//
// Black Friday Markdown Processor
// Originally based on http://github.com/tanoku/upskirt
// by Russ Ross <russ@russross.com>
//

//
//
// Markdown parsing and processing
//
//

package blackfriday

import (
	"bytes"
	"sort"
	"unicode"
)

// These are the supported markdown parsing extensions.
// OR these values together to select multiple extensions.
const (
	EXTENSION_NO_INTRA_EMPHASIS = 1 << iota
	EXTENSION_TABLES
	EXTENSION_FENCED_CODE
	EXTENSION_AUTOLINK
	EXTENSION_STRIKETHROUGH
	EXTENSION_LAX_HTML_BLOCKS
	EXTENSION_SPACE_HEADERS
)

// These are the possible flag values for the link renderer.
// Only a single one of these values will be used; they are not ORed together.
// These are mostly of interest if you are writing a new output format.
const (
	LINK_TYPE_NOT_AUTOLINK = iota
	LINK_TYPE_NORMAL
	LINK_TYPE_EMAIL
)

// These are the possible flag values for the listitem renderer.
// Multiple flag values may be ORed together.
// These are mostly of interest if you are writing a new output format.
const (
	LIST_TYPE_ORDERED = 1 << iota
	LIST_ITEM_CONTAINS_BLOCK
	LIST_ITEM_END_OF_LIST
)

// These are the possible flag values for the table cell renderer.
// Only a single one of these values will be used; they are not ORed together.
// These are mostly of interest if you are writing a new output format.
const (
	TABLE_ALIGNMENT_LEFT = 1 << iota
	TABLE_ALIGNMENT_RIGHT
	TABLE_ALIGNMENT_CENTER = (TABLE_ALIGNMENT_LEFT | TABLE_ALIGNMENT_RIGHT)
)

// The size of a tab stop.
const TAB_SIZE = 4

// These are the tags that are recognized as HTML block tags.
// Any of these can be included in markdown text without special escaping.
var block_tags = map[string]bool{
	"p":          true,
	"dl":         true,
	"h1":         true,
	"h2":         true,
	"h3":         true,
	"h4":         true,
	"h5":         true,
	"h6":         true,
	"ol":         true,
	"ul":         true,
	"del":        true,
	"div":        true,
	"ins":        true,
	"pre":        true,
	"form":       true,
	"math":       true,
	"table":      true,
	"iframe":     true,
	"script":     true,
	"fieldset":   true,
	"noscript":   true,
	"blockquote": true,
}

// This struct defines the rendering interface.
// A series of callback functions are registered to form a complete renderer.
// A single interface{} value field is provided, and that value is handed to
// each callback. Leaving a field blank suppresses rendering that type of output
// except where noted.
//
// This is mostly of interest if you are implementing a new rendering format.
// Most users will use the convenience functions to fill in this structure.
type Renderer struct {
	// block-level callbacks---nil skips the block
	blockcode  func(ob *bytes.Buffer, text []byte, lang string, opaque interface{})
	blockquote func(ob *bytes.Buffer, text []byte, opaque interface{})
	blockhtml  func(ob *bytes.Buffer, text []byte, opaque interface{})
	header     func(ob *bytes.Buffer, text []byte, level int, opaque interface{})
	hrule      func(ob *bytes.Buffer, opaque interface{})
	list       func(ob *bytes.Buffer, text []byte, flags int, opaque interface{})
	listitem   func(ob *bytes.Buffer, text []byte, flags int, opaque interface{})
	paragraph  func(ob *bytes.Buffer, text []byte, opaque interface{})
	table      func(ob *bytes.Buffer, header []byte, body []byte, opaque interface{})
	table_row  func(ob *bytes.Buffer, text []byte, opaque interface{})
	table_cell func(ob *bytes.Buffer, text []byte, flags int, opaque interface{})

	// span-level callbacks---nil or return 0 prints the span verbatim
	autolink        func(ob *bytes.Buffer, link []byte, kind int, opaque interface{}) int
	codespan        func(ob *bytes.Buffer, text []byte, opaque interface{}) int
	double_emphasis func(ob *bytes.Buffer, text []byte, opaque interface{}) int
	emphasis        func(ob *bytes.Buffer, text []byte, opaque interface{}) int
	image           func(ob *bytes.Buffer, link []byte, title []byte, alt []byte, opaque interface{}) int
	linebreak       func(ob *bytes.Buffer, opaque interface{}) int
	link            func(ob *bytes.Buffer, link []byte, title []byte, content []byte, opaque interface{}) int
	raw_html_tag    func(ob *bytes.Buffer, tag []byte, opaque interface{}) int
	triple_emphasis func(ob *bytes.Buffer, text []byte, opaque interface{}) int
	strikethrough   func(ob *bytes.Buffer, text []byte, opaque interface{}) int

	// low-level callbacks---nil copies input directly into the output
	entity      func(ob *bytes.Buffer, entity []byte, opaque interface{})
	normal_text func(ob *bytes.Buffer, text []byte, opaque interface{})

	// header and footer
	doc_header func(ob *bytes.Buffer, opaque interface{})
	doc_footer func(ob *bytes.Buffer, opaque interface{})

	// user data---passed back to every callback
	opaque interface{}
}

//

type render struct {
	mk          *Renderer
	refs        link_ref_array
	active_char [256]int
	ext_flags   uint32
	nesting     int
	max_nesting int
}

const (
	MD_CHAR_NONE = iota
	MD_CHAR_EMPHASIS
	MD_CHAR_CODESPAN
	MD_CHAR_LINEBREAK
	MD_CHAR_LINK
	MD_CHAR_LANGLE
	MD_CHAR_ESCAPE
	MD_CHAR_ENTITITY
	MD_CHAR_AUTOLINK
)


//
//
// Public interface
//
//

// Parse and render a block of markdown-encoded text.
// The renderer is used to format the output, and extensions dictates which
// non-standard extensions are enabled.
func Markdown(input []byte, renderer *Renderer, extensions uint32) []byte {
	// no point in parsing if we can't render
	if renderer == nil {
		return nil
	}

	// fill in the character-level parsers
	markdown_char_ptrs[MD_CHAR_NONE] = nil
	markdown_char_ptrs[MD_CHAR_EMPHASIS] = char_emphasis
	markdown_char_ptrs[MD_CHAR_CODESPAN] = char_codespan
	markdown_char_ptrs[MD_CHAR_LINEBREAK] = char_linebreak
	markdown_char_ptrs[MD_CHAR_LINK] = char_link
	markdown_char_ptrs[MD_CHAR_LANGLE] = char_langle_tag
	markdown_char_ptrs[MD_CHAR_ESCAPE] = char_escape
	markdown_char_ptrs[MD_CHAR_ENTITITY] = char_entity
	markdown_char_ptrs[MD_CHAR_AUTOLINK] = char_autolink

	// fill in the render structure
	rndr := new(render)
	rndr.mk = renderer
	rndr.ext_flags = extensions
	rndr.max_nesting = 16

	if rndr.mk.emphasis != nil || rndr.mk.double_emphasis != nil || rndr.mk.triple_emphasis != nil {
		rndr.active_char['*'] = MD_CHAR_EMPHASIS
		rndr.active_char['_'] = MD_CHAR_EMPHASIS
		if extensions&EXTENSION_STRIKETHROUGH != 0 {
			rndr.active_char['~'] = MD_CHAR_EMPHASIS
		}
	}
	if rndr.mk.codespan != nil {
		rndr.active_char['`'] = MD_CHAR_CODESPAN
	}
	if rndr.mk.linebreak != nil {
		rndr.active_char['\n'] = MD_CHAR_LINEBREAK
	}
	if rndr.mk.image != nil || rndr.mk.link != nil {
		rndr.active_char['['] = MD_CHAR_LINK
	}
	rndr.active_char['<'] = MD_CHAR_LANGLE
	rndr.active_char['\\'] = MD_CHAR_ESCAPE
	rndr.active_char['&'] = MD_CHAR_ENTITITY

	if extensions&EXTENSION_AUTOLINK != 0 {
		rndr.active_char['h'] = MD_CHAR_AUTOLINK // http, https
		rndr.active_char['H'] = MD_CHAR_AUTOLINK

		rndr.active_char['f'] = MD_CHAR_AUTOLINK // ftp
		rndr.active_char['F'] = MD_CHAR_AUTOLINK

		rndr.active_char['m'] = MD_CHAR_AUTOLINK // mailto
		rndr.active_char['M'] = MD_CHAR_AUTOLINK
	}

