正则表达式忽略一些字母

Question

I'm trying to solve Chasing Subs problem . I'm trying to generate that regex according to the input data. The goal is go get all substrings (including overlapped ones) with all unique letters.

I'm trying to use regexp like this:

regexp = /(?=(?<gs>.)(?<gu>[^\k<gs>])(?<gb>[^\k<gs>\k<gu>])(?<gm>[^\k<gs>\k<gu>\k<gb>])(?<ga>[^\k<gs>\k<gu>\k<gb>\k<gm>])(?<gr>[^\k<gs>\k<gu>\k<gb>\k<gm>\k<ga>])(?<gi>[^\k<gs>\k<gu>\k<gb>\k<gm>\k<ga>\k<gr>])(?<gn>[^\k<gs>\k<gu>\k<gb>\k<gm>\k<ga>\k<gr>\k<gi>])(?<ge>[^\k<gs>\k<gu>\k<gb>\k<gm>\k<ga>\k<gr>\k<gi>\k<gn>]))/
"archipelago".scan(regexp) #=> []
"archipelbgo".scan(regexp) #=> []
"brchipelbgo".scan(regexp) #=> []
"zrchipelzgo".scan(regexp) #=> [["z", "r", "c", "h", "i", "p", "e", "l", "z"]] 

Why does it behave like this? Why can't it find anything with "b" and "a" ? And why does it return only one (incorrect) result with "z" ? What am I doing wrong?

Answer 1

I don't think a regular expression is the correct tool for this problem. We could do the following, however.

def substrings(str)
  arr = str.chars
  (1..str.size).each_with_object([]) { |n,a|
    arr.each_cons(n) { |b| a << b.join if b == b.uniq } }
end

substrings("archipelago")
  #=> ["a", "r", "c", "h", "i", "p", "e", "l", "a", "g", "o", "ar", "rc", "ch", "hi",
  #    "ip", "pe", "el", "la", "ag", "go", "arc", "rch", "chi", "hip", "ipe", "pel",
  #    "ela", "lag", "ago", "arch", "rchi", "chip", "hipe", "ipel", "pela", "elag",
  #    "lago", "archi", "rchip", "chipe", "hipel", "ipela", "pelag", "elago", "archip",
  #    "rchipe", "chipel", "hipela", "ipelag", "pelago", "archipe", "rchipel", "chipela",
  #    "hipelag", "ipelago", "archipel", "rchipela", "chipelag", "hipelago", "rchipelag",
  #    "chipelago", "rchipelago"]