eb

#!/usr/bin/env bash

die() { printf %s "${@+$@$'\n'}" 1>&2 ; exit 1 ; }
see() ( { set -x; } 2>/dev/null ; "$@" )
dieifnotinpath() { command -v "$1" >/dev/null || die "$1 not in PATH"; }

dieifnotinpath 7z

optq=0

mkdir_or_die() {  # if already exists, do nothing (existing contents not disturbed)
   [[ -d "$1" ]] || mkdir "$1" || die "could not mkdir $1"
   [[ -d "$1" ]] ||               die "mkdir $1 failed"
   }
zfdin="./_consumed" ; mkdir_or_die "$zfdin"   # when making a bulk run, it's handy to be able to verify outcome by comparing the # of

excl_nonebook_files() (
   echo "eb before" ; find . -type f -print | sed 's/^/   /'
   xdnm="interesting_files"
   mkdir_or_die "$xdnm"
   rm -rf "$xdnm:?"/*   # in case dir already existed, nuke its content
   # rmv typical "advertising" files embedded in www downloads; extensions that never contain content we want to keep
   find . -type f \( \
    -iname "SHA1SUMS" -or \
    -iname "*.docx" -or \
    -iname "*.fb2" -or \
    -iname "*.html" -or \
    -iname "*.htmlz" -or \
    -iname "*.jpeg" -or \
    -iname "*.jpg"  -or \
    -iname "*.lit" -or \
    -iname "*.lrf" -or \
    -iname "*.opf" -or \
    -iname "*.pdb" -or \
    -iname "*.png" -or \
    -iname "*.rtf" -or \
    -iname "*.txt" -or \
    -iname "*.url" -or \
    -iname "*.DS_Store" \) -delete
   find . -type f ! \( \
    -iname "*.epub" -or \
    -iname "*.azw" -or \
    -iname "*.azw[34]" -or \
    -iname "*.mobi" -or \
    -iname "*.pdf" -or \
    -iname "*.zip" \) -exec mv {} "$xdnm/" \;  # *.zip: "accompanying code" files
   rmdir "$xdnm" || die "$xdnm is not empty!"  # if nothing left, rmdir will succeed
   )

do1file() (
   ((optq!=0)) || echo "arg=$1'"
   [[ -f "$1" ]] || { ((optq!=0)) || die "$1 is not a file" ; die ; }
   sanspath="${1##*/}"       # ; echo "sanspath=$sanspath'"    # https://stackoverflow.com/a/965069
   sansext="${sanspath%.*}"  # ; echo "sansext=$sansext'"
   [[ "${sanspath,,}" == *[_.]code.zip ]] && die "skipping code-zip file $sanspath"
   [[ -d "$sansext" ]] && die "dir $sansext already exists"

   7z e -o"$sansext" "$1" || die "7z extraction failed"
   (cd "$sansext" && find . -type d -empty -delete)   # `7z e ...` can create empty dirs in output dir: delete them
   # above is common to various scripts: unzips $1 into a dir of its own name (minus extension) ($sansext)

   # purpose: given a typical ebook archive file (zip, rar) extracted into $sansext, delete non-ebook files and rename ebook files $sansext.ext
   ( cd "$sansext" && excl_nonebook_files ) || die "excl_nonebook_files failed"
   ( cd "$sansext" && mv_calibre_junk ; rm -rf _calibre_junk )
   # some downloaded archives' files have timestamps newer than "now": touch them
   ( cd "$sansext" && find . -type f -newer "../$1" -printf 'from the future: %p\n' )
   ( cd "$sansext" && find . -type f -newer "../$1" -exec touch {} + )
   # move the files that're left back to cwd
   ( cd "$sansext" && find . -type f | while read -r file; do mv "$file" "../$sansext.${file##*.}" ; done )  # https://stackoverflow.com/a/8489394
   mv "$1" "$zfdin/" || die "mv $1 $zfdin/ FAILED" # move input file (which might be a zip file) into "done" dir
   rmdir "$sansext" || die "mv did not empty $sansext/"
   echo "------------------------------- eb successful -------------------------------"
   )

argerr() {  die "usage: $0 [-q] FILE..." ; }
while getopts "hq" opt; do
  case $opt in
    q) optq=1 ;;
    h) argerr ;;
   \?) argerr ;;
  esac
done
shift "$((OPTIND-1))"  # shift so that $@, $1, etc. refer to the non-option arguments

for fnm in "$@"; do
   do1file "$fnm"
done