	// first pass: look for references, copy everything else
	text := bytes.NewBuffer(nil)
	beg, end := 0, 0
	for beg < len(input) { // iterate over lines
		if end = is_ref(rndr, input[beg:]); end > 0 {
			beg += end
		} else { // skip to the next line
			end = beg
			for end < len(input) && input[end] != '\n' && input[end] != '\r' {
				end++
			}

			// add the line body if present
			if end > beg {
				expand_tabs(text, input[beg:end])
			}

			for end < len(input) && (input[end] == '\n' || input[end] == '\r') {
				// add one \n per newline
				if input[end] == '\n' || (end+1 < len(input) && input[end+1] != '\n') {
					text.WriteByte('\n')
				}
				end++
			}

			beg = end
		}
	}

	// sort the reference array
	if len(rndr.refs) > 1 {
		sort.Sort(rndr.refs)
	}

	// second pass: actual rendering
    output := bytes.NewBuffer(nil)
	if rndr.mk.doc_header != nil {
		rndr.mk.doc_header(output, rndr.mk.opaque)
	}

	if text.Len() > 0 {
		// add a final newline if not already present
		finalchar := text.Bytes()[text.Len()-1]
		if finalchar != '\n' && finalchar != '\r' {
			text.WriteByte('\n')
		}
		parse_block(output, rndr, text.Bytes())
	}

	if rndr.mk.doc_footer != nil {
		rndr.mk.doc_footer(output, rndr.mk.opaque)
	}

	if rndr.nesting != 0 {
		panic("Nesting level did not end at zero")
	}

    return output.Bytes()
}


//
// Inline parsing
// Functions to parse text within a block.
//

// closures to render active chars, each:
//   returns the number of chars taken care of
//   data is the complete block being rendered
//   offset is the number of valid chars before the data
//
// Note: this is filled in in Markdown to prevent an initilization loop
var markdown_char_ptrs [9]func(ob *bytes.Buffer, rndr *render, data []byte, offset int) int

func parse_inline(ob *bytes.Buffer, rndr *render, data []byte) {
	if rndr.nesting >= rndr.max_nesting {
		return
	}
	rndr.nesting++

	i, end := 0, 0
	for i < len(data) {
		// copy inactive chars into the output
		for end < len(data) && rndr.active_char[data[end]] == 0 {
			end++
		}

		if rndr.mk.normal_text != nil {
			rndr.mk.normal_text(ob, data[i:end], rndr.mk.opaque)
		} else {
			ob.Write(data[i:end])
		}

		if end >= len(data) {
			break
		}
		i = end

		// call the trigger
		action := rndr.active_char[data[end]]
		end = markdown_char_ptrs[action](ob, rndr, data, i)

		if end == 0 { // no action from the callback
			end = i + 1
		} else {
			i += end
			end = i
		}
	}

	rndr.nesting--
}

// single and double emphasis parsing
func char_emphasis(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	data = data[offset:]
	c := data[0]
	ret := 0

	if len(data) > 2 && data[1] != c {
		// whitespace cannot follow an opening emphasis;
		// strikethrough only takes two characters '~~'
		if c == '~' || isspace(data[1]) {
			return 0
		}
		if ret = parse_emph1(ob, rndr, data[1:], c); ret == 0 {
			return 0
		}

		return ret + 1
	}

	if len(data) > 3 && data[1] == c && data[2] != c {
		if isspace(data[2]) {
			return 0
		}
		if ret = parse_emph2(ob, rndr, data[2:], c); ret == 0 {
			return 0
		}

		return ret + 2
	}

	if len(data) > 4 && data[1] == c && data[2] == c && data[3] != c {
		if c == '~' || isspace(data[3]) {
			return 0
		}
		if ret = parse_emph3(ob, rndr, data, 3, c); ret == 0 {
			return 0
		}

		return ret + 3
	}

	return 0
}

func char_codespan(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	data = data[offset:]

	nb := 0

	// count the number of backticks in the delimiter
	for nb < len(data) && data[nb] == '`' {
		nb++
	}

	// find the next delimiter
	i, end := 0, 0
	for end = nb; end < len(data) && i < nb; end++ {
		if data[end] == '`' {
			i++
		} else {
			i = 0
		}
	}

	if i < nb && end >= len(data) {
		return 0 // no matching delimiter
	}

	// trim outside whitespace
	f_begin := nb
	for f_begin < end && (data[f_begin] == ' ' || data[f_begin] == '\t') {
		f_begin++
	}

	f_end := end - nb
	for f_end > nb && (data[f_end-1] == ' ' || data[f_end-1] == '\t') {
		f_end--
	}

	// real code span
	if rndr.mk.codespan == nil {
		return 0
	}
	if f_begin < f_end {
		if rndr.mk.codespan(ob, data[f_begin:f_end], rndr.mk.opaque) == 0 {
			end = 0
		}
	} else {
		if rndr.mk.codespan(ob, nil, rndr.mk.opaque) == 0 {
			end = 0
		}
	}

	return end

}

// '\n' preceded by two spaces
func char_linebreak(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	if offset < 2 || data[offset-1] != ' ' || data[offset-2] != ' ' {
		return 0
	}

	// remove trailing spaces from ob and render
	ob_bytes := ob.Bytes()
	end := len(ob_bytes)
	for end > 0 && ob_bytes[end-1] == ' ' {
		end--
	}
	ob.Truncate(end)

	if rndr.mk.linebreak == nil {
		return 0
	}
	if rndr.mk.linebreak(ob, rndr.mk.opaque) > 0 {
		return 1
	} else {
		return 0
	}

	return 0
}

// '[': parse a link or an image
func char_link(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	is_img := offset > 0 && data[offset-1] == '!'

	data = data[offset:]

	i := 1
	var title, link []byte
	text_has_nl := false

	// check whether the correct renderer exists
	if (is_img && rndr.mk.image == nil) || (!is_img && rndr.mk.link == nil) {
		return 0
	}

	// look for the matching closing bracket
	for level := 1; level > 0 && i < len(data); i++ {
		switch {
		case data[i] == '\n':
			text_has_nl = true

		case data[i-1] == '\\':
			continue

		case data[i] == '[':
			level++

		case data[i] == ']':
			level--
			if level <= 0 {
				i-- // compensate for extra i++ in for loop
			}
		}
	}

	if i >= len(data) {
		return 0
	}

	txt_e := i
	i++

	// skip any amount of whitespace or newline
	// (this is much more lax than original markdown syntax)
	for i < len(data) && isspace(data[i]) {
		i++
	}

	// inline style link
	switch {
	case i < len(data) && data[i] == '(':
		// skip initial whitespace
		i++

		for i < len(data) && isspace(data[i]) {
			i++
		}

		link_b := i

		// look for link end: ' " )
		for i < len(data) {
			if data[i] == '\\' {
				i += 2
			} else {
				if data[i] == ')' || data[i] == '\'' || data[i] == '"' {
					break
				}
				i++
			}
		}

		if i >= len(data) {
			return 0
		}
		link_e := i

		// look for title end if present
		title_b, title_e := 0, 0
		if data[i] == '\'' || data[i] == '"' {
			i++
			title_b = i

			for i < len(data) {
				if data[i] == '\\' {
					i += 2
				} else {
					if data[i] == ')' {
						break
					}
					i++
				}
			}

			if i >= len(data) {
				return 0
			}

			// skip whitespace after title
			title_e = i - 1
			for title_e > title_b && isspace(data[title_e]) {
				title_e--
			}

