fw5.6.1+ — Is abuse of the author field only option for library management?

[danjcla]

Is abusing either the title or author field - along the lines of the below proof-of-concept script - pretty much the only way to do off-device / non-manual organization of books on a Kindle Paperwhite 2 with firmware 5.6.1 (that hadn't previously been rooted)?

I've read a lot of stuff on these forums and that would seem to be the case, but I haven't found a post that says it point blank, so thought I'd ask, and also share the atrocious ugliness below :-)

#!/bin/bash



# This script changes the author metadata to be in this form with <> as 
# the delimiter:
# <>directory the book is in<>original author name<>

# -->> Need to be made idempotent before it'll be really usable day-to-day

# It'd also be useful to add "X of Y" to files in a directory, so it'd be easy
# to manually make sure everything in a directory is also in a collection.



EBOOKMETA="/Applications/calibre.app/Contents/MacOS/ebook-meta"
# http://manual.calibre-ebook.com/cli/ebook-meta.html

find /Volumes/Kindle -type f -name '[a-zA-Z0-9]*.mobi' -print0 | while read -d $'\0' FILE
do
  AUTHOR="$(/Applications/calibre.app/Contents/MacOS/ebook-meta "$FILE" --to-opf=/dev/null | grep ^Author | sed 's/Author(s)           : //g')"
  FOLDER="$(basename $(dirname "$FILE"))"
  $EBOOKMETA "$FILE" -a "<>$FOLDER<>$AUTHOR<>"
done

[NiLuJe]

Note that, if you're already using Calibre, a Calibre plugboard is usually the best way to achieve that kind of shenanigans . The fact that such a feature exists pretty much answers your question: yep, it's a widely used workaround, and has been for quite a while .

[danjcla]

Thanks! I'm new to all this and had no idea that feature existed. I'll def be exploring it :-)

[danjcla]

The more I look into it the less it seems like templates / the GUI (sans writing or finding some custom module) in general could do what I'm interested in doing.

That is support a fast workflow where I occasionally download a bunch of .mobi files via a web browser directly to my kindle, putting them in / creating the right categorization folder at that time, and then run some program that abuses the Author metadata to allow it to be used for categorization based on the folder containing the file.

It looks like the GUI workflow would be more like this:
1. Download .mobi files to temporary directory structure on machine running Calibre
2. Open Calibre
3-4. Import files into Calibre in batches from each temporary directory
3-4. For each batch, add tag equivalent to temporary folder name
5. Do "send to device" (I keep looking for a sync all button...)
6. Remove files, but not directories, from local temporary directory structure

Or alternatively, maybe download to kindle, then copy over to local using the GUI, then copy back (filtered through plugboard and with template that will put files in same folder hierarchy).

The two GUI options just seem... inefficient and annoying.

Also I have a laptop, work desktop, home desktop, etc., not sure how I'd keep Calibre in sync across all those machines.

But maybe this isn't the right forum for this question? If so please tell me and I'll re-ask elsewhere.

[NiLuJe]

That completely depends on how willing you are to commit to actually using Calibre to manage your library. If you don't, it might indeed not be the most efficient workflow .

(I personally keep my desktop's & laptop's Calibre library via rsync).

[danjcla]

Cool, at least good to know I'm not missing something. My major pain point is the initial collection of metadata... right when I download something I know how to categorize it, but 10 minutes later when I'm done downloading stuff to look over and maybe read, if it's all in one folder, it's hopeless.

Hmmm although thinking about it more I could just rename the file instead of put it in a subdirectory, and then the import could use that info, since "Add books -> Control the adding of books" looks like it can do anything based on file names ...

d'oh

Thanks again,
-Danny

[danjcla]

The path of least resistance seemed to be to just spend a few more minutes to fix up the script a bit; this version you can run multiple times without messing anything up, and it'll also correctly change the author metadata based on moving a file to a new directory.

#!/bin/bash

EBOOKMETA="/Applications/calibre.app/Contents/MacOS/ebook-meta"
KINDLEROOT="/Volumes/Kindle"

# --

if [ ! -x $EBOOKMETA ]; then
  echo "ERROR: $EBOOKMETA doesn't exist"
  exit 200
fi

if [ ! -d $KINDLEROOT ]; then
  echo "ERROR: $KINDLEROOT doesn't exist"
  exit 201
fi

# --

find $KINDLEROOT -type f -name '[a-zA-Z0-9]*.mobi' -print0 | while read -d $'\0' FILE
do

  AUTHOR="$(/Applications/calibre.app/Contents/MacOS/ebook-meta "$FILE" --to-opf=/dev/null | grep ^Author | sed 's/Author(s)           : //g')"
  FOLDER="$(basename $(dirname "$FILE"))"

  if [[ $AUTHOR == *".."* ]]; then
    METAAUTHOR="$(echo $AUTHOR | awk -F '\\..' '{print $3}')"
    $EBOOKMETA "$FILE" -a "..${FOLDER}..${METAAUTHOR}.."
  else
    $EBOOKMETA "$FILE" -a "..${FOLDER}..${AUTHOR}.."
  fi

done

[danjcla]

Thread moved to:

MobileRead Forums > E-Book Software > Calibre > Library Management > 'Directory tree in-place "author" -> "folder + author" script (where folder=category)'