#!/bin/sh

#
# Given https://mupdf.com/release_history.html URL generates entry for every
# news item
#


fetch()
{
	url=$1

	curl -g -L --max-redirs 2 -f -s -S -m 15 "$url"
}


extract_news_entry()
{

	xmllint --format --html --xpath '//h3' - 2>/dev/null |
	    sed -E -e 's;</?h3>;;g' -e '/^$/d' -e 's/^ +//g' -e 's/ +$//g'
}


sfeedify()
{
	awk '
BEGIN {
	author = "Artifex"
	base_link = "https://mupdf.com/release_history.html"
}

{
	date = substr($NF, 2, length($NF) - 2)
	title = substr($0, 1, length($0) - length($NF) - 1)

	id = tolower($0)
	gsub(/ /, "-", id)
	gsub(/\./, "_", id)
	gsub(/[()]/, "", id)

	link = base_link "#" id

        gsub(/[^A-Za-z0-9:+ ]/, "", date)

        # XXX: this is not portable (NetBSD date(1))
        cmd = sprintf("date -ud \"%s\" +%%s", date)
        cmd | getline timestamp
        close(cmd)

	print timestamp "\t" title "\t" link "\t" content "\t" content_type "\t" id "\t" author "\t" enclosure
}
'
}


for u in "$@"; do
	fetch "${u}" | extract_news_entry | sfeedify
done