lgen: comparison markdown.lua

-:038275cd92ed
+:836dac92eced
 #!/usr/bin/env lua
---[[
-# markdown.lua -- version 0.32
-<http://www.frykholm.se/files/markdown.lua>
-**Author:** Niklas Frykholm, <niklas@frykholm.se>
-**Date:** 31 May 2008
-This is an implementation of the popular text markup language Markdown in pure Lua.
-Markdown can convert documents written in a simple and easy to read text format
-to well-formatted HTML. For a more thourough description of Markdown and the Markdown
-syntax, see <http://daringfireball.net/projects/markdown>.
-The original Markdown source is written in Perl and makes heavy use of advanced
-regular expression techniques (such as negative look-ahead, etc) which are not available
-in Lua's simple regex engine. Therefore this Lua port has been rewritten from the ground
-up. It is probably not completely bug free. If you notice any bugs, please report them to
-me. A unit test that exposes the error is helpful.
-## Usage
-require "markdown"
-markdown(source)
-``markdown.lua`` exposes a single global function named ``markdown(s)`` which applies the
-Markdown transformation to the specified string.
-``markdown.lua`` can also be used directly from the command line:
-	lua markdown.lua test.md
-Creates a file ``test.html`` with the converted content of ``test.md``. Run:
-lua markdown.lua -h
-For a description of the command-line options.
-``markdown.lua`` uses the same license as Lua, the MIT license.
-## License
-Copyright &copy; 2008 Niklas Frykholm.
-Permission is hereby granted, free of charge, to any person obtaining a copy of this
-software and associated documentation files (the "Software"), to deal in the Software
-without restriction, including without limitation the rights to use, copy, modify, merge,
-publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons
-to whom the Software is furnished to do so, subject to the following conditions:
-The above copyright notice and this permission notice shall be included in all copies
-or substantial portions of the Software.
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
-THE SOFTWARE.
-## Version history
--	**0.32** -- 31 May 2008
-	- Fix for links containing brackets
--	**0.31** -- 1 Mar 2008
-	-	Fix for link definitions followed by spaces
--	**0.30** -- 25 Feb 2008
-	-	Consistent behavior with Markdown when the same link reference is reused
--	**0.29** -- 24 Feb 2008
-	-	Fix for <pre> blocks with spaces in them
--	**0.28** -- 18 Feb 2008
-	-	Fix for link encoding
--	**0.27** -- 14 Feb 2008
-	-	Fix for link database links with ()
--	**0.26** -- 06 Feb 2008
-	-	Fix for nested italic and bold markers
--	**0.25** -- 24 Jan 2008
-	-	Fix for encoding of naked <
--	**0.24** -- 21 Jan 2008
-	-	Fix for link behavior.
--	**0.23** -- 10 Jan 2008
-	-	Fix for a regression bug in longer expressions in italic or bold.
--	**0.22** -- 27 Dec 2007
-	-	Fix for crash when processing blocks with a percent sign in them.
--	**0.21** -- 27 Dec 2007
-	- 	Fix for combined strong and emphasis tags
--	**0.20** -- 13 Oct 2007
-	-	Fix for < as well in image titles, now matches Dingus behavior
--	**0.19** -- 28 Sep 2007
-	-	Fix for quotation marks " and ampersands & in link and image titles.
--	**0.18** -- 28 Jul 2007
-	-	Does not crash on unmatched tags (behaves like standard markdown)
--	**0.17** -- 12 Apr 2007
-	-	Fix for links with %20 in them.
--	**0.16** -- 12 Apr 2007
-	-	Do not require arg global to exist.
--	**0.15** -- 28 Aug 2006
-	-	Better handling of links with underscores in them.
--	**0.14** -- 22 Aug 2006
-	-	Bug for *`foo()`*
--	**0.13** -- 12 Aug 2006
-	-	Added -l option for including stylesheet inline in document.
-	-	Fixed bug in -s flag.
-	-	Fixed emphasis bug.
--	**0.12** -- 15 May 2006
-	-	Fixed several bugs to comply with MarkdownTest 1.0 <http://six.pairlist.net/pipermail/markdown-discuss/2004-December/000909.html>
--	**0.11** -- 12 May 2006
-	-	Fixed bug for escaping `*` and `_` inside code spans.
-	-	Added license terms.
-	-	Changed join() to table.concat().
--	**0.10** -- 3 May 2006
-	-	Initial public release.
-// Niklas
-]]
--- Set up a table for holding local functions to avoid polluting the global namespace
-local M = {}
-local MT = {__index = _G}
-setmetatable(M, MT)
-setfenv(1, M)
 ----------------------------------------------------------------------
 -- Utility functions
 ----------------------------------------------------------------------
--- Locks table t from changes, writes an error if someone attempts to change the table.
+local unpack = table.unpack or unpack
--- This is useful for detecting variables that have "accidently" been made global. Something
--- I tend to do all too much.
-function lock(t)
-	function lock_new_index(t, k, v)
-		error("module has been locked -- " .. k .. " must be declared local", 2)
-	end
-	local mt = {__newindex = lock_new_index}
-	if getmetatable(t) then mt.__index = getmetatable(t).__index end
-	setmetatable(t, mt)
-end
 -- Returns the result of mapping the values in table t through the function f
-function map(t, f)
+local function map(t, f)
-	local out = {}
+local out = {}
-	for k,v in pairs(t) do out[k] = f(v,k) end
+for k,v in pairs(t) do out[k] = f(v,k) end
-	return out
+return out
 end
--- The identity function, useful as a placeholder.
-function identity(text) return text end
 -- Functional style if statement. (NOTE: no short circuit evaluation)
-function iff(t, a, b) if t then return a else return b end end
+local function iff(t, a, b) if t then return a else return b end end
 -- Splits the text into an array of separate lines.
