固定サイズのレコードを持つバイナリファイルのレコード数を決定する方法[閉じる]

この試み：

script.sh:

#!/bin/bash

# read filename to var f from $1
f="$1"
# quit if file not found.
[ -f "$f" ] || { >&2 echo "File not found." ; exit 2; }
# total length of file (-1 to not count newline at the end of a file)
n=$(($(wc -c < "$f")-1))
# loop potential length of records, end at half total length of file.
for i in $(seq 1 $(awk -v n=$n 'BEGIN{print int((n/2)+0.5)-1}')); do
    # now read all delimiters regarding record length from a file
    # (see https://unix.stackexchange.com/a/276957/236063)
    # and `sort -u` to get only individual ones
    c=$(for ((j=i;j<n;j=j+i+1)); do printf '%s\n' $(dd ibs=1 skip=$j count=1 < "$f" 2>/dev/null); done | sort -u)
    # if we have exactly one individual delimiter, we're done and can print length of record and the delimiter and exit.
    if [ $(echo "$c" | wc -l) -eq 1 ]; then
        printf 'l=%s d=%s\n' "$i" "$c"
        exit 0
    fi
done
>&2 echo "No delimiter found".
exit 1

コンテンツfile:

abc,def,ghi

出力：

$ ./script.sh file
l=3 d=,

終了コード：

 0 success, found delimiter and record length
 1 no delimiter found
10 file not found

注：特に改行に関連するいくつかの問題を解決するには、いくつかの作業が必要だと思います。