			// check for closing quote presence
			if data[title_e] != '\'' && data[title_e] != '"' {
				title_b, title_e = 0, 0
				link_e = i
			}
		}

		// remove whitespace at the end of the link
		for link_e > link_b && isspace(data[link_e-1]) {
			link_e--
		}

		// remove optional angle brackets around the link
		if data[link_b] == '<' {
			link_b++
		}
		if data[link_e-1] == '>' {
			link_e--
		}

		// build escaped link and title
		if link_e > link_b {
			link = data[link_b:link_e]
		}

		if title_e > title_b {
			title = data[title_b:title_e]
		}

		i++

	// reference style link
	case i < len(data) && data[i] == '[':
		var id []byte

		// look for the id
		i++
		link_b := i
		for i < len(data) && data[i] != ']' {
			i++
		}
		if i >= len(data) {
			return 0
		}
		link_e := i

		// find the link_ref
		if link_b == link_e {
			if text_has_nl {
				b := bytes.NewBuffer(nil)

				for j := 1; j < txt_e; j++ {
					switch {
					case data[j] != '\n':
						b.WriteByte(data[j])
					case data[j-1] != ' ':
						b.WriteByte(' ')
					}
				}

				id = b.Bytes()
			} else {
				id = data[1:txt_e]
			}
		} else {
			id = data[link_b:link_e]
		}

		// find the link_ref with matching id
		index := sortDotSearch(len(rndr.refs), func(i int) bool {
			return !byteslice_less(rndr.refs[i].id, id)
		})
		if index >= len(rndr.refs) || !bytes.Equal(rndr.refs[index].id, id) {
			return 0
		}
		lr := rndr.refs[index]

		// keep link and title from link_ref
		link = lr.link
		title = lr.title
		i++

	// shortcut reference style link
	default:
		var id []byte

		// craft the id
		if text_has_nl {
			b := bytes.NewBuffer(nil)

			for j := 1; j < txt_e; j++ {
				switch {
				case data[j] != '\n':
					b.WriteByte(data[j])
				case data[j-1] != ' ':
					b.WriteByte(' ')
				}
			}

			id = b.Bytes()
		} else {
			id = data[1:txt_e]
		}

		// find the link_ref with matching id
		index := sortDotSearch(len(rndr.refs), func(i int) bool {
			return !byteslice_less(rndr.refs[i].id, id)
		})
		if index >= len(rndr.refs) || !bytes.Equal(rndr.refs[index].id, id) {
			return 0
		}
		lr := rndr.refs[index]

		// keep link and title from link_ref
		link = lr.link
		title = lr.title

		// rewind the whitespace
		i = txt_e + 1
	}

	// build content: img alt is escaped, link content is parsed
	content := bytes.NewBuffer(nil)
	if txt_e > 1 {
		if is_img {
			content.Write(data[1:txt_e])
		} else {
			parse_inline(content, rndr, data[1:txt_e])
		}
	}

	var u_link []byte
	if len(link) > 0 {
		u_link_buf := bytes.NewBuffer(nil)
		unescape_text(u_link_buf, link)
		u_link = u_link_buf.Bytes()
	}

	// call the relevant rendering function
	ret := 0
	if is_img {
		ob_size := ob.Len()
		ob_bytes := ob.Bytes()
		if ob_size > 0 && ob_bytes[ob_size-1] == '!' {
			ob.Truncate(ob_size - 1)
		}

		ret = rndr.mk.image(ob, u_link, title, content.Bytes(), rndr.mk.opaque)
	} else {
		ret = rndr.mk.link(ob, u_link, title, content.Bytes(), rndr.mk.opaque)
	}

	if ret > 0 {
		return i
	}
	return 0
}

// '<' when tags or autolinks are allowed
func char_langle_tag(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	data = data[offset:]
	altype := LINK_TYPE_NOT_AUTOLINK
	end := tag_length(data, &altype)
	ret := 0

	if end > 2 {
		switch {
		case rndr.mk.autolink != nil && altype != LINK_TYPE_NOT_AUTOLINK:
			u_link := bytes.NewBuffer(nil)
			unescape_text(u_link, data[1:end+1-2])
			ret = rndr.mk.autolink(ob, u_link.Bytes(), altype, rndr.mk.opaque)
		case rndr.mk.raw_html_tag != nil:
			ret = rndr.mk.raw_html_tag(ob, data[:end], rndr.mk.opaque)
		}
	}

	if ret == 0 {
		return 0
	}
	return end
}

// '\\' backslash escape
var escape_chars = []byte("\\`*_{}[]()#+-.!:|&<>")

func char_escape(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	data = data[offset:]

	if len(data) > 1 {
		if bytes.IndexByte(escape_chars, data[1]) < 0 {
			return 0
		}

		if rndr.mk.normal_text != nil {
			rndr.mk.normal_text(ob, data[1:2], rndr.mk.opaque)
		} else {
			ob.WriteByte(data[1])
		}
	}

	return 2
}

// '&' escaped when it doesn't belong to an entity
// valid entities are assumed to be anything matching &#?[A-Za-z0-9]+;
func char_entity(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	data = data[offset:]

	end := 1

	if end < len(data) && data[end] == '#' {
		end++
	}

	for end < len(data) && isalnum(data[end]) {
		end++
	}

	if end < len(data) && data[end] == ';' {
		end++ // real entity
	} else {
		return 0 // lone '&'
	}

	if rndr.mk.entity != nil {
		rndr.mk.entity(ob, data[:end], rndr.mk.opaque)
	} else {
		ob.Write(data[:end])
	}

	return end
}

func char_autolink(ob *bytes.Buffer, rndr *render, data []byte, offset int) int {
	orig_data := data
	data = data[offset:]

	if offset > 0 {
		if !isspace(orig_data[offset-1]) && !ispunct(orig_data[offset-1]) {
			return 0
		}
	}

	if !is_safe_link(data) {
		return 0
	}

	link_end := 0
	for link_end < len(data) && !isspace(data[link_end]) {
		link_end++
	}

	// Skip punctuation at the end of the link
	if (data[link_end-1] == '.' || data[link_end-1] == ',' || data[link_end-1] == ';') && data[link_end-2] != '\\' {
		link_end--
	}

	// See if the link finishes with a punctuation sign that can be closed.
	var copen byte
	switch data[link_end-1] {
	case '"':
		copen = '"'
	case '\'':
		copen = '\''
	case ')':
		copen = '('
	case ']':
		copen = '['
	case '}':
		copen = '{'
	default:
		copen = 0
	}

	if copen != 0 {
		buf_end := offset + link_end - 2

		open_delim := 1

		/* Try to close the final punctuation sign in this same line;
		 * if we managed to close it outside of the URL, that means that it's
		 * not part of the URL. If it closes inside the URL, that means it
		 * is part of the URL.
		 *
		 * Examples:
		 *
		 *      foo http://www.pokemon.com/Pikachu_(Electric) bar
		 *              => http://www.pokemon.com/Pikachu_(Electric)
		 *
		 *      foo (http://www.pokemon.com/Pikachu_(Electric)) bar
		 *              => http://www.pokemon.com/Pikachu_(Electric)
		 *
		 *      foo http://www.pokemon.com/Pikachu_(Electric)) bar
		 *              => http://www.pokemon.com/Pikachu_(Electric))
		 *
		 *      (foo http://www.pokemon.com/Pikachu_(Electric)) bar
		 *              => foo http://www.pokemon.com/Pikachu_(Electric)
		 */

		for buf_end >= 0 && orig_data[buf_end] != '\n' && open_delim != 0 {
			if orig_data[buf_end] == data[link_end-1] {
				open_delim++
			}

			if orig_data[buf_end] == copen {
				open_delim--
			}

			buf_end--
		}

		if open_delim == 0 {
			link_end--
		}
	}

	if rndr.mk.autolink != nil {
		u_link := bytes.NewBuffer(nil)
		unescape_text(u_link, data[:link_end])

		rndr.mk.autolink(ob, u_link.Bytes(), LINK_TYPE_NORMAL, rndr.mk.opaque)
	}

	return link_end
}

var valid_uris = [][]byte{[]byte("http://"), []byte("https://"), []byte("ftp://"), []byte("mailto://")}

func is_safe_link(link []byte) bool {
	for _, prefix := range valid_uris {
		if len(link) > len(prefix) && !byteslice_less(link[:len(prefix)], prefix) && !byteslice_less(prefix, link[:len(prefix)]) && isalnum(link[len(prefix)]) {
			return true
		}
	}

	return false
}

// return the length of the given tag, or 0 is it's not valid
func tag_length(data []byte, autolink *int) int {
	var i, j int

