Print invisible text

Question

Given a string as input, output a number of whitespace characters (0x0A and 0x20) equal to the length of the string.

For example, given the string Hello, World! your code would need to output exactly 13 whitespace characters and nothing else. These can be any mix of spaces and newlines.

Your code should not output any additional trailing newlines or spaces.

Testcases:

 Input -> Amount of whitespace to output
"Hello, World!" -> 13
"Hi" -> 2
" Don't
Forget about
Existing
Whitespace! " -> 45
"" -> 0
" " -> 13
"
" -> 1

Scoring:

This is code-golf so fewest bytes wins!

• manatwork: I don't get what you mean with that "0x0A". Where should that be output? Should that be kept, so "a␠b␊c" becomes "␠␠␠␊␠"?
• Mayube: @manatwork 0x0A and 0x20 are the hexadecimal values for the Newline and Space characters respectively
• manatwork: "output a number of whitespace characters (0x0A and 0x20)" – Where in the output should those newline characters be?
• Mayube: These can be any mix of spaces and newlines Your output can be any mix of spaces and newlines, you can just output spaces if you want, like everyone else, or you can just output newlines. It's up to you
• Luis Mendo: Can we assume the input will only have printable characters?

Answer 1

Whitespace, (削除) 311 (削除ここまで) (削除) 150 (削除ここまで) (削除) 77 (削除ここまで) (削除) 68 (削除ここまで) (削除) 65 (削除ここまで) (削除) 46 (削除ここまで) (削除) 41 (削除ここまで) 38 bytes

^{-3 bytes thanks to Kevin Cruijssen
-27 bytes thanks to Ephphatha}

Try it online!

A visible format

'\n \n \t \t \n \n \n \t\n\t \t\t\t\n\t \n\t\n \n \n\n'

Explanation (s = space, t = tab, n = new line)

nssn # label(NULL) - loop start
ssststsn # push 10 in the stack -> [10]
sns # duplicate the top of the stack -> [10, 10]
sns # duplicate the top of the stack -> [10, 10, 10]
tnts # read a single char from input, pop the stack and store at that address -> [10, 10] [10:<char that was read>]
ttt # pop the stack and put the value at that adress on stack -> [10,<char>] [10:<char>]
ntssn # jump to label(0) - since label(0) is not defined, the interpreter jumps to the end of the program - if the top of the stack (char) is 0 -> [10] [10:<char>]
tnss # pop the top of the stack and print as ascii -> [] [10:<char>]
nsnn # go back to the label(NULL)

• Mayube: Assuming this does actually work, this definitely wins my vote for most creative answer
• Erik the Outgolfer: Wait where is the answer? Is it invisible too?
• user63187: WHAT BLACK MAGIC IS THIS. Your code is not even there! -1
• Rod: @Christopher more like WHITEspace MAGIC
• Draco18s no longer trusts SE: I knew someone would answer this question with a whitespace program

Answer 2

Japt, 1 byte

ç

Try it online!

• Mayube: Does japt seriously have a builtin for this? Damn...
• Tom: @Mayube well it has a builtin to replace all characters in a string with another, and the default replacement is a space ;)
• Oliver: Very nice! For those running the program, you can add the -Q flag into the input to put quotes around the output. TIO

Answer 3

Haskell, 7 bytes

(>>" ")

Try it online! Usage: (>>" ") "Hello, world!".

Given two lists (and strings are lists of characters in Haskell) the >> operator will repeat the second list as many times as the first list has elements. Setting " " as second argument means we concatenate as many spaces as the input string is long.

---

Alternative (same byte count):

(' '<$)

Try it online! Usage: (' '<$) "Hello, world!".

Given some value and a list, the <$ operator replaces each value in the list with the given value. Thus 5 <$ "abc" results in [5,5,5], and ' ' <$ "abc" in " ".

The function can equivalently be written as (<$)' ', in case you want to find some more marine creatures in my code.

• Taylor Raine: Its like an adorable little finless fish
• Connor D: dang this blows my haskell answer out of the water haha

Answer 4

brainfuck, 18 bytes

++++++++++>,[<.>,]

Try it online!

Prints one newline for each byte of input. Printing spaces instead would add 4 bytes.

• lynn: For posterity: Mika Lammi posted a clever 16-byte answer that got buried. ,[>++++++++++.,]

Answer 5

Python, 19 bytes

lambda s:' '*len(s)

Answer 6

Brainfuck, 16 bytes

Prints newlines.

,[>++++++++++.,]

Answer 7

Retina, (削除) 3 (削除ここまで) 4 bytes

S\`.

Old version, doesn't work because Retina prints a trailing line feed.

.
 

(The second line contains a space).

• Digital Trauma: The retina TIO is quite easy to use. Here is your answer
• Martin Ender: Unfortunately, Retina prints a trailing linefeed by default. You'll need to prepend \` to avoid that. Then it's shorter to use S\`. though, which replaces each character with a linefeed (because it splits the input around each character).
• TheLethalCoder: @MartinEnder Ahhh wasn't sure if that was a Retina or TIO thing. Thanks for the help on saving a byte there though!

