Introduction

My car speedometer was hacked! Instead of showing me how fast i'm driving, it just shows: "Vroooom!" Please help me know how fast i'm going.

Challenge

Take a string as input, and check if it matches the regex /^[Vv]ro*m!$/m. In English that means any line of the string must start with a capital or lowercase v, then a lowercase r, then any amount (including zero) of the lowercase letter o, then the exact string m!. There may be other lines, but the Vroom string must be on it's own line. If you find a match, then you must count the amount of o's in the Vroom string and output it. If you don't find a match however, you should output any default value that can't be outputted otherwise (like -1 or an empty string) Reminders Scoring This is , so the shortest code in bytes wins. However, I will not mark any answer as accepted. Test cases Input Vrom!  Output 1 Input vrooooooom!  Output 7 Input Hello, Vroom!  Output (none) Input Foo bar boo baz Vrooom! hi  Output 3 Input Vrm!ooo  Output (none) Input PPCG puzzlers pie  Output (none) Input hallo vROOOm!  Output (none) 32 Answers sed 4.2.2, 20 bytes -nr options required at the command-line. s/^[Vv]r(o*)m!$/\1/p


This outputs the speed in unary as the number of os.

Try it online!

• Can't tell Vrm! from Vram! – l4m2 Nov 3 '18 at 8:56

lambda x:len(re.search('^[Vv]r(o*)m!$',x,8).group(1))  Try it online! Basic regex and grouping, uses re.MULTILINE flag (which has a value of 8) and re.search to ensure it works for multiline inputs. Raises an exception when no match is found. Thanks to @ovs for the -3 bytes from (re.M == 8) tip. • Welcome to PPCG! I reformatted your answer to make it appear a bit nicer, if you're unhappy with my edit you can always roll back. Btw. I suggest to link to something like tio.run such that people can easily test your answer. – BMO Nov 2 '18 at 20:30 • re.M has the value 8, so can just use re.search(regex,x,8) – ovs Nov 3 '18 at 16:24 R, 62 60 58 44 bytes nchar(grep("^[Vv]ro*m!$",readLines(),v=T))-4


Try it online!

@Giuseppe with 14 bytes golfed.

Original approach with explanation:

function(x)attr(el(regexec("(?m)[Vv]r(o*)m!$",x,,T)),"m")[2]  Try it online! R has seven pattern matching functions. The more commonly used ones are grep, grepl, and sub, but here's a nice use for regexec. regexec gives you a bunch of things, one of which is the length of any captured substring, in this case the (o*) part of the multiline regex. The attr(el .... "m")[2] stuff is a golfy way to get the desired number. Returns NA if there is no match. • 58 bytes with a different approach – digEmAll Nov 3 '18 at 15:33 • I have a 44 byte approach... Not gonna post it unless you want me to. – Giuseppe Nov 4 '18 at 1:53 • @Giuseppe not sure why not? Especially if it's fundamentally different. – ngm Nov 4 '18 at 2:01 JavaScript (Node.js), 41 bytes a=>(l=/[Vv]r(o*)m!/.exec(a))&&l[1].length  Try it online! • This fails for vroooooooooooom!x\nvrom! – BMO Nov 2 '18 at 20:27 • Seeing as we're allowed exit with an error if no match is found, you could do this for -3 bytes, fixing the problem @BMO mentioned above in the process. – Shaggy Nov 2 '18 at 22:32 • Uncrossed-out 41 is definitely still 41 – Redwolf Programs Nov 3 '18 at 2:13 • @Shaggy What's the space in [1]. length for? – l4m2 Nov 3 '18 at 8:58 • @l4m2, a typo! I didn't spot it on my phone 'cause length was breaking to a newline anyway. – Shaggy Nov 3 '18 at 10:53 Powershell, 625853 48 bytes bytes "$($args|sls '(?m-i)^[Vv]ro*m!$'|% M*)".Length-4


returns numbers of o in a first Vroom!, or -4 if Vroom! not found.

