Sample Awk pre-processor for string matching.

#! /bin/bash

HERE=~/"SandBox/Similar"
#.. Mount this first.
MUSIC="/media/paul/Windows/40_Music"

ALL="./Music.all"
TRACK="./Track.all"
RANK="./Track.rank"
UNIQ="./Track.uniq"

function Tracks {

AWK='
BEGIN {
    stderr = "cat 1>&2"; FS = "/";
    reBox = "[[][^]]*[]]";
    reSfx = "[.](WMA|wma|WAV|wav|MP3|mp3)$";
}
function Prune (tx) {
    #.. Early return for unhelpful filenames.
    #.. Like: AlbumArt_{6FE0559F-2C09-4AF7-8774-2912DC3CEDD7}_Large.jpg
    if (tx ~ /^AlbumArt.*Large.jpg$/) { ++CC["Delete Art-Large.jpg"]; return; }
    if (tx ~ /^AlbumArt.*Small.jpg$/) { ++CC["Delete Art-Small.jpg"]; return; }
    #.. Like: desktop.ini.
    if (tx !~ reSfx) { ++CC["Delete " tx]; return; }
    #.. Like: 08 Track 8.mp3
    if (tx ~ "^[0-9]*[ ]?Track [1-9][0-9]?") {
        ++CC["Delete unlabelled track"]; return;
    }
    #.. Remove generic texts before printing.
    #.. Like: 01 Love Minus Zero-No Limit.mp3
    if (sub (/^[0-9]+[ ]*/, "", tx)) ++CC["Omit track prefix"];
    #.. Like: Poetry in Motion>>.mp3
    if (sub (/[.][Mm][Pp]3$/, "", tx))    ++CC["Omit .mp3 suffix"];
    if (sub (/[.][Ww][Mm][Aa]$/, "", tx)) ++CC["Omit .wma suffix"];
    if (sub (/[.][Ww][Aa][Vv]$/, "", tx)) ++CC["Omit .wav suffix"];
    #.. Like: [Alternate Take]
    while (match (tx, reBox)) {
        ++CC["Omit " substr (tx, RSTART, RLENGTH)];
        tx = substr (tx, 1, RSTART-1) substr (tx, RSTART+RLENGTH);
    }
    print tx;
}
#.. Only consider filename -- last field.
{ Prune( $(NF)); }

function End (Local, j, K) {
    asorti (CC, K);
    for (j = 1; j in K; j++)
        printf ("%8d  %s\n", CC[K[j]], K[j]) | stderr;
}
END { End( ); }
'
    awk -f <( printf '%s\n' "${AWK}" )
}

#### Script Body Starts Here.

    #.. Compare the .mp3 tracks against each other.

    [ x ] && {
        echo 1>&2 ".... Extracting MP3_Tree"
        ( cd "${MUSIC}/MP3_Tree" && find * -type f ) > "${ALL}"
        Tracks < "${ALL}" > "${TRACK}"
        echo 1>&2 ".... One-file comparison"
        time ./Sim -f -b -t 0.75 -p -w < "${TRACK}" > "${RANK}"
        uniq -c < "${RANK}" > "${UNIQ}"
    }

    #.. Compare all the .wma tracks against the .mp3 ones.

    [ x ] && {
        echo 1>&2 ".... Extracting CD_RIPS"
        ( cd "${MUSIC}/CD_RIPS"  && find * -type f ) | Tracks > "./WMA.track"
        echo 1>&2 ".... Extracting MP3_Tree"
        ( cd "${MUSIC}/MP3_Tree" && find * -type f ) | Tracks > "./MP3.track"
        echo 1>&2 ".... Two-file comparison"
        time ./Sim -p -f -b -t 0.75 "./WMA.track" "./MP3.track" > "./Pair.rank"
        uniq -c < "./Pair.rank" > "./Pair.uniq"
    }