website/removeBadSymbols.sh

#!/bin/bash
# this is a shebang. It guarantees that this script is executed with bash, not the currently open shell
# problem with shanging to sh: I can't get test to work with * wildcards :(

#todo: also handle hidden files -- actually doable but not a good idea
#todo: bei jeder Aenderung auch usage (doc) anpassen
#todo: translate entire script into english

# options
# -h --help                      display help
# -a --noask --no-interactive    do not prompt the user to confirm renaming
# -l --onlylog                   just logs what would have been done but not doing anything
# the following two MUST be present # TODO: not making them mandatory
# -o [file] --outputfile [file]  specifies the output (log) file (standard is rn.log)
# -v [file] --voutputfile [file] specifies the verbose output log file (standard rnv.log)
# -rm                            removes files of the form AlbumArt_*, Folder.jpg, Thumbs.db, desktop.ini
# -c                             converts file names to lower case

#Constants
# weil im folgenden ein Apostroph vorkommt, kann man diesen String nicht einfach einfuegen
STRICHSTR=s/\'//g
# auch hier muss " escaped werden
ANFUERSTR=s/\"//g
# und hier ist es `
ANDSTRSTR=s/\`//g
# und hier ist es &
AMBERSSTR=s/\&/_und_/g
# und hier ist es *
STERNXSTR=s/\*/x/g

#Functions
vLog()
{
  echo "$1" >> "$VLOGFILE"
}

log()
{
  echo "$1" >> "$NVLOGFILE"
  echo "$1" >> "$VLOGFILE"
}

rename()
# moves $1 to $2 if rn is 1, 
# not anymore: otherwise set variable new to $1
{
  log "rn \"$1\" to \"$2\""
  if test $rn -eq 1
  then
    mv -- "$1" "$2"
  # else # not necessary since $new is not used anymore
    # new="$1"
  fi
}


remove()
# deletes $1 if rn is 1, otherwise only logging
{
  if test -f "$1"
    then
    #echo "$1"
    log "rm \"$1\""
    if test $rn -eq 1
    then
      trash "$1"
    fi
  fi
}

usage()
# displays help message and exit
{
  echo ""
  echo "help page for $0"
  echo "---------------------------------"
  echo "this script should not be called directly but by the Wrapper script WrapperRemoveBadSymbols.sh"
  echo "Usage: bash $0 [OPTIONS] DIR"
  echo "  $0 removes a bunch of symbols that should not be in file names in all subdirs and files of DIR:"
  echo "  (with lyrics) -> remove"
  echo "  lyrics -> remove"
  echo "  blank -> _"
  echo "  ' - ' -> '-'"
  echo "  ' (Apostroph) -> remove"
  echo "  '\"'' (Anfuerungsstriche) -> remove"
  echo "  ´ -> remove"
  echo "  \& -> _und_"
  echo "  * -> x"
  echo "  # -> _"
  echo "  | -> l (small L since it looks similar)"
  echo "  ~ -> -"
  echo "  : -> _"
  echo "  , -> _"
  echo "  (){}[] -> remove"
  echo "  € -> EUR"
  echo "  $ -> USD"
  echo "  ä, ö, ü, Ä, Ü, Ö, ß -> ae, ..."
  echo "  remove various accents (not all, you can add missing ones)"
  echo "  consecutive _ and - and _ around - -> remove"
  echo ""
  echo "options"
  echo "  -h, --help                      display this help"
  echo "  -a, --noask --no-interactive    do not prompt the user to confirm renaming (standard is asking for each file)"
  echo "  -l, --onlylog                   just logs what would have been done but not doing anything"
  echo "  -o [file], --outputfile [file]  specifies the output (log) file (standard is rn.log)"
  echo "  -v [file], --voutputfile [file] specifies the verbose output log file (standard rnv.log)"
  echo "  -rm                             removes files of the form AlbumArt* and Folder.jpg and Thumbs.db"
  echo "  -c                              change everything to lowercase (doesn't work for foreign letters if diacritics remain"
  echo "  --                              finishes the list of options, next argument is taken as the dir"
  #echo "  -i, --image                     adds date taken in beginning of file name if jpg, .png, (not yet: .mp4, .mov, ...)"
  echo ""
  echo "error handling"
  echo "  if DIR is no valid directory, the scripts eith error code 1"
  exit
}

folderloc()
# returns the argument as absolute path
# arg1: if arg1 starts with /, return arg1, otherwise return $(pwd)/arg1
{
  case "$1" in
    /* ) echo "$1"
        break
      ;;
    * ) echo "$(pwd)/$1"
        break
      ;;
  esac
}