Notes:

• sls is alias for Select-String;
• (?m-i) inside regexp means:
• Use multiline mode. ^ and $ match the beginning and end of a line, instead of the beginning and end of a string. • Use case-sensitive matching • |% M* is shortcut for the property Matches, which gives a first match because we don't use -AllMatches parameter. Test script: $f = {

"$($args|sls '(?m-i)^[Vv]ro*m!$'|% M*)".Length-4 } @( ,('Vrom!',1) ,('vrooooooom!',7) ,('Hello, Vroom!',-4) ,('Foo bar boo baz Vrooom! hi',3) ,('Vrm!ooo',-4) ,('PPCG puzzlers pie',-4) ,('hallo vROOOm!',-4) ,(' Vrooom! Vrooooom! ',3) # undefined behavior. ,('vrm!',0) # :) ) | % {$n,$expected =$_
$result = &$f $n "$($result-eq$expected): $result" }  Output: True: 1 True: 7 True: -4 True: 3 True: -4 True: -4 True: -4 True: 3 True: 0  Retina, 21 bytes L$m^[Vv]r(o*)m!.1


Try it online! Explanation: L lists matches, so if the regex fails to match then output is empty. $ causes the result to be the substitution rather than the match. m makes it a multiline match (the equivalent to the trailing m in the question). The . in the substitution makes it output the length of the capture in decimal. SNOBOL4 (CSNOBOL4), 99 82 bytes I INPUT POS(0) ('V' | 'v') 'r' ARBNO('o') @X 'm!' RPOS(0) :F(I) OUTPUT =X - 2 END  Try it online! Pretty direct SNOBOL translation of the spec, reads each line until it finds one that matches ^[Vv]ro*m!$, then outputs the length of the o* bit.

Enters an infinite loop if no Vroom! can be found.

• Is all that whitespace necessary? Wow. – FireCubez Nov 2 '18 at 18:51
• @FireCubez yep, that's what you get with a 50+ year-old language: weird whitespace requirements. It uses space/tab as concatenation and you must surround operators with whitespace as well. – Giuseppe Nov 2 '18 at 18:53

{-!/^^[V|v]r(o)*m\!/+$0}  Try it online! C (gcc), 188 183 bytes Why use regexes when you can use a state machine instead? :-) a,b;f(char*s){for(a=b=0;a<5;s++){!a&*s==86|*s=='v'?a++:a==1&*s=='r'?a++:a==2?*s-'o'?*s-'m'?0:a++:b++:a==3&*s==33?a++:!*s&a==4?a++:*s-10?(a=-1):a-4?(a=0):a++;if(!*s)break;}s=a<5?-1:b;}  Try it online! PowerShell, 83 bytes ($args-split"n"|%{if(($x=[regex]::Match($_,"^[Vv]ro*m!$")).success){$x}}).length-4


Try it online!

-splits the input $args on newlines, pipes those into a for loop. Each iteration, we check whether our [regex]::Match is a .success or not. If so, we leave $x (the regex results object) on the pipeline. Outside the loop, we take the .length property -- if it's the regex results object, this is the length of the match (e.g., "Vroom!" would be 6); if it's not a regex results object, the length is zero. We then subtract 4 to remove the counts for the Vrm! and leave that on the pipeline. Output is implicit. Outputs a -4 if no match is found.

