Re: (Long post) Metaphone Algorithm In AWK

Final iteration (for me). Please post any improvements in this group.

# Metaphone Algorithm in AWK v6: Michael Sanders - 2024
# usage example: echo poluphloisboiotatotic | awk -f metaphone.awk
# valid output should be: POLUPHLOISBOIOTATOTIC : PLFLSBTTTK
# see also: https://en.wikipedia.org/wiki/Metaphone

{ print $0 " : " metaphone($0) }

function isvowel(char) { return char ~ /[AEIOU]/ }

function metaphone(w, m, c, n, z, i) {

  w = toupper(w)
  # strip non-alphabetic characters
  gsub(/[^A-Z]/, "", w)
  z = length(w)

  # handle initial letters
  if (substr(w, 1, 2) ~ /^(KN|GN|PN|WR|PS)/) {
    w = substr(w, 2)
    z--
  }

  for (i = 1; i <= z; i++) {
    c = substr(w, i, 1)
    n = (i < z) ? substr(w, i + 1, 1) : ""

    # skip duplicate letters except for 'C'
    if (i > 1 && c == substr(w, i - 1, 1) && c != "C") continue

    # handle vowels: retain only if it's the 1st letter
    if (isvowel(c)) {
      if (i == 1) m = m c
    }
    # consonants...
    else if (c == "B") {
      if (!(i == z && substr(w, i - 1, 1) == "M")) m = m "B"
    }
    else if (c == "C") {
      if (substr(w, i, 2) == "CH") {
        m = m "X"
        i++
      } else if (substr(w, i, 2) ~ /^(CI|CE|CY)/) {
        m = m "S"
      } else {
        m = m "K"
      }
    }
    else if (c == "D") {
      if (substr(w, i, 2) == "DG" && isvowel(substr(w, i + 2, 1))) {
        m = m "J"
        i += 2
      } else {
        m = m "T"
      }
    }
    else if (c == "F") {  # Handling for 'F'
      m = m "F"
    }
    else if (c == "G") {
      if (substr(w, i, 2) == "GH" && (i == 1 || !isvowel(substr(w, i - 1, 1)))) {
        i++
      } else if (substr(w, i, 2) == "GN" || (i == z && c == "G")) {
        continue
      } else if (substr(w, i, 3) ~ /^(GIA|GIE|GEY)/) {
        m = m "J"
      } else {
        m = m "K"
      }
    }
    else if (c == "H") {
      if (i == 1 || substr(w, i - 1, 1) !~ /[CSPTG]/) {
        if (i < z && !isvowel(n)) {
          m = m "H"
        }
      }
    }
    else if (c == "K") {
      if (i == 1 || substr(w, i - 1, 1) != "C") m = m "K"
    }
    else if (c == "P") {
      if (substr(w, i, 2) == "PH") {
        m = m "F"
        i++
      } else {
        m = m "P"
      }
    }
    else if (c == "Q") {
      m = m "K"
    }
    else if (c == "R") {
      if (i == 1 || isvowel(substr(w, i - 1, 1))) {
        m = m "R"
      }
    }
    else if (c == "S") {
      if (substr(w, i, 2) == "SH") {
        m = m "X"
        i++
      } else if (substr(w, i, 3) == "TIA" || substr(w, i, 3) == "TIO") {
        m = m "X"
        i += 2
      } else {
        m = m "S"
      }
    }
    else if (c == "T") {
      if (substr(w, i, 2) == "TH") {
        m = m "0" # add '0' for 'TH' digraph to distinguish from regular 'T'
        i++
      } else if (substr(w, i, 3) == "TIA" || substr(w, i, 3) == "TIO") {
        m = m "X"
        i += 2
      } else {
        m = m "T"
      }
    }
    else if (c == "V") {
      m = m "F"
    }
    else if (c == "W" || c == "Y") {
      if (i < z && isvowel(n)) m = m c
    }
    else if (c == "X") {
      m = m "KS"
    }
    else if (c == "Z") {
      m = m "S"
    }
    # handle M, N, L, J, G more generally
    else if (c == "M" || c == "N" || c == "L" || c == "J" || c == "G") {
      m = m c
    }
  }

  return m
}

# eof

--
:wq
Mike Sanders