	// a valid tag can't be shorter than 3 chars
	if len(data) < 3 {
		return 0
	}

	// begins with a '<' optionally followed by '/', followed by letter or number
	if data[0] != '<' {
		return 0
	}
	if data[1] == '/' {
		i = 2
	} else {
		i = 1
	}

	if !isalnum(data[i]) {
		return 0
	}

	// scheme test
	*autolink = LINK_TYPE_NOT_AUTOLINK

	// try to find the beggining of an URI
	for i < len(data) && (isalnum(data[i]) || data[i] == '.' || data[i] == '+' || data[i] == '-') {
		i++
	}

	if i > 1 && data[i] == '@' {
		if j = is_mail_autolink(data[i:]); j != 0 {
			*autolink = LINK_TYPE_EMAIL
			return i + j
		}
	}

	if i > 2 && data[i] == ':' {
		*autolink = LINK_TYPE_NORMAL
		i++
	}

	// complete autolink test: no whitespace or ' or "
	switch {
	case i >= len(data):
		*autolink = LINK_TYPE_NOT_AUTOLINK
	case *autolink != 0:
		j = i

		for i < len(data) {
			if data[i] == '\\' {
				i += 2
			} else {
				if data[i] == '>' || data[i] == '\'' || data[i] == '"' || isspace(data[i]) {
					break
				} else {
					i++
				}
			}

		}

		if i >= len(data) {
			return 0
		}
		if i > j && data[i] == '>' {
			return i + 1
		}

		// one of the forbidden chars has been found
		*autolink = LINK_TYPE_NOT_AUTOLINK
	}

	// look for something looking like a tag end
	for i < len(data) && data[i] != '>' {
		i++
	}
	if i >= len(data) {
		return 0
	}
	return i + 1
}

// look for the address part of a mail autolink and '>'
// this is less strict than the original markdown e-mail address matching
func is_mail_autolink(data []byte) int {
	nb := 0

	// address is assumed to be: [-@._a-zA-Z0-9]+ with exactly one '@'
	for i := 0; i < len(data); i++ {
		if isalnum(data[i]) {
			continue
		}

		switch data[i] {
		case '@':
			nb++

		case '-', '.', '_':
			break

		case '>':
			if nb == 1 {
				return i + 1
			} else {
				return 0
			}
		default:
			return 0
		}
	}

	return 0
}

// look for the next emph char, skipping other constructs
func find_emph_char(data []byte, c byte) int {
	i := 1

	for i < len(data) {
		for i < len(data) && data[i] != c && data[i] != '`' && data[i] != '[' {
			i++
		}
		if i >= len(data) {
			return 0
		}
		if data[i] == c {
			return i
		}

		// do not count escaped chars
		if i != 0 && data[i-1] == '\\' {
			i++
			continue
		}

		if data[i] == '`' {
			// skip a code span
			tmp_i := 0
			i++
			for i < len(data) && data[i] != '`' {
				if tmp_i == 0 && data[i] == c {
					tmp_i = i
				}
				i++
			}
			if i >= len(data) {
				return tmp_i
			}
			i++
		} else {
			if data[i] == '[' {
				// skip a link
				tmp_i := 0
				i++
				for i < len(data) && data[i] != ']' {
					if tmp_i == 0 && data[i] == c {
						tmp_i = i
					}
					i++
				}
				i++
				for i < len(data) && (data[i] == ' ' || data[i] == '\t' || data[i] == '\n') {
					i++
				}
				if i >= len(data) {
					return tmp_i
				}
				if data[i] != '[' && data[i] != '(' { // not a link
					if tmp_i > 0 {
						return tmp_i
					} else {
						continue
					}
				}
				cc := data[i]
				i++
				for i < len(data) && data[i] != cc {
					if tmp_i == 0 && data[i] == c {
						tmp_i = i
					}
					i++
				}
				if i >= len(data) {
					return tmp_i
				}
				i++
			}
		}
	}
	return 0
}

func parse_emph1(ob *bytes.Buffer, rndr *render, data []byte, c byte) int {
	i := 0

	if rndr.mk.emphasis == nil {
		return 0
	}

	// skip one symbol if coming from emph3
	if len(data) > 1 && data[0] == c && data[1] == c {
		i = 1
	}

	for i < len(data) {
		length := find_emph_char(data[i:], c)
		if length == 0 {
			return 0
		}
		i += length
		if i >= len(data) {
			return 0
		}

		if i+1 < len(data) && data[i+1] == c {
			i++
			continue
		}

		if data[i] == c && !isspace(data[i-1]) {

			if rndr.ext_flags&EXTENSION_NO_INTRA_EMPHASIS != 0 {
				if !(i+1 == len(data) || isspace(data[i+1]) || ispunct(data[i+1])) {
					continue
				}
			}

			work := bytes.NewBuffer(nil)
			parse_inline(work, rndr, data[:i])
			r := rndr.mk.emphasis(ob, work.Bytes(), rndr.mk.opaque)
			if r > 0 {
				return i + 1
			} else {
				return 0
			}
		}
	}

	return 0
}

func parse_emph2(ob *bytes.Buffer, rndr *render, data []byte, c byte) int {
	render_method := rndr.mk.double_emphasis
	if c == '~' {
		render_method = rndr.mk.strikethrough
	}

	if render_method == nil {
		return 0
	}

	i := 0

	for i < len(data) {
		length := find_emph_char(data[i:], c)
		if length == 0 {
			return 0
		}
		i += length

		if i+1 < len(data) && data[i] == c && data[i+1] == c && i > 0 && !isspace(data[i-1]) {
			work := bytes.NewBuffer(nil)
			parse_inline(work, rndr, data[:i])
			r := render_method(ob, work.Bytes(), rndr.mk.opaque)
			if r > 0 {
				return i + 2
			} else {
				return 0
			}
		}
		i++
	}
	return 0
}

func parse_emph3(ob *bytes.Buffer, rndr *render, data []byte, offset int, c byte) int {
	i := 0
	orig_data := data
	data = data[offset:]

	for i < len(data) {
		length := find_emph_char(data[i:], c)
		if length == 0 {
			return 0
		}
		i += length

		// skip whitespace preceded symbols
		if data[i] != c || isspace(data[i-1]) {
			continue
		}

		switch {
		case (i+2 < len(data) && data[i+1] == c && data[i+2] == c && rndr.mk.triple_emphasis != nil):
			// triple symbol found
			work := bytes.NewBuffer(nil)

			parse_inline(work, rndr, data[:i])
			r := rndr.mk.triple_emphasis(ob, work.Bytes(), rndr.mk.opaque)
			if r > 0 {
				return i + 3
			} else {
				return 0
			}
		case (i+1 < len(data) && data[i+1] == c):
			// double symbol found, hand over to emph1
			length = parse_emph1(ob, rndr, orig_data[offset-2:], c)
			if length == 0 {
				return 0
			} else {
				return length - 2
			}
		default:
			// single symbol found, hand over to emph2
			length = parse_emph2(ob, rndr, orig_data[offset-1:], c)
			if length == 0 {
				return 0
			} else {
				return length - 1
			}
		}
	}
	return 0
}