• sls "^[Vv]ro*m!$"? – mazzy Nov 2 '18 at 18:58 • @mazzy How would that work for multiline input? Your only input is one string, and so sls will give back ('','Vroom!','') for example. – AdmBorkBork Nov 2 '18 at 19:06 • it's not completed solution. I mean, you can try sls instead [regex]::Match – mazzy Nov 2 '18 at 19:08 • @mazzy Maybe you should post it as a separate solution. – AdmBorkBork Nov 2 '18 at 19:13 Stax, 16 bytes ∞╠mQ╛3mQ->n▀÷↕┐ò  Run and debug it Haskell, 7571 69 bytes f s=[length n-2|r<-lines s,n<-scanr(:)"m!"$'o'<$r,v<-"Vv",r==v:'r':n]  Try it online! No regex. Instead builds all valid Vrooom!-strings up to a sufficient length and compares the lines of the input against them, collecting the number of os in a list. Thus for invalid inputs an empty list is returned. C (gcc), 104 100 bytes s;main(c,n){for(;gets(&s);sscanf(&s,"v%*[o]%nm%c%c",&n,&c,&c)-1||c-33?:printf("%d",n-2))s=s-768|32;}  Try it online! Output the n for each valid line, exactly in the requirement(nothing if no valid line, the n if exactly one) int s; // Use as a char[] main(c){ while(gets(&s)) { s=s-768|32; // byte 0: 'V'=>'v'; byte 1: 'r'=>'o', 'o'=>'l' if (sscanf(&s,"v%[o]m%c%c",&s,&c,&c)==2 && c=='!') { // The last '%c' get nothing if it's EndOfLine printf("%d",strlen(&s)-1)) } } }  • It's so funny that the regex answer is longer than this – Windmill Cookies Nov 3 '18 at 11:12 • @WindmillCookies GCC need extra code to support regex – l4m2 Nov 3 '18 at 11:29 • hmm. seems like names related to regexes are extremely long – Windmill Cookies Nov 3 '18 at 11:32 Japt, 18 bytes fè^[Vv]*m!$ ®èo


Try it online!

Saved a byte by taking input as an array of lines.

Includes an unprintable character between ] and *.

Explanation:

fè                   Get the line(s) that match
^[Vv]*m!$ The provided RegEx with a little compression ®èo Count the number of "o" in that line if it exists  • 19 bytes – Shaggy Nov 2 '18 at 22:37 • Also 19 bytes – Shaggy Nov 2 '18 at 23:13 • Actually, as input can be an array of lines, you can knock the first byte off my first comment above. – Shaggy Nov 3 '18 at 12:07 • @Shaggy I can't find anywhere in the question that specifies that input can be an array of lines, and it doesn't seem to be listed in the default I/O methods that a multiline string can be taken as an array of lines. It seems likely to be reasonable, but I'll wait for confirmation first. – Kamil Drakari Nov 3 '18 at 17:34 C (gcc), 138 124 bytes Here's the boring regex way. #include<regex.h> f(char*s){regmatch_t m[9];regcomp(m+2,"^[Vv]r(o*)m!$",5);s=regexec(m+2,s,2,m,0)?-1:m[1].rm_eo-m[1].rm_so;}


Try it online!

IJjb.nm:rd0"^vro*m!$"1.z/J\o  Try it online! Pyth, 20 bytes /R\o:#"^Vro*m!$"1cQb


Outputs as a list containing only the number of 'o's, or an empty list if there's no Vroom.
Try it here

Explanation

/R\o:#"^Vro*m!$"1cQb cQb Split on newlines. :#"^Vro*m!$"1     Filter the ones that match the regex.
/R\o                  Count the os in each remaining element.


Pip, 21 bytes

a~,^[Vv]r(o*)m!$#$1


Try it online!

Match regex ^[Vv]r(o*)m!$ in multiline mode; output length of capture group. sfk, 94 bytes xex -i -case "_[lstart][char of Vv]r[chars of o]m![lend]_[part 4]o_" +linelen +calc "#text-1"  Try it online! Gives -1 when you're not vrooming. Red, 104 bytes func[s][n:""if parse/case s[opt[thru"^/"]["V"|"v"]"r"copy n any"o""m!"opt["^/"to end]][print length? n]]  Try it online! A straightforward solution. Red's parse is cool and readable, but too long compared to regex Red [] f: func [ s ] [ n: "" if parse/case s [ opt [ thru newline ] [ "V" | "v" ] "r" copy n any "o" "m!" opt [ newline to end ] ] [ print length? n ] ]  J, 35 bytes (]{~0<{.)(1{'^[Vv]r(o*)m!'rxmatch])  Returns negative 1 if the pattern doesn't match. JavaScript, 9073 61 bytes _=>_.replace(/^[Vv]r(o*)m!$|[^\1]/mg,(m,a)=>a||'').length||-1


Try it online!

