#!/bin/awk -f
# Convert sfeed(1) formatted files into an HTML webpage.
# Usage: sfeed_html.awk -- FILES...

## Set up variables
BEGIN {
	# Sidebar
	if (! ASIDE) {
		ASIDE = envor("SFEED_HTML_ASIDE", "/tmp/sfeed-aside.html")
	}
	print > ASIDE	# clear out ASIDE
	in_aside = 0	# Don't show the feed in the sidebar
	# Formatting
	if (! DATEFMT) {
		ddf = "<time title=\"%F %R UTC\">"
		ddf = ddf "<span class=\"date\">%m-%d</span>" " "
		ddf = ddf "<span class=\"time\">%R</span>"
		ddf = ddf "</time>"
		DATEFMT = envor("SFEED_HTML_DATEFMT", ddf)
	}
	if (! TITLE) {
		TITLE = envor("SFEED_HTML_TITLE", "Planet ACDW")
	}
	if (! SEPARATOR) {
		SEPARATOR = envor("SFEED_HTML_SEPARATOR", "//")
	}
	if (! ALT_LINK_STAMP) {
		ALT_LINK_STAMP = envor("SFEED_ALT_LINK_STAMP", "&")
	}
	if (! ENCLOSURE_STAMP) {
		ENCLOSURE_STAMP = envor("SFEED_ENCLOSURE_STAMP", "@")
	}
	if (! SILO_STAMP) {
		SILO_STAMP = envor("SFEED_SILO_STAMP", "%")
	}
	# Limiting posts...
	## by time
	if (! NOW) {
		datecmd = "date +%s"
		datecmd | getline NOW
		close(datecmd)
	}
	if (! FRESHDAYS) {
		FRESHDAYS = envor("SFEED_FRESHDAYS", 1.5)
	}
	if (! STALEDAYS) {
		STALEDAYS = envor("SFEED_STALEDAYS", 4)
	}
	fresh_secs = FRESHDAYS * 24 * 60 * 60
	stale_secs = STALEDAYS * 24 * 60 * 60
	fresh_age = (NOW - fresh_secs)
	stale_age = (NOW - stale_secs)
	## by number
	if (! LIMIT) {
		# If LIMIT == -1, ignore time limit as well.
		LIMIT = envor("SFEED_LIMIT", 20)
	}
	# Alternate URLs for siloed content
	if (! YOUTUBE_ALT_URL) {
		YOUTUBE_ALT_URL = envor("SFEED_YOUTUBE_ALT_URL", "https://piped.kavin.rocks")
	}
	if (! TWITTER_ALT_URL) {
		TWITTER_ALT_URL = envor("SFEED_TWITTER_ALT_URL", "https://nitter.net")
	}
	if (! REDDIT_ALT_URL) {
		REDDIT_ALT_URL = envor("SFEED_REDDIT_ALT_URL", "https://libreddit.de")
	}
	if (! MEDIUM_ALT_URL) {
		MEDIUM_ALT_URL = envor("SFEED_MEDIUM_ALT_URL", "https://scribe.rip")
	}
	# Paywalls
	PAYWALL_LADDER = "https://archive.ph/newest/"
	PAYWALLED[1] = "washingtonpost\\.com"
	PAYWALLED[2] = "nytimes\\.com"
	# Awk and convenience constants
	FS = "\t"
	STDERR = "/dev/stderr"
}

BEGIN {
	html_top()
}

FNR == 1 {
	if (file_count) {
		# End the previous file before beginning the current one
		end_file()
	}
	# Filename
	name = FILENAME
	sub(/.*\//, "", name)
	safe_name = name
	sub(/\//, "_", safe_name)
	dir = FILENAME
	sub(/[^\/]*$/, "", dir)
	if (name ~ /\[yt\]$/) {
		sub(/ \[yt\]$/, "", name)
		yt = " class=\"yt\""
	} else {
		yt = ""
	}
	# extra metadata, yaaaaay
	feed_url = ""
	site_url = ""
	URLS = envor("SFEED_DATA", ENVIRON["HOME"] "/.sfeed") "/urls/" safe_name
	getline feed_url < URLS
	getline site_url < URLS
	getline feed_tags < URLS
	# put more lines to read here
	close(URLS)
	# State variables
	stamp = "."
	buf = ""
	ni = 0
	stale_total += stale_count
	stale_count = 0
	fresh_total += fresh_count
	fresh_count = 0
	item_total += item_count
	item_count = 0
	file_count++
	siloed = 0
}

