#!/bin/sh

## wag is a tool to generate static web sites
## 
## Commands:

unset CDPATH
export LC_ALL=C IFS=' 	
'

cmd=$(command -v "$0")

dest=./public

footer() {
	:
}

# front parses front matter
front() {
	{
		read a b && [ "$a" = "---" ] && [ "$b" = "" ] || return
		while read a b; do
			[ "$a" = "---" ] && [ "$b" = "" ] && break
			eval "front_${a%:}=\"$b\""
		done
	} < "$1"
}

## gen generates site content from source files
gen() {
	[ -d "$dest" ] || mkdir -p "$dest"
	for f in *.md; do
		[ "$f" = "*.md" ] && continue
		g="$dest/${f%.md}.html"
		[ "$f" -ot "$g" ] && continue
		echo "f: $f $g"
		front "$f"
		{
			header
			md2html "$f"
			footer
		} > "$g"
	done
}

header() {
	cat <<- EOT
		<title>$front_title</title>
		<meta charset="utf-8">
		<style>body {max-width: 52em; margin: 1em auto; padding: 1em}</style>
	EOT
}

## help prints this program documentation
help() { awk '/^## / {print substr($0, 4)}' "$cmd"; }

lipsum='Lorem ipsum dolor sit amet, consectetur adipiscing elit,
sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.
Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris
nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in
reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla
pariatur. Excepteur sint occaecat cupidatat non proident, sunt in
culpa qui officia deserunt mollit anim id est laborum.'

# An overly simplified http request parser for static web sites.
http_request() {
	read -r cmd uri proto
	case $uri in */) uri="${uri}index.html" ;; esac
	while true; do
		read -r line || break
		[ ${#line} = 0 ] && break
	done
	printf 'HTTP/1.1 200 OK\n\n' && cat "${uri#/}"
}

md2h() { want="$2" got=$(echo "$1" | md2html -); }
 	
md2html() {
	tmp=$(mktemp -u)
	trap "rm -f '$tmp'" EXIT
	{
		cat "${1:--}" | tee "$tmp" | awk '/^[ 	]*\[[^]]+\]:/'
		cat "$tmp"
	} |
	awk '
	function newblock(nblock) {
		if (text)
			print "<" block ">" text "</" block ">"
		text = ""
		out = 1
		block = nblock ? nblock : "p"
	}

	function subinline(tgl, inl) {
		while (match($0, tgl)) {
			if (inline[ni] == inl)
				ni -= sub(tgl, "</" inl ">")
			else if (sub(tgl, "<" inl ">"))
				inline[++ni] = inl
		}
	}

	function dolink(href, lnk) {
		# Undo escaped html in uris
		gsub(/&amp;/, "\\&", href)
		gsub(/&lt;/, "<", href)
		gsub(/&gt;/, ">", href)
		# & can be tricky, and not standard:
		gsub(/&/, "\\\\\\&", href)
		gsub(/&/, "\\\\\\&", lnk)
		return "<a href=\"" href "\">" lnk "</a>"
	}

	BEGIN {
		ni = 0	 # inlines
		nl = 0   # nested lists
		out = 0  # 0 if no output so far
		text = ""
		block = "p"
	}

	# Skip front matter.
	out == 0 && $0 == "---" {
		do
			getline
		while ($0 != "---")
		next
	}

	# Escape HTML.
	esc != "false" {
		gsub("&", "\\&amp;")
		gsub("<", "\\&lt;")
		gsub(">", "\\&gt;")
	}
	
	# Internal references.
	match($0, /^[    ]*\[[^]]+\]:/) > 0 {
		k = substr($0, RSTART+1, RLENGTH-3)
		v = substr($0, RLENGTH+1)
		sub(/^[ 	]/, "", v)
		sub(/[ 	]$/, "", v)
		ref[substr($0, RSTART+1, RLENGTH-3)] = v
		next
	}

	# Horizontal rules.
	/^[ 	]*([-*_] ?)+[ 	]*$/ && text == "" {
		print "<hr>"
		next
	}

	# Tables. Syntax:
	# 		Right Align| 	Center Align	|Left Align
	/([ 	]\|)|(\|[ 	])/ {
		if (block != "table")
			newblock("table")
		nc = split($0, cells, "|")
		$0 = "<tr>\n"
		for (i = 1; i <= nc; i++) {
			align = "left"
			if (sub(/^[ 	]+/, "", cells[i])) {
				if (sub(/[ 	]+$/, "", cells[i]))
					align = "center"
				else
					align = "right"
			}
			sub(/[ 	]+$/, "", cells[i])
			$0 = $0 "<td align=\"" align "\">" cells[i] "</td>\n"
		}
		$0 = $0 "</tr>"
	}

	# Ordered and unordered (possibly nested) lists.
	/^[ 	]*([*+-]|(([0-9]+[.-]?)+))[ 	]/ {
		newblock("li")
		nnl = 1
		while (match($0, /^[ 	]/)) {
			sub(/^[ 	]/, "")
			nnl++
		}
		while (nl > nnl)
			print "</" list[nl--] ">"
		while (nl < nnl) {
			list[++nl] = "ol"
			if (match($0, /^[*+-]/))
				list[nl] = "ul"
			print "<" list[nl] ">"
		}
		sub(/^([*+-]|(([0-9]+[.-]?)+))[ 	]/, "")
	}