Replace characters that are not captured at (o*) with empty string, return length of string containing only "o" or -1 if resulting string is empty.

Ruby, 32 bytes

->n{n=~/^[Vv]r(o*)m!$/m;$1.size}


Matches the string to the regex, then uses Ruby's magic regex group variables to get the first group's size.

Call it like so:

x=->n{n=~/^[Vv]r(o*)m!$/m;$1.size}
x["Vrooooooooooooooooooooom!"] # returns 21


Ruby, 28 29 bytes

p$_[/^[vV]r(o*)m!$/].count ?o


Try it online!

Multi-line strings require three more bytes. I'm not sure if that's a hard requirement. If so, I'll update this.

->l{l[/^[Vv]r(o*)m!$/].count ?o}  • How can you test multi-line strings? – Laikoni Nov 3 '18 at 9:17 • Fail on VROM! – l4m2 Nov 3 '18 at 10:59 Clojure, 90 bytes #(do(def a(clojure.string/replace % #"(?ms).*^[Vv]r(o*)m!$.*""$1"))(if(= a %)-1(count a)))  Try it online! This anonymous function returns the number of "o"s in the vroom string, or -1 if there is no valid vroom string. Readable version (fn [s] (def a (clojure.string/replace s #"(?ms).*^[Vv]r(o*)m!$.*" "$1")) (if (= a s) -1 (count a)))  Explanation #"(?ms).*^[Vv]r(o*)m!$.*" ; This regex matches any string that contains a valid vroom string. The first capturing group contains only the "o"s in the vroom string
(clojure.string/replace s #"(?ms).*^[Vv]r(o*)m!$.*" "$1") ; Replaces a match of the above regex with its first capturing group. The resulting string is stored in the variable a
(if (= a s) -1 (count a))) ; a equals s if and only if there is no valid vroom string, so if a equal s we return -1. If there is a valid vroom string, a contains only the "o"s from the vroom string, so we return the length of a


perl -nE, 35 bytes

$s=length$1if/^[Vv]r(o*)m!$/}{say$s


This uses the Eskimo greeting (}{) which abuses the a quick on how the -n option is dealt with by perl.

Java 8, 109 bytes

s->{int r=-1;for(var l:s.split("\n"))r=l.matches("[Vv]ro*m\\!")?l.replaceAll("[^o]","").length():r;return r;}


Try it online.

Explanation:

s->{                             // Method with String parameter and integer return-type
int r=-1;                      //  Result-integer, starting at -1
for(var l:s.split("\n"))       //  Loop over the lines:
r=l.matches("[Vv]ro*m\\!")?  //   If the current line matches the regex:
l.replaceAll("[^o]","").length()
//    Change r to the amount of "o"'s in it
:                          //   Else:
r;                        //    Leave the result r unchanged
return r;}                     //  Return the result


C# (.NET Core), 134 122 bytes

for(var a="";a!=null;a=Console.ReadLine())if(new Regex(@"^[Vv]ro*m!$").Match(a).Success)Console.Write(a.Count(x=>x=='o'));  Try it online! -12 bytes: moved null check into for loop and removed brackets Ungolfed: for(var a = ""; a != null; a = Console.ReadLine()) // initialize a, and while a isn't null, set to new line from console if(new Regex(@"^[Vv]ro*m!$")                        // set regex
.Match(a).Success)              // check if the line from the console matches
Console.Write(a.Count(x => x == 'o'));              // write the number of 'o's to the console