//
// Block parsing
// Functions to parse block-level elements.
//

// parse block-level data
func parse_block(ob *bytes.Buffer, rndr *render, data []byte) {
	if rndr.nesting >= rndr.max_nesting {
		return
	}
	rndr.nesting++

	for len(data) > 0 {
		if is_atxheader(rndr, data) {
			data = data[parse_atxheader(ob, rndr, data):]
			continue
		}
		if data[0] == '<' && rndr.mk.blockhtml != nil {
			if i := parse_htmlblock(ob, rndr, data, true); i > 0 {
				data = data[i:]
				continue
			}
		}
		if i := is_empty(data); i > 0 {
			data = data[i:]
			continue
		}
		if is_hrule(data) {
			if rndr.mk.hrule != nil {
				rndr.mk.hrule(ob, rndr.mk.opaque)
			}
			var i int
			for i = 0; i < len(data) && data[i] != '\n'; i++ {
			}
			data = data[i:]
			continue
		}
		if rndr.ext_flags&EXTENSION_FENCED_CODE != 0 {
			if i := parse_fencedcode(ob, rndr, data); i > 0 {
				data = data[i:]
				continue
			}
		}
		if rndr.ext_flags&EXTENSION_TABLES != 0 {
			if i := parse_table(ob, rndr, data); i > 0 {
				data = data[i:]
				continue
			}
		}
		if prefix_quote(data) > 0 {
			data = data[parse_blockquote(ob, rndr, data):]
			continue
		}
		if prefix_code(data) > 0 {
			data = data[parse_blockcode(ob, rndr, data):]
			continue
		}
		if prefix_uli(data) > 0 {
			data = data[parse_list(ob, rndr, data, 0):]
			continue
		}
		if prefix_oli(data) > 0 {
			data = data[parse_list(ob, rndr, data, LIST_TYPE_ORDERED):]
			continue
		}

		data = data[parse_paragraph(ob, rndr, data):]
	}

	rndr.nesting--
}

func is_atxheader(rndr *render, data []byte) bool {
	if data[0] != '#' {
		return false
	}

	if rndr.ext_flags&EXTENSION_SPACE_HEADERS != 0 {
		level := 0
		for level < len(data) && level < 6 && data[level] == '#' {
			level++
		}
		if level < len(data) && data[level] != ' ' && data[level] != '\t' {
			return false
		}
	}
	return true
}

func parse_atxheader(ob *bytes.Buffer, rndr *render, data []byte) int {
	level := 0
	for level < len(data) && level < 6 && data[level] == '#' {
		level++
	}
	i, end := 0, 0
	for i = level; i < len(data) && (data[i] == ' ' || data[i] == '\t'); i++ {
	}
	for end = i; end < len(data) && data[end] != '\n'; end++ {
	}
	skip := end
	for end > 0 && data[end-1] == '#' {
		end--
	}
	for end > 0 && (data[end-1] == ' ' || data[end-1] == '\t') {
		end--
	}
	if end > i {
		work := bytes.NewBuffer(nil)
		parse_inline(work, rndr, data[i:end])
		if rndr.mk.header != nil {
			rndr.mk.header(ob, work.Bytes(), level, rndr.mk.opaque)
		}
	}
	return skip
}

func is_headerline(data []byte) int {
	i := 0

	// test of level 1 header
	if data[i] == '=' {
		for i = 1; i < len(data) && data[i] == '='; i++ {
		}
		for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
			i++
		}
		if i >= len(data) || data[i] == '\n' {
			return 1
		} else {
			return 0
		}
	}

	// test of level 2 header
	if data[i] == '-' {
		for i = 1; i < len(data) && data[i] == '-'; i++ {
		}
		for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
			i++
		}
		if i >= len(data) || data[i] == '\n' {
			return 2
		} else {
			return 0
		}
	}

	return 0
}

func parse_htmlblock(ob *bytes.Buffer, rndr *render, data []byte, do_render bool) int {
	var i, j int

	// identify the opening tag
	if len(data) < 2 || data[0] != '<' {
		return 0
	}
	curtag, tagfound := find_block_tag(data[1:])

	// handle special cases
	if !tagfound {

		// HTML comment, laxist form
		if len(data) > 5 && data[1] == '!' && data[2] == '-' && data[3] == '-' {
			i = 5

			for i < len(data) && !(data[i-2] == '-' && data[i-1] == '-' && data[i] == '>') {
				i++
			}
			i++

			if i < len(data) {
				j = is_empty(data[i:])
			}

			if j > 0 {
				size := i + j
				if do_render && rndr.mk.blockhtml != nil {
					rndr.mk.blockhtml(ob, data[:size], rndr.mk.opaque)
				}
				return size
			}
		}

		// HR, which is the only self-closing block tag considered
		if len(data) > 4 && (data[1] == 'h' || data[1] == 'H') && (data[2] == 'r' || data[2] == 'R') {
			i = 3
			for i < len(data) && data[i] != '>' {
				i++
			}

			if i+1 < len(data) {
				i++
				j = is_empty(data[i:])
				if j > 0 {
					size := i + j
					if do_render && rndr.mk.blockhtml != nil {
						rndr.mk.blockhtml(ob, data[:size], rndr.mk.opaque)
					}
					return size
				}
			}
		}

		// no special case recognized
		return 0
	}

	// look for an unindented matching closing tag
	//      followed by a blank line
	i = 1
	found := false

	// if not found, try a second pass looking for indented match
	// but not if tag is "ins" or "del" (following original Markdown.pl)
	if curtag != "ins" && curtag != "del" {
		i = 1
		for i < len(data) {
			i++
			for i < len(data) && !(data[i-1] == '<' && data[i] == '/') {
				i++
			}

			if i+2+len(curtag) >= len(data) {
				break
			}

			j = htmlblock_end(curtag, rndr, data[i-1:])

			if j > 0 {
				i += j - 1
				found = true
				break
			}
		}
	}

	if !found {
		return 0
	}

	// the end of the block has been found
	if do_render && rndr.mk.blockhtml != nil {
		rndr.mk.blockhtml(ob, data[:i], rndr.mk.opaque)
	}

	return i
}

func find_block_tag(data []byte) (string, bool) {
	i := 0
	for i < len(data) && ((data[i] >= '0' && data[i] <= '9') || (data[i] >= 'A' && data[i] <= 'Z') || (data[i] >= 'a' && data[i] <= 'z')) {
		i++
	}
	if i >= len(data) {
		return "", false
	}
	key := string(data[:i])
	if block_tags[key] {
		return key, true
	}
	return "", false
}

func htmlblock_end(tag string, rndr *render, data []byte) int {
	// assume data[0] == '<' && data[1] == '/' already tested

	// check if tag is a match
	if len(tag)+3 >= len(data) || bytes.Compare(data[2:2+len(tag)], []byte(tag)) != 0 || data[len(tag)+2] != '>' {
		return 0
	}

	// check white lines
	i := len(tag) + 3
	w := 0
	if i < len(data) {
		if w = is_empty(data[i:]); w == 0 {
			return 0 // non-blank after tag
		}
	}
	i += w
	w = 0

	if rndr.ext_flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
		if i < len(data) {
			w = is_empty(data[i:])
		}
	} else {
		if i < len(data) {
			if w = is_empty(data[i:]); w == 0 {
				return 0 // non-blank line after tag line
			}
		}
	}

	return i + w
}

func is_empty(data []byte) int {
	var i int
	for i = 0; i < len(data) && data[i] != '\n'; i++ {
		if data[i] != ' ' && data[i] != '\t' {
			return 0
		}
	}
	return i + 1
}

func is_hrule(data []byte) bool {
	// skip initial spaces
	if len(data) < 3 {
		return false
	}
	i := 0
	if data[0] == ' ' {
		i++
		if data[1] == ' ' {
			i++
			if data[2] == ' ' {
				i++
			}
		}
	}