Answer 8

sed, 7 bytes

s/./ /g

Try it online!

Answer 9

C#, (削除) 28 (削除ここまで) 24 bytes

s=>"".PadLeft(s.Length);

Old version using the string constructor for 28 bytes:

s=>new string(' ',s.Length);

• LiefdeWen: Wanted to do exactly the same
• TheLethalCoder: @StefanDelport Gotta be quick with C# when I'm around :) There's Linq approaches to do the same but they're all a LOT longer...
• Bbrk24: Given adequate type context, couldn't this just be s=>new(' ',s.Length)?

Answer 10

JavaScript ES6, 22 bytes

a=>a.replace(/./g," ")

f=a=>a.replace(/./g," ");
var test = f("Hello, World!");
console.log(test, test.length);

• ETHproductions: Huh, I thought "oh darn, it'd have to be s=>s.replace(/[^]/g," "), a byte longer than the other solution". It didn't occur to me that newlines are allowed in the output :P

Answer 11

Retina, 5 bytes

\`.
¶

Try it online! Changes everything into newlines. The \` suppresses the extra newline Retina would normally output.

Answer 12

Mathematica, 21 bytes

StringReplace[_->" "]

• CalculatorFeline: If charlist input was allowed, this could be #/._->" "&. Sadly, the input is a string and Characters[] makes it one byte longer than your solution :(
• Ian Miller: Doesn't this need a # and a & in it? E.g. StringReplace[#,_->" "]&
• alephalpha: @IanMiller Not in Mathematica 10.4 or 11. reference.wolfram.com/language/ref/StringReplace.html
• Ian Miller: Ah ok. I only have 10.3. Maybe time to upgrade...

Answer 13

C, 31 bytes

f(char*c){puts(""),*c++&&f(c);}

• Tas: How does this differ from your other C answer? Clearly this one is shorter, but should you have simply edited the other one? Should it just be one answer with two solutions?
• sigvaldm: @Tas First of all, I guess in some sense I feel this is not as good as the other one eventhough it's shorter, because it doesn't actually compile as-is. It's just a function so you need to write some main routine around it. However, it is shorter and others seems to post just functions. Clearly it's two very different solutions. One is not the refinement of the other, so to me it makes sense that it should be two different answers. However, I'm new to this community. Is the consensus that one user only posts one answer? If so I will do that next time.
• Oskar Skog: Should the comma really be a comma and not a semicolon?
• cat: @OskarSkog well, in this case it doesn't matter that much because there is no lhs
• sigvaldm: @OskarSkog Yes, it should be a comma. As @cat says, it doesn't really matter in this case but I chose comma for variation :) The comma operator evaluates two expressions (e.g. i++, j++ in a for loop) and returns the rightmost one. An important detail is that the recursion has to stop somehow. && doesn't evaluate it's rhs if it's lhs is false. *c++ evaluates false when it points to the null-termination of the string.

Answer 14

PHP, 28 Bytes

for(;a&$argn[$i++];)echo" ";

Try it online!

PHP, 29 Bytes

<?=str_pad('',strlen($argn));

Try it online!

Answer 15

Excel VBA, _{(削除) 17 (削除ここまで)} 15 Bytes

Anonymous VBE immediate window funtion that takes input from cell [A1] and outputs spaces of length of the input to the VBE immediate window

?Spc([Len(A1)])

Answer 16

05AB1E, 3 bytes

vð?

Try it online!

v # For each character...
 ð? # Output a space without a newline

Other 3 byte solutions (Thanks Magic Octopus Urn and Kevin Cruijssen for most of these)

v¶? # For each character print a newline (without adding a newline)
võ, # For each character print the empty string with a newline
g×ばつ # Get the length, concatenate that many copies of space
g×ばつ # Get the length, concatenate that many copies of newline
Sð: # Split, push a space, replace each char in input with a space
ðs∍ # Push ' ', swap, make the string of spaces as long as the input was
võJ # For each char, push a space and ''.join(stack)
v¶J # For each char, push a newline and ''.join(stack)
€ðJ # For each char, push a space. Then ''.join(stack)
ۦJ # For each char, push a newline. Then ''.join(stack)

• Magic Octopus Urn: Other solution: gð×, the rest I came up with were above 3 like: õ‚.B¤
• Magic Octopus Urn: Another fun one: Sð:
• Magic Octopus Urn: More fun: ðs∍
• Kevin Cruijssen: Some more alternative 3-byters: võJ/v¶J; €ðJ/ۦJ. And since a sequence of characters as I/O is allowed by default when strings I/O is asked, some 2-byte versions are possible: €ð/ۦ/εð/ε¶ and ð:/¶:. Although since this is a pretty old challenge and all other answers use actual strings, I could understand if you kept it as string I/O.