• -10 bytes with C# 6's null-coalescing and conditional operators, also unnecessary {} when using just one statement in the for loop: for(var a="";;a=Console.ReadLine())Console.WriteLine(new Regex(@"^[Vv]ro*m!$").Match(a??"").Success?a.Count(x =>x=='o'):-1); – Ivan García Topete Nov 2 '18 at 21:37 • Also, this needs using System.Linq; using System.Text.RegularExpressions;, not sure if this is important lol – Ivan García Topete Nov 2 '18 at 21:46 • The code you provided doesn't actually work, as it will not only output a -1 for every line that it doesn't work on, but it will output -1s forever as there is no check for null. – Meerkat Nov 5 '18 at 14:24 • No, it won't. a = Console.ReadLine() makes the loop, so each time you're requesting input for the loop to go on, if there is no input, the loop is just waiting, not printing -1 forever – Ivan García Topete Nov 5 '18 at 15:28 • Proof of concept. Even if it did work as you said, a never-ending loop is not ideal behavior. Regardless, I've moved the null check into the for loop, which removes the brackets (and makes the code shorter than your suggestion). – Meerkat Nov 5 '18 at 16:01 05AB1E, 39 37 bytes |ʒć„VvsåsÁÁD…m!rÅ?s¦¦¦Ù'oså)P}Dgi'o¢  Although 05AB1E is a golfing languages, regex-based challenges are definitely not its strong suite, since it has no regex-builtins. Outputs [] if no match was found. Explanation: | # Get the input split by newlines ʒ # Filter it by: ć # Head extracted: Pop and push the remainder and head-character # i.e. "vrm!" → "rm!" and "v" # i.e. "Vroaom!" → "roaom!" and "V" „Vvså # Is this head character in the string "Vv"? # i.e. "v" → 1 (truthy) # i.e. "V" → 1 (truthy) s # Swap so the remainder is at the top of the stack again ÁÁ # Rotate it twice to the right # i.e. "rm!" → "m!r" # i.e. "roaom!" → "m!roao" D # Duplicate it …m!rÅ? # Does the rotated remainder start with "m!r"? # i.e. "m!r" → 1 (truthy) # i.e. "m!roao" → 1 (truthy) s¦¦¦ # Remove the first three characters from the duplicated rotated remainder # i.e. "m!r" → "" # i.e. "m!roao" → "oao" Ù # Uniquify, leaving only distinct characters # i.e. "" → "" # i.e. "oao" → "oa" 'oså '# Is this uniquified string in the string "o"? # i.e. "" → 1 (truthy) # i.e. "oa" → 0 (falsey) )P # Check if all three checks above are truthy # i.e. [1,1,1] → 1 (truthy) # i.e. [1,1,0] → 0 (falsey) } # Close the filter D # After the filter, duplicate the list gi # If its length is 1: # i.e. ["vrm!"] → 1 (truthy) # i.e. [] → 0 (falsey)  # Push the value in this list to the stack # i.e. ["vrm!"] → "vrm!" 'o¢ '# And count the amount of "o" in it (and output implicitly) # i.e. "vrm!" → 0 # (Implicit else:) # (Implicitly output the duplicated empty list) # i.e. []  C++, MSVC, 164 159 bytes -5 bytes thanks to Zacharý It compiles even with the regex header only #include<regex> using namespace std;int f(vector<string>i){smatch m;for(auto&e:i)if(regex_match(e,m,regex("^[Vv]ro*m!$")))return m[0].str().size()-4;return-1;}


Tests :

std::cout << "Vrom!" << " -> " << f({ "Vrom!" }) << '\n';
std::cout << "vrooooooom!" << " -> " << f({ "vrooooooom!" }) << '\n';
std::cout << "Hello, Vroom!" << " -> " << f({ "Hello, Vroom!" }) << '\n';
std::cout << "Foo bar boo baz \\n Vrooom! \\n hi" << " -> " << f({ "Foo bar boo baz", "Vrooom!", "hi" }) << '\n';
std::cout << "Vrm!ooo" << " -> " << f({ "Vrm!ooo" }) << '\n';
std::cout << "PPCG puzzlers pie" << " -> " << f({ "PPCG puzzlers pie" }) << '\n';
std::cout << "hallo \\n vROOOm!" << " -> " << f({ "hallo", "vROOOm!" }) << '\n';

• I think using namespace std;` would save a few bytes – Zacharý Nov 5 '18 at 14:30