{
	# Skip if we have too many items
	if ((LIMIT > 0) && (++ni > LIMIT)) {
		next
	}
	# Collect fields
	timestamp = $1
	title = $2
	silo = $3
	link = silo_link(silo)
	content = unescape($4)
	content_type = $5
	id = $6
	author = $7
	enclosure = $8
	category = $9
	# Skip if the item is too old
	if ((LIMIT >= 0) && (timestamp < stale_age)) {
		next
	}
	# Otherwise, we're showing it
	stale_count++
	# Is this item fresh?
	if (timestamp >= fresh_age) {
		is_fresh = 1
		stamp = "!"
		fresh_count++
	} else {
		is_fresh = 0
	}
	# Debugging comment
	# bufprint("<!-- timestamp title silolink type id author enclosure category -->")
	# bufprint("<!-- " $1 " -- " $2 " -- " $3 " -- " $5 " -- " $6 " -- " $7 " -- " $8 " -- " $9 " -->")
	# Print!
	bufprint("<tr class=\"entry " (is_fresh ? "fresh" : "stale") "\">")
	# Timestamp
	datecmd = "date -d \"@" timestamp "\" +'" DATEFMT "'"
	if (timestamp) {
		datecmd | getline timestamp
		close(datecmd)
	}
	bufprint("<td class=\"entry-timestamp\"><time>" timestamp "</time></td>")
	# Extra links
	bufprint("<td class=\"entry-extra\">")
	if (siloed) {
		# "siloed" links like youtube, facebook, etc. --- I convert them
		# to more privacy-friendly links, but sometimes those don't
		# work. TODO: also try to circumvent paywalls.
		stamp = SILO_STAMP
		print_link(silo, "Silo: " silo, stamp)
	}
	if (enclosure) {
		# enclosures --- podcast files, etc.
		stamp = ENCLOSURE_STAMP
		print_link(enclosure, "Enclosure: " enclosure, stamp)
	}
	# Test id against silo here, since silo is the raw link value.
	if ((id != silo) && (id != enclosure) && (id ~ /^https?:/)) {
		# alternate links (comments, etc.)
		stamp = ALT_LINK_STAMP
		print_link(silo_link(id), "alternate link", stamp)
	}
	bufprint("</td>")
	# Title
	bufprint("<td class=\"entry-title" silo_class(silo) "\">")
	# bufprint("<details><summary>", 0)
	print_link(link, "", title)
	# bufprint("</summary>")
	# bufprint(content)
	# bufprint("</details>")
	bufprint("</td>")
	# End row
	bufprint("</tr>")
	item_count++
}

END {
	# End the last file and the #list section
	end_file()
	print "</section>"
	# Sidebar
	print "<aside><ul id=\"feedlist\">"
	system("cat " ASIDE)
	close(ASIDE)
	print "</ul></aside>"
	# End of <main>
	print "</main>"
	# footer
	printf "<footer>"
	print "Generated by <a href=\"https://codemadness.org/sfeed-simple-feed-parser.html\">sfeed</a>"
	print "using <a href=\"https://git.acdw.net/sfeed/\">this configuration</a>."
	print "<a href=\"mailto:planet@me.acdw.net\">suggest&nbsp;a&nbsp;feed!</a></footer>"
	# end of HTML
	print "<script src=\"script.js\"> </script>"
	print "</body>"
	print "</html>"
	printf(SEPARATOR) > STDERR
}


function bufprint(text, sep)
{
	buf = buf text (sep ? sep : ((sep == 0) ? "" : "\n"))
	sep = ""
}

function end_file()
{
	if (! item_count) {
		return 1
	}
	# Header
	printf "<section id=\"%s\" class=\"feed %s\">\n", name, (fresh_count ? "fresh" : "stale")
	printf "<header><h2%s><a class=\"anchor\" href=\"#%s\">#</a> %s</h2>\n", yt, name, name
	# Feed links
	printf "<nav class=\"flinks\">"
	if (site_url) {
		#print_link(site_url, "", (yt ? "chan" : "site"), "site-url")
		printf "<a class=\"site-url\" href=\"%s\">%s</a>\n", site_url, (yt ? "chan" : "site")
	}
	if (feed_url) {
		#print_link(feed_url, "", "feed", "feed-url")
		printf "<a class=\"feed-url\" href=\"%s\">feed</a>\n", feed_url
	}
	# Top link
	#print_link("#", "back to top", "top", "top")
	printf "<a class=\"top\" href=\"#\">top</a>"
	printf "</nav>"
	printf "</header>\n"
	# Feed entries
	printf "<table class=\"entries\">\n"
	printf "%s", buf
	printf "</table>\n</section>\n"
	# Sidebar
	if (stale_count) {
		printf("<li class=\"%s\">", (fresh_count ? "fresh" : "stale")) >> ASIDE
		printf("<a href=\"#%s\"%s>%s</a></li>\n", name, (yt ? " class=\"yt\"" : ""), name) >> ASIDE
	}
	# Log
	printf("%s", stamp) > STDERR
}

function envor(var, def)
{
	return (ENVIRON[var] ? ENVIRON[var] : def)
}

function html_escape(t)
{
	gsub(/</, "\\&lt;", t)
	gsub(/>/, "\\&gt;", t)
	gsub(/&/, "\\&amp;", t)
	return t
}

