bandono / kobo-annot-import

With Kobo e-Reader we can highlight EPUB book as annotation that will be kept in the reader storage as XML files.

This Python script uses combination of find,rsync, and grep to extract the annotations and save them as individual Markdown file using the annotation text (in my case as sort of title) for naming.

I'm using Mac OSX so, plugging my Kobo Clara HD will mount it to /Volumes/KOBOeReader. I use one line where:

1. I change dir to the e-Reader (cd)
2. find annotation files
3. Compare those files whether they are newer or the same with the ones kept on my laptop
4. Copy only newer files to my laptop (rsync)
5. Pipe the list of these newer files for processing the XML into individual markdown files

cd /Volumes/KOBOeReader && find ./ -name "*.annot" -type f -print0 2>/dev/null | rsync -av --files-from=- --from0 --no-relative /Volumes/KOBOeReader <directory where I keep my XML files on my laptop> | grep annot | /<pyenv-version>/python kobo-annot-import.py

In the Python script I still initiate folder for XML and markdown files.

# Define annotation folder
annot_folder = '' # replace with annotation local folder
# Define result folder root
folder_name = 'result' # replace with target folder

It will check whether same markdown with the name Some-Annotation-Text-as-Title-example.md exists and prevent from overwriting it, assuming the one on the laptop has already been modified. It will then append to Some-Annotation-Text-as-Title-example (1).md if content doesn't match newer annotation.

This is a work in progress. Always backup properly if not sure on result.

kobo-annot-import's Issues

When creating newer annotation on KOBO same UUID maybe used again

Duplicate UUIDs were found to have different annotation content. So can't safely use UUID as file name of the generated annotation file.

Here is example UUID case: 93b5dc68-f29c-40a6-9d85-e478e26f7988

Since this is beyond us as user, simply just change naming to use other than UUID.