# considering options
# saving all options to give them recursively
ops=""
# standard behavior: ask the user is file is to be renamed
# 0 meens not asking
ask=1
# standard behavior: actually rename
# 0 meens not renaming but just log what would have be done
rn=1
# remove album pictures
# standard behavior: not removing
rmAlbumPics=0
# if file names should be converted to lower case
# standard behavior: no
lowercase=0

# the log file for verbose logging
#VLOGFILE="$1"
VLOGFILE="$(pwd)"/../rnv.log
# the log file for short logging
#NVLOGFILE="$2"
NVLOGFILE="$(pwd)"/../rn.log

#directory to be checked
dir="$(pwd)" # standard if not specified

while [ "$1" != "" ]; do
    case $1 in
        -h | --help )           usage;;
        -o | --outputfile )     # ops="$ops -o ../$2"
                                shift
                                NVLOGFILE="$(folderloc "$1")"
                                ;;
        -v | --voutputfile )    #ops="$ops -v ../$2"
                                shift                                
                                VLOGFILE="$(folderloc "$1")"
                                ;;
        -a | --noask | --no-interactive ) ops="$ops -a"   
                                ask=0
                                ;;
        -l | --onlylog )        ops="$ops -l"
                                rn=0
                                ;;
        -rm )                   ops="$ops -rm"
                                rmAlbumPics=1
                                ;;
        -c )                    lowercase=1
                                ;;
        -- )                    shift
                                dir="$(folderloc "$1")"
                                break
                                ;;
        * )                     dir="$(folderloc "$1")"
                                break
                                ;;
    esac
    shift # must be done in any case: the next argument is $1
done

# echo "options"
# echo "asking=$ask"
# echo "rename=$rn"
# echo "image=$image"
# echo "log file=$NVLOGFILE"
# echo "verbose log file=$VLOGFILE"
# echo "direc=$dir"

#for testing purposes:
# exit

improveNames()
# this function does the work for one folder and goes into subfolders recursively
# assumes it is already in the right place
{
  vLog "start with folder: $(pwd)"
  printf '.'
  # delete unnecessary album pictures and other useless files
  if test $rmAlbumPics -eq 1
  then
    for alb in AlbumArt*.jpg
    do
      remove "$alb"
    done
    remove Folder.jpg
    remove Thumbs.db
    remove desktop.ini
  fi

  for fil in ./* # ../{.[^.],..?,}*
  # that is: .NOTDOTanything &
  # ..anything but something &
  # something, not starting with dot
  # also working on hidden files is not good, e.g. there are .git and many config files
  do
    # * als einziger Dateiname kommt auch bei leerem Ordner vor
    if test -e "$fil" # file to be moved exists
      then
      :
      # do nothing
    else
      continue # continue with next file
    fi