function html_top()
{
	# Print HTML header
	fortunecmd = "fortune"
	fortunecmd | getline LOGO_TITLE
	close(fortunecmd)
	sub(/"/, "", LOGO_TITLE)
	datecmd = "date -u +'" DATEFMT "'"
	datecmd | getline UPDATE_TIME
	close(datecmd)
	print "<!DOCTYPE html>"
	print "<html>"
	# <head>
	print "<head>"
	print "<meta charset=\"utf-8\">"
	print "<meta http-equiv=\"X-UA-Compatible\" content=\"IE=edge\">"
	print "<meta http-equiv=\"Content-Type\" content=\"text/html;charset=UTF-8\">"
	print "<meta name=\"viewport\" content=\"width=device-width, initial-scale=1\">"
	print "<title>" TITLE "</title>"
	print "<link rel=\"stylesheet\" type=\"text/css\" href=\"style.css\">"
	print "<link rel=\"shortcut icon\" type=\"image/png\" href=\"mars-eyes.png\">"
	print "<link rel=\"alternate\" type=\"application/atom+xml\" title=\"rss (full)\" href=\"feeds.xml\">"
	print "<link rel=\"alternate\" type=\"application/atom+xml\" title=\"rss (short)\" href=\"feeds-short.xml\">"
	print "<link rel=\"alternate\" type=\"application/xml\" title=\"opml\" href=\"feeds.opml\">"
	print "<link rel=\"alternate\" type=\"text/plain\" title=\"twtxt\" href=\"feeds.txt\">"
	print "<link rel=\"alternate\" type=\"text/plain\" title=\"ass\" href=\"feed.ass\">"
	print "</head>"
	print "<body>"
	# <header>
	print "<header id=\"page-header\">"
	print "<h1>"
	printf "<a id=\"logo\" href=\"index.html\"><img src=\"mars-eyes.png\""
	printf " title=\"%s\" ", LOGO_TITLE
	printf "width=\"40\" height=\"39\" "
	print "alt=\"mars, but with eyes\" /></a>"
	print TITLE "</h1>"
	print "<p id=\"last-updated\">last updated at <time>" UPDATE_TIME "</time></p>"
	print "</header>"
	# <nav>
	print "<nav id=\"page-nav\">"
	print "<a href=\"#\" id=\"stalebutton\">hide stale</a>", SEPARATOR
	print "<a href=\"feeds-short.xml\">rss</a>", SEPARATOR
	print "<a href=\"feeds.html\">all</a>", "(<a href=\"feeds.xml\">rss</a>)", SEPARATOR
	print "<a href=\"feeds.txt\">twtxt</a>", SEPARATOR
	print "<a href=\"feed.ass\">ass</a>", SEPARATOR
	print "<a href=\"feeds.opml\">opml</a>"
	print "</nav>"
	# first part of <main>
	print "<main>"
	print "<section id=\"list\">"
}

function print_link(href, title, text, class)
{
	l = "<a href=\"" href "\""
	if (class) {
		l = l " class=\"" class "\""
	}
	l = l " title=\"" title "\""
	l = l " rel=\"nofollow\" target=\"_blank\">"
	l = l text
	l = l "</a>"
	bufprint(l, 0)
}

function silo_class(link)
{
	if (link ~ /youtube\.com/) {
		return " youtube"
	}
	if (link ~ /facebook\.com/) {
		return " facebook"
	}
	if (link ~ /twitter\.com/) {
		return " twitter"
	}
	if (link ~ /reddit\.com/) {
		return " reddit"
	}
	if (link ~ /meedium\.com/) {
		return "medium"
	}
	return ""
}

function silo_link(link)
{
	ret = link
	http = "https?://[^\\.]*\\.?"
	siloed = 1
	if (ret ~ (http "youtube\\.com")) {
		sub(http "youtube\\.com", YOUTUBE_ALT_URL, ret)
	} else if (ret ~ (http "i\\.reddit\\.com")) {
		sub(http "i\\.reddit\\.com", REDDIT_ALT_URL, ret)
	} else if (ret ~ (http "twitter\\.com")) {
		sub(http "twitter\\.com", TWITTER_ALT_URL, ret)
	} else if (ret ~ (http "medium\\.com")) {
		sub(http "medium\\.com", MEDIUM_ALT_URL, ret)
	} else {
		siloed = 0
		for (site in PAYWALLED) {
			if (link ~ (http PAYWALLED[site])) {
				ret = PAYWALL_LADDER link
				siloed = 1
				break
			}
		}
	}
	return ret
}

function unescape(t)
{
	# t = html_escape(t)
	gsub(/\\t/, "\t", t)
	gsub(/\\n/, "\n", t)
	gsub(/\\\\/, "\\", t)
	return t
}