arxiv2bibtex

#!/bin/bash

# Script for generating a bibtex entry from a given arXiv identifier
#      
# Usage:    arxiv2bibtex arXiv
#           arxiv2bibtex http://arxiv.org/abs/arXiv
#      
# Example:  arxiv2bibtex 1412.3740 >> ~/World.bib
# Example:  arxiv2bibtex math/0602029 >> references.bib (old arXiv format)


# Issue help message if necessary
if [ $# = 0 ] || [ "x$1" = "x--help" ]; then
	>&2 echo "Usage: $(basename $0) arXiv"
	>&2 echo "or:    $(basename $0) http://arxiv.org/abs/arXiv"
	>&2 echo "will try and retrieve a bibtex entry for the arXiv identifier given."
	>&2 echo
	>&2 echo "Example: $(basename $0) 1701.06092"
	>&2 echo "         $(basename $0) 1701.06092v1"
	>&2 echo "         $(basename $0) http://arxiv.org/abs/1701.06092v1"
	>&2 echo "         $(basename $0) math/0602029"
	>&2 echo "         $(basename $0) http://arxiv.org/abs/math/0602029"
	exit 1
fi


# Unify the calling syntax
arxivpattern='([0-9]{4}\.[0-9]{4,}([v][0-9]*)?|[a-z]{1,}/[0-9]{7})$'
if [[ $1 =~ $arxivpattern ]]; then
	arxiv=${BASH_REMATCH[1]}
else
	echo "Failed to find arXiv pattern."
	exit 1
fi

# echo $arxiv


# arxivnodots="${arxiv//./}"
# 
# URL="http://adsabs.harvard.edu/cgi-bin/bib_query?arXiv:$arxiv"
# 
# # Get URL of bibtex entry
# URL=$(wget -O - $URL | grep 'Bibtex entry for' | perl -ne '/<a href="(.*?)">/ and print "$1\n"')
# 
# wget -O - $URL

# Better?
URL="https://arxiv2bibtex.org/?q=$arxiv&format=biblatex"

ENTRY=$(wget -O - $URL -q | xmllint  --xpath '//div[@id="biblatex"]/textarea/text()' -)


# If we have bibtool (https://ctan.org/pkg/bibtool) we can use it to format the key
if [ -x "$(command -v bibtool)" ] ; then
	ENTRY=$(echo "$ENTRY" | bibtool -f "%n(author)%4d(year)_preprint" \
		-- "new.entry.type {online}" \
		-- 'fmt.name.name = ""' \
		-- 'add.field {type = "Preprint"}' \
		-- 'print.align.key = 0'\
		-- 'print.line.length = 1000'\
		-- 'print.use.tab = 0'\
		)

	ENTRY=${ENTRY/#$'\n'}   # Remove a leading newline.
fi

echo "$ENTRY"

echo # Add an additional newline