Answer 17

JavaScript (ES6), 23 bytes

s=>" ".repeat(s.length)

Answer 18

Octave, 14 bytes

@(a)["" 0*a+32]

Answer 19

CJam, 4 bytes

q,S*

Try it online!

Explanation

q e# Read input
 , e# Length
 S* e# Repeat space that many times

Answer 20

V, 2 bytes

Ò 

Try it online!

Note the trailing space!

Answer 21

C, 45 bytes

Using main. Compile with gcc, ignore warnings.

main(c,v)char**v;{while(*(v[1]++))puts("");}

Usage:

$./a.out "Hello, World!"

• CalculatorFeline: Any reason why you can't put char**v in main(c,v)?
• sigvaldm: @CalculatorFeline At least GCC 6.3.1 compiling simply with gcc main.c doesn't seem to allow mixing ANSI function definition with K&R function definition, so main(c,char**v) won't compile. I either have to do main(int c,char**v) or main(c,v)char**v; of which the latter is 3 bytes shorter. You wouldn't by chance know any flag or something which allows mixing these styles?
• Cody Gray: No, you can't mix 'em. There's no flag that allows that. K&R style is long obsolete, used only for code golfing and obfuscation purposes.
• CalculatorFeline: And I'm guessing removing char**v entirely doesn't compile either.
• sigvaldm: @CalculatorFeline If you omit char** entirely the compiler will interpret it as int. If I'm not mistaken you get an error trying to dereference an int and even if you didn't the program wouldn't do what you expected it to do since an int consumes several chars and therefore you never get a NULL value.

Answer 22

Excel, 18 bytes

=REPT(" ",LEN(A1))

Pretty boring and one byte longer than the VBA answer.

Answer 23

><>, 7 bytes

i0(?;ao

The program is a loop

i //Push a character from the input onto the stack
 0 //Add a 0 to the stack
 ( //Pop the top two values of the stack, and push a 1 if the second is less than the first (In this case, the input has all been read), else push a 0
 ? //Pop the top of the stack. If the value is a 0, skip the next instruction
 ; // Terminate the program
 a // Add a newline to the stack
 o // Pop the top character of the stack and print it

Answer 24

Hexagony, (削除) 12 (削除ここまで) 11 bytes

-1 byte thanks to Martin Ender

,<.;.M@.>~8

Try it online!

Here is the expanded hex:

 , < . 
 ; . M @
. > ~ 8 .
 . . . .
 . . .

While there is input, this code runs:

, # Get input
 < # Turn right (SE) if we didn't get EOF
 M8 # Set the memory edge to 778 which is 10 (mod 256)
 ; # Print as a character (newline)
 > # Redirect East
 ~ # Multiply by -1. This makes the pointer go to the top when it runs off the edge
 8 # Effectively a no-op.

When EOF is reached:

, # Get input
 < # Turn left (NE)
 8 # Effectively a no-op
 @ # End program

• Martin Ender: You can print a linefeed in three bytes with M8; (which gives 778 = 10 (mod 256)). That should allow you to move the ~ where the ; is right now, saving a byte.

Answer 25

Pyth, 3 bytes

*dl

Try it!

Python equivalent: len(input())*" "

jmk

Try this!

Python equivalent: "\n".join(map("", input()))

smb

Try that!

Python equivalent: "".join(map("\n",input())

VQk

Try!

Python equivalent: For N in input():print("")

• KarlKastor: @StanStrum: It outputs newlines. Newlines are also allowed per the challenge description above. You can see this better from this example input

Answer 26

Perl 5, 7 bytes

-1 byte thanks to @Xcali

6 bytes of code + -p flag.

y// /c

Try it online!

Quite straight forward : replaces every character with a space.

• Xcali: y// /c is one byte shorter.

Answer 27

Python 2, 25 bytes

exec'print;'*len(input())

-2 bytes thanks to Loovjo
-2 bytes in the invalid code thanks to totallyhuman :p
-3 bytes

• xenia: You can remove the parens after exec since it's a keyword in Python 2
• hyperneutrino: @Loovjo Oh right, Python 2. Thanks!
• totallyhuman: I know this is old and stuff but exec'print;'*len(input()) works.
• totallyhuman: Oh, my bad, didn't think of that.
• hyperneutrino: @TheMatt it's probably not in the problem specs but it's one of the default acceptable input methods. Try looking on meta, I don't want to go looking for it right now

Answer 28

PHP, 36 bytes

<?=str_repeat('
',strlen($argv[1]));

Try it online!

Outputs newlines because spaces are too mainstream

• Jörg Hülsermann: $argn instead of $argv[1] save 4 bytes. Run with the -F option

Answer 29

Jelly, 2 bytes

6ṁ

Try it online!

Answer 30

C (tcc), 31 bytes

I opted to output newlines since it's shorter...

f(char*s){for(;*s++;puts(""));}

Try it online!