gappi

#!/bin/bash
#
# Get A Phil Papers Item (gappi); version 0.3
# by Kevin C. Klement ([email protected])
# Feel free to email Kevin with suggestions or questions
# License: GPLv3
#

# get the name of the script
SCRIPTNAME="$(basename "$0")"

# function to show usage info
usage_info() {
    cat > /dev/stderr << EOF
USAGE: $SCRIPTNAME [-n NN] "search item 1" "search item 2" ...

For each search item, gappi will fetch the BibTeX record  
for the first [or first NN] PhilPapers result for that search item. Each
item's individual search parameters should be enclosed in quotes.

You can use the additional search parameters listed at. 

   http://philpapers.org/help/search.html

E.g.:
$SCRIPTNAME "@authors Strawson @title On Referring"

Note that it is "@authors" with an s, not "@author", even
when there is only one.

Use an backslash escaped \" for a literal quotation mark.
EOF
}

MATCHES=1  # default number of matches to return

# if there are no search items given, then show usage info
if [[ "$#" == 0 ]] ; then
    usage_info
fi

# loop for each search item
while [[ "$#" -gt 0 ]] ; do
    # help users who ask for it 
    if [[ "$1" == "-help" ]] ||
       [[ "$1" == "--help" ]] ||
       [[ "$1" == "-h" ]] ||
       [[ "$1" == "-?" ]] ; then
        usage_info
        exit 0
    fi
    # -n NN sets the number of matches to be returned
    if [[ "$1" == "-n" ]] ; then
        shift 
        if [[ "$1" -gt 0 ]] ; then
            MATCHES="$1"
            shift 
        else
            echo "The option '-n' must be followed by a postive integer."
            usage_info
            exit 0
        fi
    fi
        
    # replace spaces with %20 for URLS
    SEARCHSTRING="${1// /%20}"
    # get BibTeX entry for that match, send to stdout
    curl --silent --data-urlencode -G \
        -d "search_header=search_header.html&searchStr=$SEARCHSTRING&filterMode=keywords&limit=$MATCHES&newWindow=on&sort=relevance&showCategories=on&proOnly=on&format=bib" \
        "http://philpapers.org/asearch.pl" 2>/dev/null 
    # add a line of padding
    echo ""
    # move to next search item
    shift
done
exit 0

Simplifies Kevin Klement's gappi script---just one call to curl to fetch the bibtex entry for each search item---and adds a command line option, -n NN, to specify the number of bibtex entries to return for each search term.

Thanks. I'll go ahead and incorporate these changes on my site, only fixing the typo "pos[i]tive", and having that go to /dev/stderr, since it's an error message, and also exiting with an error code "exit 1" instead of "exit 0" in that case.