Audio Diary

This is a frugal audio diary, which receives e-mails with attached voice messages, performs speech recognition, creates a RSS feed from the results, and adds a stylesheet to create a web page from the feed.

Installation

1. Prepare local directory that is reachable on your web server; this will be the feed directory of the diary; it should be owned by user nobody and group nogroup; note the URL and the absolute path of the directory
2. Place files enqueue.py, daemon.py, and tohtml5.xsl into a separate directory; the user nobody should be able to execute files there; install whisper.cpp into the same directory, also download the medium model to the same location; should you install whisper.cpp to another directory, please adjust the paths in the file daemon.py; install ffmpeg to convert audio files to the format preferred by whisper.cpp (for example on Debian: apt-get install ffmpeg)
3. In file aliases, replace username with the user part of the e-mail address you will send e-mails with audio attachments to, and the /feed_dir with the absolute path of the feed directory; append the resulting file to the system-wide aliases file (for example on Debian: /etc/aliases) and update the mail aliases database by executing newaliases
4. In file audiodiary.service, replace /feed_dir with the absolute path of the feed directory, and https://feed_url with it's URL; copy the file to the systemd services directory (for example on Debian: /etc/systemd/system), start it using command systemctl start audiodiary and test the setup by sending an e-mail with audio attachment to the e-mail address of your choice
5. Persist the setup by executing command systemctl enable audiodiary
6. Access the URL you entered in step 4; if you see list of files and your test file, everything works as expected; access the file index.xml therein; you should see a web page with your test file embedded and with the recognised text beside it