Is it Covfefey?

Question

Based on this challenge, you must determine if a string is covfefey, that is, could it have been produced as output from a covfefifier?

The string will be composed of only alphabet chars (^[a-z]\*$, ^[A-Z]\*$ are possible schemes for input. change the regexs appropriately if using caps)

To do this, there are a few checks you must do:

Do the last four characters follow this scheme:

([^aeiouy][aeiouy])\1

example:

fefe

follows this scheme.

is this consonant at the start of the ending the voiced or voiceless version of a consonant immediately before it?

consonant pairings follow this key. find the consonant occuring before the end part consonant we had, and check whether the end part consonant is there

b: p
c: g
d: t
f: v
g: k
h: h
j: j
k: g
l: l
m: m
n: n
p: b
q: q
r: r
s: z
t: d
v: f
w: w
x: x
z: s

does the first part of the word, the part that does not include the ([^aeiouy][aeiouy])\1, follow this regex:

[^aeiouy]*[aeiouy]+[^aeiouy]

That is, does it contain one group of vowels, and exactly one consonant after that vowel?

If it fulfills all these criteria, it is covfefey.

Output a truthy value if covfefey, otherwise falsy

Test cases

covfefe -> truthy
barber  -> falsy
covefefe -> falsy
pressisi -> falsy
preszizi -> truthy
prezsisi -> truthy
president covfefe -> You don't need to handle this input as it contains a space.
cofvyvy -> truthy
businnene -> falsy (the first part does not follow the regex)

Answer 1

JavaScript (ES6), 109 108 bytes

Returns true for covfefey / null or false otherwise.

s=>(m=s.match`^[^aeiouy]*[aeiouy]+(.)((.)[aeiouy])\\2$`)&&((a="bcdfgszkvtgp")[11-a.search(b=m[1])]||b)==m[3]

Test cases

let f =

s=>(m=s.match`^[^aeiouy]*[aeiouy]+(.)((.)[aeiouy])\\2$`)&&((a="bcdfgszkvtgp")[11-a.search(b=m[1])]||b)==m[3]

;[
  "covfefe",
  "barber",
  "covefefe",
  "pressisi",
  "preszizi",
  "prezsisi",
  "cofvyvy",
  "exxaxa"
]
.map(s => console.log(s, '-->', f(s)))

Answer 2

Jelly, 56 bytes

5ịØYiị“ßȷ%Hẹrȧq’œ?ØY¤⁾cgy
Ṛµ4,2ị=ÇȦ
Çȧe€ØyµḄ%⁴⁼5ȧŒgLe5,6

A monadic link taking a list of lowercase characters and returning 1 for truthy and 0 for falsey.

Try it online! or see the test suite.

How?

5ịØYiị“ßȷ%Hẹrȧq’œ?ØY¤⁾cgy - Link 1, convert 5th character: list of characters, r
5ị                        - index into r at 5
  ØYi                     - first index of that in "BCDF...XZbcdfghjklmnpqrstvwxz"
                    ¤     - nilad followed by link(s) as a nilad:
      “ßȷ%Hẹrȧq’          -   base 250 number = 1349402632272870364
                  ØY      -   yield consonants-y = "BCDF...XZbcdfghjklmnpqrstvwxz"
                œ?        -   nth permutation    = "BCDF...XZpctvkhjglmnbqrzdfwxs"
     ị                    - index into that (converts a 'b' to a 'p' etc..)
                     ⁾cg  - literal ['c','g']
                        y - translate (convert a 'c' to a 'g' - a special case)

Ṛµ4,2ị=ÇȦ - Link 2, check the 3 consonants at the end: list of characters, w
Ṛ         - reverse w
 µ        - monadic chain separation, call that r
  4,2     - 4 paired with 2 = [4,2]
     ị    - index into r
       Ç  - call the last link (1) as a monad (get 5th char converted)
      =   - equal? (vectorises)
        Ȧ - any and all (0 if empty or contains a falsey value, else 1)

Çȧe€ØyµḄ%⁴⁼5ȧŒgLe5,6 - Main link: list of characters, w
Ç                    - call last link as a monad (1 if the 3 consonants are good; else 0)
    Øy               - vowel+y yield = "AEIOUYaeiouy"
  e€                 - exists in? for €ach c in r
 ȧ                   - logical and (list of isVowels; or 0 if link 2 returned 0)
      µ              - monadic chain separation, call that v
       Ḅ             - convert from binary (0 yields 0)
         ⁴           - literal 16
        %            - modulo (0 yields 0)
          ⁼5         - equals 5? (1 if ending is consonant,vowel,consonant,vowel; else 0)
             Œg      - group runs of v (e.g. [0,0,0,1,1,0,1,0,1]->[[0,0,0],[1,1],[0],[1],[0],[1]])
            ȧ        - logical and (the grouped runs if AOK so far, else 0)
               L     - length (number of runs)
                 5,6 - 5 paired with 6 = [5,6]
                e    - exists in (the number of runs must be 5 or 6 to comply)

Answer 3

Java 8, 211 210 166 163 161 156 bytes

s->{int l=s.length()-5;return"bpbdtdfvfgkgszshhjjllmmnnqqrrwwxxcg".contains(s.substring(l,l+2))&s.matches("[^x]*[x]+(.)((.)[x])\\2".replace("x","aeiouy"));}

-2 bytes by replacing the regex [^x]*[x]+[^x]([^x][x])\\1 with [^x]*[x]+(.)((.)[x])\\2 thanks to @Arnauld's JavaScript answer.

Explanation:

Try it here.

s->{                   // Method with String parameter and boolean return-type
  int l=s.length()-5;  //  Temp integer value with the length of the input-String - 5
  return"bpbdtdfvfgkgszshhjjllmmnnqqrrwwxxcg".contains(s.substring(l,l+2))
                       //   Return if the correct pair is present at the correct location
                       //    (i.e. `vf` at index l-5 + l-6 in the String `convfefe`
   &s.matches("[^x]*[x]+(.)((.)[x])\\2".replace("x","aeiouy"));
                       //    and if the String matches the regex-pattern
}                      // End of method

Answer 4

Perl, 80 bytes

$_=/[aeiouy]+(*COMMIT)(.)((.)[aeiouy])\2$/&&$3eq$1=~y/bcdfgkpstvz/pgtvkgbzdfs/r

Run with perl -pe.

Due to their length, it’s not often that I get to use the exotic regex verbs in golf, but this challenge is a perfect fit!

Whenever the engine backtracks to (*COMMIT), the entire match fails. This guarantees that the part that comes before it, [aeiouy]+, can only ever match the first thing it tries to match, namely the first continuous stretch of vowels. Thus, [aeiouy]+(*COMMIT)(.) is equivalent to ^[^aeiouy]*[aeiouy]+([^aeiouy]) (equivalent in terms of which strings match, but matches are on different positions).

Answer 5

PHP , 121 Bytes

$v=aeiouy;echo preg_match("#^[^$v]*[$v]+([^$v])(([^$v])[$v])\\2$#",$argn,$t)&$t[3]==strtr($t[1],bcdfgkpstvz,pgtvkgbzdfs);

Try it online!