	# Multi line list items.
	block == "li" {
		sub(/^( *)|(	*)/, "")
	}

	# Code blocks.
	/^(    |	)/ {
		if (block != "code")
			newblock("code")
		sub(/^(    |	)/, "")
		text = text $0 "\n"
		next
	}

	# Paragraphs.
	/^$/ {
		newblock()
		while (nl > 0)
			print "</" list[nl--] ">"
	}

	# Headers.
	/^#+ / {
		newblock()
		match($0, /#+/)
		n = RLENGTH
		if (n > 6)
			n = 6
		text = substr($0, RLENGTH + 1)
		sub(/^ */, "", text)
		block = "h" n
		next
	}

	# Alternate headers (underlined).
	/^=+$/ {
		block = "h" 1
		next
	}

	/^-+$/ {
		block = "h" 2
		next
	}

	{
		# Images.
		while (match($0, /!\[[^]]+\]\([^)]+\)/)) {
			split(substr($0, RSTART, RLENGTH), a, /(!\[)|\)|(\]\()/)
			sub(/!\[[^]]+\]\([^)]+\)/, "<img src=\"" a[3] "\" alt=\"" a[2] "\">")
		}
		# Links.
		while (match($0, /\[[^]]+\]\([^)]+\)/)) {
			split(substr($0, RSTART, RLENGTH), a, /[[)]|(\]\()/)
			sub(/\[[^]]+\]\([^)]+\)/, dolink(a[3], a[2]))
		}
		# Internal references.
		while (match($0, /\[[^]]+\]/)) {
			k = substr($0, RSTART+1, RLENGTH-2)
			sub(/\[[^]]+\]/, dolink(ref[k], k))
		}
		# Auto links (uri matching is poor).
		na = split($0, a, /(^\()|[ 	]|([,.)]([ 	]|$))/)
		for (i = 1; i <= na; i++)
			if (match(a[i], /^(((https?|ftp|file|news|irc):\/\/)|(mailto:)).+$/))
				sub(a[i], dolink(a[i], a[i]))
		# Inline.
		subinline("(\\*\\*)|(__)", "strong")
		subinline("\\*", "em")
		subinline("`", "code")
		text = text (text ? " " : "") $0
	}

	END {
		while (ni > 0)
			text = text "</" inline[ni--] ">"
		newblock()
		while (nl > 0)
			print "</" list[nl--] ">"
	}'
}

#serve() { while true; do busybox nc -kl -p 1500 -e "$cmd http_request"; done; }
serve() { WAG_FUN=http_request busybox nc -kl -p 1500 -e "$cmd"; }

test() {
	fail=0 pass=0 skip=0 tfilter="$*"

	test_run md2h 'abc __def__ ghi' '<p>abc <strong>def</strong> ghi</p>'
	test_run md2h 'abc **def** ghi' '<p>abc <strong>def</strong> ghi</p>'
	test_run md2h 'abc *def* ghi' '<p>abc <em>def</em> ghi</p>'
	test_run md2h 'abc ***def*** ghi' '<p>abc <strong><em>def</strong></em> ghi</p>'
	test_run md2h 'abc `def` ghi' '<p>abc <code>def</code> ghi</p>'
	test_run md2h '# h1' '<h1>h1</h1>'
	test_run md2h '## h2' '<h2>h2</h2>'
	test_run md2h 'h1
==' '<h1>h1</h1>'
	test_run md2h 'h2
--' '<h2>h2</h2>'
	test_run md2h 'abc [github] def' '<p>abc <a href="">github</a> def</p>'
	test_run md2h 'abc [github](https://github.com) def' '<p>abc <a href="https://github.com">github</a> def</p>'
	test_run md2h 'abc [github] def

[github]: https://github.com' '<p>abc <a href="https://github.com">github</a> def</p>'
	test_run md2h '---
Title: front matter test
---

Hello [world].

---
Bye.

[world]: http://example.com' '<p>Hello <a href="http://example.com">world</a>.</p>
<hr>
<p>Bye.</p>'

	echo "Total: $((pass + fail + skip)), Passed: $pass, Failed: $fail, Skip: $skip"
	return "$fail"
}

test_run() {
	eval "test_$1=\$((test_$1 + 1)); ti=\"\$test_$1\""
	[ "$tfilter" ] &&
	case "$1#$ti" in
		$tfilter) ;;
		*) skip=$((skip + 1)); return 0 ;;
	esac
	"$@"
	[ "$got" = "$want" ] && {
		pass=$((pass + 1))
		return 0
	}
	fail=$((fail + 1))
	printf "%s FAIL\n  Got: %s\n Want: %s\n" "$1#$ti" "$got" "$want" >&2
	return 1
}

[ "$WAG_FUN" ] && { $WAG_FUN; exit; }

# Execute command line
[ "$1" ] || help && "$@"