	// look at the hrule char
	if i+2 >= len(data) || (data[i] != '*' && data[i] != '-' && data[i] != '_') {
		return false
	}
	c := data[i]

	// the whole line must be the char or whitespace
	n := 0
	for i < len(data) && data[i] != '\n' {
		switch {
		case data[i] == c:
			n++
		case data[i] != ' ' && data[i] != '\t':
			return false
		}
		i++
	}

	return n >= 3
}

func is_codefence(data []byte, syntax **string) int {
	i, n := 0, 0

	// skip initial spaces
	if len(data) < 3 {
		return 0
	}
	if data[0] == ' ' {
		i++
		if data[1] == ' ' {
			i++
			if data[2] == ' ' {
				i++
			}
		}
	}

	// look at the hrule char
	if i+2 >= len(data) || !(data[i] == '~' || data[i] == '`') {
		return 0
	}

	c := data[i]

	// the whole line must be the char or whitespace
	for i < len(data) && data[i] == c {
		n++
		i++
	}

	if n < 3 {
		return 0
	}

	if syntax != nil {
		syn := 0

		for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
			i++
		}

		syntax_start := i

		if i < len(data) && data[i] == '{' {
			i++
			syntax_start++

			for i < len(data) && data[i] != '}' && data[i] != '\n' {
				syn++
				i++
			}

			if i == len(data) || data[i] != '}' {
				return 0
			}

			// string all whitespace at the beginning and the end
			// of the {} block
			for syn > 0 && isspace(data[syntax_start]) {
				syntax_start++
				syn--
			}

			for syn > 0 && isspace(data[syntax_start+syn-1]) {
				syn--
			}

			i++
		} else {
			for i < len(data) && !isspace(data[i]) {
				syn++
				i++
			}
		}

		language := string(data[syntax_start : syntax_start+syn])
		*syntax = &language
	}

	for i < len(data) && data[i] != '\n' {
		if !isspace(data[i]) {
			return 0
		}
		i++
	}

	return i + 1
}

func parse_fencedcode(ob *bytes.Buffer, rndr *render, data []byte) int {
	var lang *string
	beg := is_codefence(data, &lang)
	if beg == 0 {
		return 0
	}

	work := bytes.NewBuffer(nil)

	for beg < len(data) {
		fence_end := is_codefence(data[beg:], nil)
		if fence_end != 0 {
			beg += fence_end
			break
		}

		var end int
		for end = beg + 1; end < len(data) && data[end-1] != '\n'; end++ {
		}

		if beg < end {
			// verbatim copy to the working buffer, escaping entities
			if is_empty(data[beg:]) > 0 {
				work.WriteByte('\n')
			} else {
				work.Write(data[beg:end])
			}
		}
		beg = end
	}

	if work.Len() > 0 && work.Bytes()[work.Len()-1] != '\n' {
		work.WriteByte('\n')
	}

	if rndr.mk.blockcode != nil {
		syntax := ""
		if lang != nil {
			syntax = *lang
		}

		rndr.mk.blockcode(ob, work.Bytes(), syntax, rndr.mk.opaque)
	}

	return beg
}

func parse_table(ob *bytes.Buffer, rndr *render, data []byte) int {
	header_work := bytes.NewBuffer(nil)
	i, columns, col_data := parse_table_header(header_work, rndr, data)
	if i > 0 {
		body_work := bytes.NewBuffer(nil)

		for i < len(data) {
			pipes, row_start := 0, i
			for ; i < len(data) && data[i] != '\n'; i++ {
				if data[i] == '|' {
					pipes++
				}
			}

			if pipes == 0 || i == len(data) {
				i = row_start
				break
			}

			parse_table_row(body_work, rndr, data[row_start:i], columns, col_data)
			i++
		}

		if rndr.mk.table != nil {
			rndr.mk.table(ob, header_work.Bytes(), body_work.Bytes(), rndr.mk.opaque)
		}
	}

	return i
}

func parse_table_header(ob *bytes.Buffer, rndr *render, data []byte) (size int, columns int, column_data []int) {
	i, pipes := 0, 0
	column_data = []int{}
	for i = 0; i < len(data) && data[i] != '\n'; i++ {
		if data[i] == '|' {
			pipes++
		}
	}

	if i == len(data) || pipes == 0 {
		return 0, 0, column_data
	}

	header_end := i

	if data[0] == '|' {
		pipes--
	}

	if i > 2 && data[i-1] == '|' {
		pipes--
	}

	columns = pipes + 1
	column_data = make([]int, columns)

	// parse the header underline
	i++
	if i < len(data) && data[i] == '|' {
		i++
	}

	under_end := i
	for under_end < len(data) && data[under_end] != '\n' {
		under_end++
	}

	col := 0
	for ; col < columns && i < under_end; col++ {
		dashes := 0

		for i < under_end && (data[i] == ' ' || data[i] == '\t') {
			i++
		}

		if data[i] == ':' {
			i++
			column_data[col] |= TABLE_ALIGNMENT_LEFT
			dashes++
		}

		for i < under_end && data[i] == '-' {
			i++
			dashes++
		}

		if i < under_end && data[i] == ':' {
			i++
			column_data[col] |= TABLE_ALIGNMENT_RIGHT
			dashes++
		}

		for i < under_end && (data[i] == ' ' || data[i] == '\t') {
			i++
		}

		if i < under_end && data[i] != '|' {
			break
		}

		if dashes < 3 {
			break
		}

		i++
	}

	if col < columns {
		return 0, 0, column_data
	}

	parse_table_row(ob, rndr, data[:header_end], columns, column_data)
	size = under_end + 1
	return
}

func parse_table_row(ob *bytes.Buffer, rndr *render, data []byte, columns int, col_data []int) {
	i, col := 0, 0
	row_work := bytes.NewBuffer(nil)

	if i < len(data) && data[i] == '|' {
		i++
	}

	for col = 0; col < columns && i < len(data); col++ {
		for i < len(data) && isspace(data[i]) {
			i++
		}

		cell_start := i

		for i < len(data) && data[i] != '|' {
			i++
		}

		cell_end := i - 1

		for cell_end > cell_start && isspace(data[cell_end]) {
			cell_end--
		}

		cell_work := bytes.NewBuffer(nil)
		parse_inline(cell_work, rndr, data[cell_start:cell_end+1])

		if rndr.mk.table_cell != nil {
			cdata := 0
			if col < len(col_data) {
				cdata = col_data[col]
			}
			rndr.mk.table_cell(row_work, cell_work.Bytes(), cdata, rndr.mk.opaque)
		}

		i++
	}

	for ; col < columns; col++ {
		empty_cell := []byte{}
		if rndr.mk.table_cell != nil {
			cdata := 0
			if col < len(col_data) {
				cdata = col_data[col]
			}
			rndr.mk.table_cell(row_work, empty_cell, cdata, rndr.mk.opaque)
		}
	}

	if rndr.mk.table_row != nil {
		rndr.mk.table_row(ob, row_work.Bytes(), rndr.mk.opaque)
	}
}

// returns blockquote prefix length
func prefix_quote(data []byte) int {
	i := 0
	for i < len(data) && i < 3 && data[i] == ' ' {
		i++
	}
	if i < len(data) && data[i] == '>' {
		if i+1 < len(data) && (data[i+1] == ' ' || data[i+1] == '\t') {
			return i + 2
		}
		return i + 1
	}
	return 0
}