    # with lyrics entfernen
    new="$(echo "$fil" | sed 's/(with lyrics)//g')"
    # lyrics entfernen
    new="$(echo "$new" | sed 's/lyrics//g')"
    # Leerzeichen entfernen:
    new="$(echo "$new" | sed 's/ /_/g')"
    # Apostroph entfernen:
    new="$(echo "$new" | sed $STRICHSTR)"
    # anderes Apostroph entfernen:
    new="$(echo "$new" | sed $ANDSTRSTR)"
    # Anfuerungstriche entfernen:
    new="$(echo "$new" | sed $ANFUERSTR)"
    # wieder anderes Apostroph entfernen:
    new="$(echo "$new" | sed 's/´//g')"
    # Ambersands (&) entfernen:
    new="$(echo "$new" | sed $AMBERSSTR)"
    # Sternchen (*) entfernen:
    new="$(echo "$new" | sed $STERNXSTR)"
    # Rauten als Teil von Csharp:
    new="$(echo "$new" | sed 's/C#/C_sharp/g')"
    new="$(echo "$new" | sed 's/c#/c_sharp/g')"
    # Rauten entfernen:
    new="$(echo "$new" | sed 's/#/_/g')"
    # vertical bars |
    new="$(echo "$new" | sed 's/|/l/g')"
    # comma, entfernen
    new="$(echo "$new" | sed 's/,/_/g')"
    # parenthesis
    new="$(echo "$new" | sed 's/{//g')"
    new="$(echo "$new" | sed 's/}//g')"
    new="$(echo "$new" | sed 's/(//g')"
    new="$(echo "$new" | sed 's/)//g')"
    # [] need escaping since they are part of regexps
    new="$(echo "$new" | sed 's/\[//g')"
    new="$(echo "$new" | sed 's/\]//g')"
    # Tilde ~
    new="$(echo "$new" | sed 's/~/-/g')"
    # :
    new="$(echo "$new" | sed 's/:/_/g')"
    # diacritics
    # y as sed command says replace one character with the corresponding
    # here you might want to add letters when the need arises
    while true
    do
      new2=$new
      new="$(echo "$new" | sed 'y/ãāáǎàēéěèȩêīíǐìĩïōóǒòũūúǔùǖǘǚǜşļĻķĶḩģĢḨņŅŗŖĀÁǍÀĒÉĚÈĪÍǏÌŌÓǑÒŪÚǓÙǕǗǙǛ/aaaaaeeeeeeiiiiiioooouuuuuüüüüslLkKhgGHnNrRAAAAEEEEIIIIOOOOUUUUÜÜÜÜ/')"
      if test "$new2" = "$new"
      then
        break
      fi
    done    
    # äa
    new="$(echo "$new" | sed 's/ä/ae/g')"
    # Äa
    new="$(echo "$new" | sed 's/Ä/Ae/g')"
    # öa
    new="$(echo "$new" | sed 's/ö/oe/g')"
    # Öa
    new="$(echo "$new" | sed 's/Ö/Oe/g')"
    # üa
    new="$(echo "$new" | sed 's/ü/ue/g')"
    # Üa
    new="$(echo "$new" | sed 's/Ü/Ue/g')"
    # ßs
    new="$(echo "$new" | sed 's/ß/ss/g')"
    # $
    new="$(echo "$new" | sed 's/\$/USD/g')"
    # €
    new="$(echo "$new" | sed 's/€/EUR/g')"
    # different jpeg extensions:
    new="$(echo "$new" | sed 's/\.JPG/.jpg/g')"
    new="$(echo "$new" | sed 's/\.JPEG/.jpg/g')"
    new="$(echo "$new" | sed 's/\.jpeg/.jpg/g')"
    if test $lowercase -eq 1 # if should be converted to upper case
    then
      new="$(echo "$new" | tr '[:upper:]' '[:lower:]')"
    fi
    #log "before complicated: $fil -> $new"
    for i in "_" "\." "-"
    do
      # remove _ around i
      isymbostr="s/_$i/$i/g"
      new="$(echo "$new" | sed $isymbostr)"
      #log "step 1 $i: $new"
      # _ um i entfernen:
      isymbostr="s/$i""_/$i/g"
      #echo $isymbostr
      new="$(echo "$new" | sed $isymbostr)"
      #log "step 2 $i: $new; $isymbostr"
      # Remove several consecutive
      while true
      do
        isymbostr="s/$i$i/$i/g"
        new2="$(echo "$new" | sed $isymbostr)"
        if [ $new2 = $new ]
          then break
        else
          new=$new2
        fi
      done
      #log "step 3 $i: $new"
    done

    if test "$new" = "." -o "$new" = ".."
    then
      log "dont rename bc would result in $new : $(pwd)/$fil"
    elif test "$fil" = "$new"
    then
      vLog "dont rename bc no change: $fil"
    elif [[ "$fil" == ./__*__.py || "$fil" == ./__*__.html || "$fil" == ./__*__.txt ]]
    then
      vLog "dont rename bc probably special python file: $fil"
    elif [[ "$fil" == *.class ]]
    then
      # sollten unproblematisch sein, beinhalten aber $s
      vLog "dont rename bc probably special java class file: $fil"

      # existiert eine Datei mit dem neuen Namen bereits?
    elif test -e "$new" # bisher: -f (file ex. and is regular file but folders are problematic as well?!)
    then
      log "a file with name \"$new\" already exists, -> dont rename: $(pwd)/$fil"
    elif test $ask -eq 0
    then
      rename "$fil" "$new"
    else
      # Abfrage ob man wirklich umbenennen will
      echo "rename $fil to $new? enter, y, yes = yes"
      read q
      if [ "$q" = "" -o "$q" = "y" -o "$q" = "yes" ]
      then
        rename "$fil" "$new"
      fi
    fi
  done
  # renaming done, now iterate over new names
  for dir in ./*; do #./{.[^.],..?,}*; do
    if test -d "$dir"
    then
      cd "$dir"
      improveNames # arbeite im Unterordner
      cd ..
      vLog "went back to folder $(pwd)"
    fi
  done
  
}

#actual script

# check if the argument is correct:
if [ ! -d "$dir" ]
then
  log "exiting bc there is no such dir"
  #echo "exiting bc there is no such dir"
  exit 1
fi

# vLog "move to $dir"
cd "$dir"

# reset log files
echo "" > "$NVLOGFILE"
echo "" > "$VLOGFILE"

improveNames

# merke: in Bedingungsklammern [] von if muss am Anfang und am Ende ein " " (blank) stehen!!
# merke: Hinweise zu Bedingungen in if abrufbar unter "man test"
# merke: bei Zuweisungen mit = darf zwischen Variablenname und = kein Leerzweichen sein! 
# (sonst wird Variablenname als command interpretiert)