x0d",'l',108);} 

Prints out:

Hello world!


• If you're aiming for golf, main(){printf("He%clo world!%c",108,10);} should work, and saves you a few chars. Jan 17, 2014 at 19:23
• @FireFly you're right, you'd save me 3 characters. Your suggestion works perfectly, too. Jan 17, 2014 at 19:43

# PHP

32 Chars

Note how I am not using a character more than twice, since l != L

He<?=strtolower("LLO WOR");?>ld!


Also note that, despite of Stack Overflow deleting it in the representation, there's a line break after the !.

• You are using three spaces, but I guess this can be fixed removing the ones near the PHP tags! :P Jan 18, 2014 at 1:05
• That was fast, although I realized before of reading your comment (from another answer actually) Jan 18, 2014 at 1:07
• Isn't the output "H" supposed to be uppercase? Jan 18, 2014 at 2:47
• Completely true, missed that also ): Jan 18, 2014 at 2:49

# XQuery, 19 chars

"Hello Wor&#x6C;d!"


## GolfScript, 21 characters

'He'[108]+"lo world!"


108 is the ASCII code for l.

First, I push He on the stack. Then, He gets popped and becomes Hel. Then I push lo world!` on the stack. Now there are two elements on the stack. Because at the end of a GolfScript program, everything of the stack is outputted, this program outputs:

Hello world!

followed by a newline, because Golfscript always outputs a newline.