// parse a blockquote fragment
func parse_blockquote(ob *bytes.Buffer, rndr *render, data []byte) int {
	out := bytes.NewBuffer(nil)
	work := bytes.NewBuffer(nil)
	beg, end := 0, 0
	for beg < len(data) {
		for end = beg + 1; end < len(data) && data[end-1] != '\n'; end++ {
		}

		if pre := prefix_quote(data[beg:]); pre > 0 {
			beg += pre // skip prefix
		} else {
			// empty line followed by non-quote line
			if is_empty(data[beg:]) > 0 && (end >= len(data) || (prefix_quote(data[end:]) == 0 && is_empty(data[end:]) == 0)) {
				break
			}
		}

		if beg < end { // copy into the in-place working buffer
			work.Write(data[beg:end])
		}
		beg = end
	}

	parse_block(out, rndr, work.Bytes())
	if rndr.mk.blockquote != nil {
		rndr.mk.blockquote(ob, out.Bytes(), rndr.mk.opaque)
	}
	return end
}

// returns prefix length for block code
func prefix_code(data []byte) int {
	if len(data) > 0 && data[0] == '\t' {
		return 1
	}
	if len(data) > 3 && data[0] == ' ' && data[1] == ' ' && data[2] == ' ' && data[3] == ' ' {
		return 4
	}
	return 0
}

func parse_blockcode(ob *bytes.Buffer, rndr *render, data []byte) int {
	work := bytes.NewBuffer(nil)

	beg, end := 0, 0
	for beg < len(data) {
		for end = beg + 1; end < len(data) && data[end-1] != '\n'; end++ {
		}

		if pre := prefix_code(data[beg:end]); pre > 0 {
			beg += pre
		} else {
			if is_empty(data[beg:end]) == 0 {
				// non-empty non-prefixed line breaks the pre
				break
			}
		}

		if beg < end {
			// verbatim copy to the working buffer, escaping entities
			if is_empty(data[beg:end]) > 0 {
				work.WriteByte('\n')
			} else {
				work.Write(data[beg:end])
			}
		}
		beg = end
	}

	// trim all the \n off the end of work
	workbytes := work.Bytes()
	n := 0
	for len(workbytes) > n && workbytes[len(workbytes)-n-1] == '\n' {
		n++
	}
	if n > 0 {
		work = bytes.NewBuffer(workbytes[:len(workbytes)-n])
	}

	work.WriteByte('\n')

	if rndr.mk.blockcode != nil {
		rndr.mk.blockcode(ob, work.Bytes(), "", rndr.mk.opaque)
	}

	return beg
}

// returns unordered list item prefix
func prefix_uli(data []byte) int {
	i := 0
	for i < len(data) && i < 3 && data[i] == ' ' {
		i++
	}
	if i+1 >= len(data) || (data[i] != '*' && data[i] != '+' && data[i] != '-') || (data[i+1] != ' ' && data[i+1] != '\t') {
		return 0
	}
	return i + 2
}

// returns ordered list item prefix
func prefix_oli(data []byte) int {
	i := 0
	for i < len(data) && i < 3 && data[i] == ' ' {
		i++
	}
	if i >= len(data) || data[i] < '0' || data[i] > '9' {
		return 0
	}
	for i < len(data) && data[i] >= '0' && data[i] <= '9' {
		i++
	}
	if i+1 >= len(data) || data[i] != '.' || (data[i+1] != ' ' && data[i+1] != '\t') {
		return 0
	}
	return i + 2
}

// parse ordered or unordered list block
func parse_list(ob *bytes.Buffer, rndr *render, data []byte, flags int) int {
	work := bytes.NewBuffer(nil)

	i, j := 0, 0
	for i < len(data) {
		j = parse_listitem(work, rndr, data[i:], &flags)
		i += j

		if j == 0 || flags&LIST_ITEM_END_OF_LIST != 0 {
			break
		}
	}

	if rndr.mk.list != nil {
		rndr.mk.list(ob, work.Bytes(), flags, rndr.mk.opaque)
	}
	return i
}

// parse a single list item
// assumes initial prefix is already removed
func parse_listitem(ob *bytes.Buffer, rndr *render, data []byte, flags *int) int {
	// keep track of the first indentation prefix
	beg, end, pre, sublist, orgpre, i := 0, 0, 0, 0, 0, 0

	for orgpre < 3 && orgpre < len(data) && data[orgpre] == ' ' {
		orgpre++
	}

	beg = prefix_uli(data)
	if beg == 0 {
		beg = prefix_oli(data)
	}
	if beg == 0 {
		return 0
	}

	// skip leading whitespace on first line
	for beg < len(data) && data[beg] == ' ' {
		beg++
	}

	// skip to the beginning of the following line
	end = beg
	for end < len(data) && data[end-1] != '\n' {
		end++
	}

	// get working buffers
	work := bytes.NewBuffer(nil)
	inter := bytes.NewBuffer(nil)

	// put the first line into the working buffer
	work.Write(data[beg:end])
	beg = end

	// process the following lines
	in_empty, has_inside_empty := false, false
	for beg < len(data) {
		end++

		for end < len(data) && data[end-1] != '\n' {
			end++
		}

		// process an empty line
		if is_empty(data[beg:end]) > 0 {
			in_empty = true
			beg = end
			continue
		}

		// calculate the indentation
		i = 0
		for i < 4 && beg+i < end && data[beg+i] == ' ' {
			i++
		}

		pre = i
		if data[beg] == '\t' {
			i = 1
			pre = 8
		}

		// check for a new item
		chunk := data[beg+i : end]
		if (prefix_uli(chunk) > 0 && !is_hrule(chunk)) || prefix_oli(chunk) > 0 {
			if in_empty {
				has_inside_empty = true
			}

			if pre == orgpre { // the following item must have the same indentation
				break
			}

			if sublist == 0 {
				sublist = work.Len()
			}
		} else {
			// only join indented stuff after empty lines
			if in_empty && i < 4 && data[beg] != '\t' {
				*flags |= LIST_ITEM_END_OF_LIST
				break
			} else {
				if in_empty {
					work.WriteByte('\n')
					has_inside_empty = true
				}
			}
		}

		in_empty = false

		// add the line into the working buffer without prefix
		work.Write(data[beg+i : end])
		beg = end
	}

	// render li contents
	if has_inside_empty {
		*flags |= LIST_ITEM_CONTAINS_BLOCK
	}

	workbytes := work.Bytes()
	if *flags&LIST_ITEM_CONTAINS_BLOCK != 0 {
		// intermediate render of block li
		if sublist > 0 && sublist < len(workbytes) {
			parse_block(inter, rndr, workbytes[:sublist])
			parse_block(inter, rndr, workbytes[sublist:])
		} else {
			parse_block(inter, rndr, workbytes)
		}
	} else {
		// intermediate render of inline li
		if sublist > 0 && sublist < len(workbytes) {
			parse_inline(inter, rndr, workbytes[:sublist])
			parse_block(inter, rndr, workbytes[sublist:])
		} else {
			parse_inline(inter, rndr, workbytes)
		}
	}

