reflacker-threaded.sh v01

#!/bin/bash

#######
#
# NAME:
#      reflacker-threaded.sh - A Bash script for re-encoding FLAC files with ReplayGain
#
# SYNOPSIS:
#      reflacker-threaded.sh ROOT_OF_SOURCE_TREE
#
# DESCRIPTION:
#      Re-encode a single folder or an entire folder tree of FLAC files, add ReplayGain,
#      detect and remove ID3 tags while preserving any existing VORBIS comment tags, export
#      and purge embedded album art, with per-folder multithreading.
#
# DEPENDS:
#      file, flac, (GNU) parallel
#
# RECOMMENDS:
#      metaflac, id3v2, date
#
# VERSION: 01
#
#######

flac_options="-V8 --force --totally-silent --padding=4096" #

parallel_jobs="0" #  "0" - Run as many jobs as possible (or 1 job per CPU core anyway)
                  #  "N" - Run up to N jobs in parallel
#  'parallel -j'  # "-N" - Subtract N from number of CPU cores, run this many jobs in parallel
                  # "+N" - Add N to number of CPU cores, run this many jobs in parallel

# enable the following options with "1", any other value disables them
   unembed_artwork="1" # NB: only preserves art embedded in the _first_ flac in a given directory. see lines 93, 95, and 97 for how PADDING blocks are handled, or could be
check_for_id3_tags="1" # NB: currently, after removing id3 tags there's no padding in output, using --padding in $flac_options is advised
   add_replay_gain="1" # attempts to only add replaygain tags to files that don't already have them

command -v file >/dev/null 2>&1                             || { printf '\nfile (the program) not found, aborting.\n\n' ;exit 1 ; }
[[ "$( file -b "$( command -v parallel )" )" == "Perl"* ]]  || { printf '\nGNU Parallel not found, aborting.\n\n' ;exit 1 ; }
command -v flac >/dev/null 2>&1                             || { printf '\nflac not found, aborting.\n\n' ;exit 1 ; }
if [[ "$unembed_artwork" == "1" || "$check_for_id3_tags" == "1" || "$add_replay_gain" == "1" ]] ;then
    if ! command -v metaflac >/dev/null 2>&1 ;then
        printf '\nmetaflac not found, disabling _EVERY_FEATURE_ except threaded flac tree re-encoding.\n\n'
        unembed_artwork="0" ;check_for_id3_tags="0" ;add_replay_gain="0"
    fi
fi
if [[ "$check_for_id3_tags" == "1" ]] ;then
    if ! command -v id3v2 >/dev/null 2>&1 ;then
        printf '\nid3v2 (the program) not found. Checking for and removing ID3 tags is disabled.\n\n'
        check_for_id3_tags="0"
    fi
fi

[[ "$check_for_id3_tags" == "1" ]] || printf '\nDetection of ID3 tags currently disabled.\nflac will refuse to operate on any FLAC file with an ID3 tag.\n\n'

declare -x flac_options parallel_jobs unembed_artwork check_for_id3_tags add_replay_gain # how we get our opaque-as-fuck function to see these variables/values

readarray -t -d '' dirs_in_tree < <( find "$1" -type d -print0 ) # https://unix.stackexchange.com/questions/263883/bash-pipe-find-output-into-readarray/263885#263885

fail_message() { printf '  ______      _____ _      ______ _____  _
 |  ____/\   |_   _| |    |  ____|  __ \| |
 | |__ /  \    | | | |    | |__  | |  | | |
 |  __/ /\ \   | | | |    |  __| | |  | | |
 | | / ____ \ _| |_| |____| |____| |__| |_|
 |_|/_/    \_\_____|______|______|_____/(_)

'
}

actions() {

    [[ ! -d "$1" ]] && exit

    if [[ "$( ls "$1" | grep -ic \\.flac )" -lt "1" ]] ;then # SC2010: Don't use ls | grep. Use a glob or a for loop with a condition to allow non-alphanumeric filenames.
        exit
    else
        shopt -s nocaseglob
        flacfiles+=( "$1"/*.flac )
        shopt -u nocaseglob
        printf '\n'
    fi

