How to process paper pile-exported bibfiles so that emacs / helm-bibtex / ivy-bibtex can work with them.

paperpile.sh

#!/bin/bash

# Get the newest paperpile and fix the format

# Argument 1: Paperpile folder name to look for 

# This is messy because paperpile insists on putting spaces in file names and gnu make can't handle that :-(

# This is also messy because paperpile has such a weird file format; not really compatible with ivy-bibtex / helm-bibtex in emacs. 

sed_command1='s!file *= "All Papers!file     = "/Users/hkarl/Google Drive/Paperpile/All Papers!'
sed_command2='s!,Starred Papers!!'

# find newest download file; careful with the spaces
# beware of: http://mywiki.wooledge.org/ParsingLs

download=`ls -t ~/Downloads/Paperpile*.bib | grep $1 | head -1`

if [ -z "${download}" ]
then
    echo "no download file -- nothing to do"
else
    echo "download file found: ${download}"
    
    if [ -f paperpile.bib ]
    then
	echo "paperpile exists -- comparing date"
	if [ "${download}" -nt paperpile.bib  ]
	then
	    echo "download is younger; copy it"
	    sed "${sed_command1}"  "${download}" | sed "${sed_command2}" > paperpile.bib
	else
	    echo "no newer download found"
	fi
    else
	echo "download, but no local file"
	sed "${sed_command1}"  "${download}" | sed "${sed_command2}" > paperpile.bib
    fi
fi