	// render li itself
	if rndr.mk.listitem != nil {
		rndr.mk.listitem(ob, inter.Bytes(), *flags, rndr.mk.opaque)
	}

	return beg
}

func parse_paragraph(ob *bytes.Buffer, rndr *render, data []byte) int {
	i, end, level := 0, 0, 0

	for i < len(data) {
		for end = i + 1; end < len(data) && data[end-1] != '\n'; end++ {
		}

		if is_empty(data[i:]) > 0 {
			break
		}
		if level = is_headerline(data[i:]); level > 0 {
			break
		}

		if rndr.ext_flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
			if data[i] == '<' && rndr.mk.blockhtml != nil && parse_htmlblock(ob, rndr, data[i:], false) > 0 {
				end = i
				break
			}
		}

		if is_atxheader(rndr, data[i:]) || is_hrule(data[i:]) {
			end = i
			break
		}

		i = end
	}

	work := data
	size := i
	for size > 0 && work[size-1] == '\n' {
		size--
	}

	if level == 0 {
		tmp := bytes.NewBuffer(nil)
		parse_inline(tmp, rndr, work[:size])
		if rndr.mk.paragraph != nil {
			rndr.mk.paragraph(ob, tmp.Bytes(), rndr.mk.opaque)
		}
	} else {
		if size > 0 {
			beg := 0
			i = size
			size--

			for size > 0 && work[size] != '\n' {
				size--
			}

			beg = size + 1
			for size > 0 && work[size-1] == '\n' {
				size--
			}

			if size > 0 {
				tmp := bytes.NewBuffer(nil)
				parse_inline(tmp, rndr, work[:size])
				if rndr.mk.paragraph != nil {
					rndr.mk.paragraph(ob, tmp.Bytes(), rndr.mk.opaque)
				}

				work = work[beg:]
				size = i - beg
			} else {
				size = i
			}
		}

		header_work := bytes.NewBuffer(nil)
		parse_inline(header_work, rndr, work[:size])

		if rndr.mk.header != nil {
			rndr.mk.header(ob, header_work.Bytes(), level, rndr.mk.opaque)
		}
	}

	return end
}


//
// Link references
//
// This section implements support for references that (usually) appear
// as footnotes in a document, and can be referenced anywhere in the document.
// The basic format is:
//
//    [1]: http://www.google.com/ "Google"
//    [2]: http://www.github.com/ "Github"
//
// Anywhere in the document, the reference can be linked by referring to its
// label, i.e., 1 and 2 in this example, as in:
//
//    This library is hosted on [Github][2], a git hosting site.

// References are parsed and stored in this struct.
type link_ref struct {
	id    []byte
	link  []byte
	title []byte
}

// The list of all reference links is stored in this type.
type link_ref_array []*link_ref

// Find the length of a list of references.
// This implements an interface needed for sorting.
func (elt link_ref_array) Len() int {
	return len(elt)
}

// Test if one reference is less than another (case-insensitive).
// This implements an interface needed for sorting.
func (elt link_ref_array) Less(i, j int) bool {
	return byteslice_less(elt[i].id, elt[j].id)
}

// Compare two []byte values (case-insensitive), returning
// true if a is less than b.
func byteslice_less(a []byte, b []byte) bool {
	// adapted from bytes.Compare in stdlib
	m := len(a)
	if m > len(b) {
		m = len(b)
	}
	for i, ac := range a[0:m] {
		// do a case-insensitive comparison
		ai, bi := unicode.ToLower(int(ac)), unicode.ToLower(int(b[i]))
		switch {
		case ai > bi:
			return false
		case ai < bi:
			return true
		}
	}
	switch {
	case len(a) < len(b):
		return true
	case len(a) > len(b):
		return false
	}
	return false
}

// Swap two references.
// This implements an interface needed for sorting.
func (elt link_ref_array) Swap(i, j int) {
	elt[i], elt[j] = elt[j], elt[i]
}

// Check whether or not data starts with a reference link.
// If so, it is parsed and stored in the list of references
// (in the render struct).
// Returns the number of bytes to skip to move past it, or zero
// if there is the first line is not a reference.
func is_ref(rndr *render, data []byte) int {
	// up to 3 optional leading spaces
	if len(data) < 4 {
		return 0
	}
	i := 0
	for i < 3 && data[i] == ' ' {
		i++
	}
	if data[i] == ' ' {
		return 0
	}

	// id part: anything but a newline between brackets
	if data[i] != '[' {
		return 0
	}
	i++
	id_offset := i
	for i < len(data) && data[i] != '\n' && data[i] != '\r' && data[i] != ']' {
		i++
	}
	if i >= len(data) || data[i] != ']' {
		return 0
	}
	id_end := i

	// spacer: colon (space | tab)* newline? (space | tab)*
	i++
	if i >= len(data) || data[i] != ':' {
		return 0
	}
	i++
	for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
		i++
	}
	if i < len(data) && (data[i] == '\n' || data[i] == '\r') {
		i++
		if i < len(data) && data[i] == '\n' && data[i-1] == '\r' {
			i++
		}
	}
	for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
		i++
	}
	if i >= len(data) {
		return 0
	}

	// link: whitespace-free sequence, optionally between angle brackets
	if data[i] == '<' {
		i++
	}
	link_offset := i
	for i < len(data) && data[i] != ' ' && data[i] != '\t' && data[i] != '\n' && data[i] != '\r' {
		i++
	}
	link_end := i
	if data[link_offset] == '<' && data[link_end-1] == '>' {
		link_offset++
		link_end--
	}

	// optional spacer: (space | tab)* (newline | '\'' | '"' | '(' )
	for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
		i++
	}
	if i < len(data) && data[i] != '\n' && data[i] != '\r' && data[i] != '\'' && data[i] != '"' && data[i] != '(' {
		return 0
	}

	// compute end-of-line
	line_end := 0
	if i >= len(data) || data[i] == '\r' || data[i] == '\n' {
		line_end = i
	}
	if i+1 < len(data) && data[i] == '\r' && data[i+1] == '\n' {
		line_end++
	}

	// optional (space|tab)* spacer after a newline
	if line_end > 0 {
		i = line_end + 1
		for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
			i++
		}
	}

	// optional title: any non-newline sequence enclosed in '"() alone on its line
	title_offset, title_end := 0, 0
	if i+1 < len(data) && (data[i] == '\'' || data[i] == '"' || data[i] == '(') {
		i++
		title_offset = i

		// look for EOL
		for i < len(data) && data[i] != '\n' && data[i] != '\r' {
			i++
		}
		if i+1 < len(data) && data[i] == '\n' && data[i+1] == '\r' {
			title_end = i + 1
		} else {
			title_end = i
		}

		// step back
		i--
		for i > title_offset && (data[i] == ' ' || data[i] == '\t') {
			i--
		}
		if i > title_offset && (data[i] == '\'' || data[i] == '"' || data[i] == ')') {
			line_end = title_end
			title_end = i
		}
	}
	if line_end == 0 { // garbage after the link
		return 0
	}

	// a valid ref has been found
	if rndr == nil {
		return line_end
	}
	item := &link_ref{id: data[id_offset:id_end], link: data[link_offset:link_end], title: data[title_offset:title_end]}
	rndr.refs = append(rndr.refs, item)

	return line_end
}


//
//
// Miscellaneous helper functions
//
//


// Test if a character is a punctuation symbol.
// Taken from a private function in regexp in the stdlib.
func ispunct(c byte) bool {
	for _, r := range []byte("!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~") {
		if c == r {
			return true
		}
	}
	return false
}

// this is sort.Search, reproduced here because an older
// version of the library had a bug
func sortDotSearch(n int, f func(int) bool) int {
	// Define f(-1) == false and f(n) == true.
	// Invariant: f(i-1) == false, f(j) == true.
	i, j := 0, n
	for i < j {
		h := i + (j-i)/2 // avoid overflow when computing h
		// i ≤ h < j
		if !f(h) {
			i = h + 1 // preserves f(i-1) == false
		} else {
			j = h // preserves f(j) == true
		}
	}
	// i == j, f(i-1) == false, and f(j) (= f(i)) == true  =>  answer is i.
	return i
}

// Test if a character is a whitespace character.
func isspace(c byte) bool {
	return c == ' ' || c == '\t' || c == '\n' || c == '\r' || c == '\f' || c == '\v'
}

// Test if a character is a letter or a digit.
func isalnum(c byte) bool {
	return (c >= '0' && c <= '9') || (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z')
}

// Replace tab characters with spaces, aligning to the next TAB_SIZE column.
func expand_tabs(ob *bytes.Buffer, line []byte) {
	i, tab := 0, 0

	for i < len(line) {
		org := i
		for i < len(line) && line[i] != '\t' {
			i++
			tab++
		}

		if i > org {
			ob.Write(line[org:i])
		}

		if i >= len(line) {
			break
		}

		for {
			ob.WriteByte(' ')
			tab++
			if tab%TAB_SIZE == 0 {
				break
			}
		}

		i++
	}
}