    # embedded artwork? make a copy before we CURSE IT! TAKE BACK THOSE 2 MEGABYTES!
    [[ "$unembed_artwork" == "1" ]] && {
        if metaflac --list "${flacfiles[0]}" |grep -A 1 "METADATA block" |grep -q PICTURE ;then
            printf 'Exporting embedded artwork... '
            if metaflac --export-picture-to="$1"/embed_art.unk "${flacfiles[0]}" ;then
                if grep -q JPEG <<< "$( file -b "$1"/embed_art.unk )" ;then
                    printf 'JPEG format detected.\n'
                    mv "$1"/embed_art.unk "$1"/embed_art.jpg
                elif grep -q PNG <<< "$( file -b "$1"/embed_art.unk )" ;then
                    printf 'PNG format detected.\n'
                    mv "$1"/embed_art.unk "$1"/embed_art.png
                elif grep -q GIF <<< "$( file -b "$1"/embed_art.unk )" ;then
                    printf 'GIF format detected.\n'
                    mv "$1"/embed_art.unk "$1"/embed_art.gif
                else
                    printf 'unable to determine filetype, saved as '\''embed_art.unk'\''.\n'
                fi
                metaflac --dont-use-padding --remove --block-type=PICTURE,PADDING "${flacfiles[@]}"
                # use this one instead I guess, if you aren't adding padding in $flac_options:
                #metaflac --dont-use-padding --remove --block-type=PICTURE "${flacfiles[@]}"
                # probably not needed at all, definitely not if all padding is removed in previous command:
                #metaflac --list --block-type=PADDING "${flacfiles[@]}" |grep -q "is last: false" && metaflac --sort-padding "${flacfiles[@]}"
            else
                printf 'FAILED!\n'
            fi
        fi
    }

    # ID3 tags? CURSE THEM!
    [[ "$check_for_id3_tags" == "1" ]] && {
        if grep -q ID3 <<< "$( file -b "${flacfiles[@]}" )" ;then
            for id3flac in "${flacfiles[@]}" ;do
                id3temp="$(basename "$id3flac")"
                if metaflac --export-tags-to="$1/$id3temp".metaflac "$id3flac" ;then
                    id3v2 -D "$id3flac"
                    metaflac --dont-use-padding --remove-all "$id3flac"
                    metaflac --import-tags-from="$1/$id3temp".metaflac "$id3flac" && rm "$1/$id3temp".metaflac
                else
                    printf 'Tag export failed, aborting ID3 tag removal for %s.\n' "$id3flac"
                fi
            done
        fi
    }

    if [[ "$add_replay_gain" == "1" && "$( metaflac --list --block-type=VORBIS_COMMENT "${flacfiles[0]}" |grep -c REPLAYGAIN )" -ne "5" ]] ;then
        printf 'Re-encoding FLAC file(s) with Replay Gain.\n'
        if flac $flac_options --replay-gain "${flacfiles[@]}" ;then
            printf 'Success! @ %s.\n\n' "$( date +%_I:%M%P 2>/dev/null || printf 'timestamp failure! ...meh' )"
        else
            fail_message
        fi
    else
        printf 'Re-encoding FLAC file(s). 5 Replay Gain tags already exist.\n'
        if flac $flac_options "${flacfiles[@]}" ;then
            printf 'Success! @ %s.\n\n' "$( date +%_I:%M%P 2>/dev/null || printf 'timestamp failure! ...meh' )"
        else
            fail_message
        fi
    fi
}

SHELL="$( type -p bash )" # https://stackoverflow.com/a/23815646
export SHELL              # SC2155: Declare and assign separately to avoid masking return values
export -f fail_message actions
trap "killall -HUP parallel" INT # supposed to complete any active jobs and not start any new ones, instead: "^Cparallel: no process found" -- functions are so not frustrating!
parallel --will-cite --bar --tag -j "$parallel_jobs" actions ::: "${dirs_in_tree[@]}"



# TODO/IDEAS/PONDERINGS
# how can we tell when a folder has failed and needs re-doing?
# or tell which folders have been done already if we stop in the middle and start again?

# can we "mkfifo failed_folders" and "printf '%s\n' "$1" >> failed_folders" ??
#     if potentially thousands of iterations of the script are being run, seems pretty clear mkfifo shouldn't run in every one
#     can all the iterations even write to the same fifo? _without_ breaking somehow??
# is a simple text file in /tmp easier? does that work when more than one iteration of the script tries to add a line at the same time?

# does flac do the same as "metaflac --sort-padding" to any existing padding (or even just
# replace any existing padding when re-encoding a flac file and using the -P/--padding option ??

# flac --force leaves behind source files when their extension includes uppercase characters