-function split(text, sep)
+local function split(text, sep)
-	sep = sep or "\n"
+sep = sep or "\n"
-	local lines = {}
+local lines = {}
-	local pos = 1
+local pos = 1
-	while true do
+while true do
-		local b,e = text:find(sep, pos)
+local b,e = text:find(sep, pos)
-		if not b then table.insert(lines, text:sub(pos)) break end
+if not b then table.insert(lines, text:sub(pos)) break end
-		table.insert(lines, text:sub(pos, b-1))
+table.insert(lines, text:sub(pos, b-1))
-		pos = e + 1
+pos = e + 1
-	end
+end
-	return lines
+return lines
 end
 -- Converts tabs to spaces
-function detab(text)
+local function detab(text)
-	local tab_width = 4
+local tab_width = 4
-	local function rep(match)
+local function rep(match)
-		local spaces = -match:len()
+local spaces = -match:len()
-		while spaces<1 do spaces = spaces + tab_width end
+while spaces<1 do spaces = spaces + tab_width end
-		return match .. string.rep(" ", spaces)
+return match .. string.rep(" ", spaces)
-	end
+end
-	text = text:gsub("([^\n]-)\t", rep)
+text = text:gsub("([^\n]-)\t", rep)
-	return text
+return text
 end
 -- Applies string.find for every pattern in the list and returns the first match
-function find_first(s, patterns, index)
+local function find_first(s, patterns, index)
-	local res = {}
+local res = {}
-	for _,p in ipairs(patterns) do
+for _,p in ipairs(patterns) do
-		local match = {s:find(p, index)}
+local match = {s:find(p, index)}
-		if #match>0 and (#res==0 or match[1] < res[1]) then res = match end
+if #match>0 and (#res==0 or match[1] < res[1]) then res = match end
-	end
+end
-	return table.unpack(res)
+return unpack(res)
 end
 -- If a replacement array is specified, the range [start, stop] in the array is replaced
 -- with the replacement array and the resulting array is returned. Without a replacement
 -- array the section of the array between start and stop is returned.
-function splice(array, start, stop, replacement)
+local function splice(array, start, stop, replacement)
-	if replacement then
+if replacement then
-		local n = stop - start + 1
+local n = stop - start + 1
-		while n > 0 do
+while n > 0 do
-			table.remove(array, start)
+table.remove(array, start)
-			n = n - 1
+n = n - 1
-		end
+end
-		for i,v in ipairs(replacement) do
+for _,v in ipairs(replacement) do
-			table.insert(array, start, v)
+table.insert(array, start, v)
-		end
+end
-		return array
+return array
-	else
+else
-		local res = {}
+local res = {}
-		for i = start,stop do
+for i = start,stop do
-			table.insert(res, array[i])
+table.insert(res, array[i])
-		end
+end
-		return res
+return res
-	end
+end
 end
 -- Outdents the text one step.
-function outdent(text)
+local function outdent(text)
-	text = "\n" .. text
+text = "\n" .. text
-	text = text:gsub("\n  ? ? ?", "\n")
+text = text:gsub("\n  ? ? ?", "\n")
-	text = text:sub(2)
+text = text:sub(2)
-	return text
+return text
 end
 -- Indents the text one step.
-function indent(text)
+local function indent(text)
-	text = text:gsub("\n", "\n    ")
+text = text:gsub("\n", "\n    ")
-	return text
+return text
 end
 -- Does a simple tokenization of html data. Returns the data as a list of tokens.
 -- Each token is a table with a type field (which is either "tag" or "text") and
 -- a text field (which contains the original token data).
-function tokenize_html(html)
+local function tokenize_html(html)
-	local tokens = {}
+local tokens = {}
-	local pos = 1
+local pos = 1
-	while true do
+while true do
-		local start = find_first(html, {"<!%-%-", "<[a-z/!$]", "<%?"}, pos)
+local start = find_first(html, {"<!%-%-", "<[a-z/!$]", "<%?"}, pos)
-		if not start then
+if not start then
-			table.insert(tokens, {type="text", text=html:sub(pos)})
+table.insert(tokens, {type="text", text=html:sub(pos)})
-			break
+break
-		end
+end
-		if start ~= pos then table.insert(tokens, {type="text", text = html:sub(pos, start-1)}) end
+if start ~= pos then table.insert(tokens, {type="text", text = html:sub(pos, start-1)}) end
-		local _, stop
+local _, stop
-		if html:match("^<!%-%-", start) then
+if html:match("^<!%-%-", start) then
-			_,stop = html:find("%-%->", start)
+_,stop = html:find("%-%->", start)
-		elseif html:match("^<%?", start) then
+elseif html:match("^<%?", start) then
-			_,stop = html:find("?>", start)
+_,stop = html:find("?>", start)
-		else
+else
-			_,stop = html:find("%b<>", start)
+_,stop = html:find("%b<>", start)
-		end
+end
-		if not stop then
+if not stop then
-			-- error("Could not match html tag " .. html:sub(start,start+30))
+-- error("Could not match html tag " .. html:sub(start,start+30))
-		 	table.insert(tokens, {type="text", text=html:sub(start, start)})
+table.insert(tokens, {type="text", text=html:sub(start, start)})
-			pos = start + 1
+pos = start + 1
-		else
+else
-			table.insert(tokens, {type="tag", text=html:sub(start, stop)})
+table.insert(tokens, {type="tag", text=html:sub(start, stop)})
-			pos = stop + 1
+pos = stop + 1
-		end
+end
-	end
+end
-	return tokens
+return tokens
 end
 ----------------------------------------------------------------------
 -- Hash
 ----------------------------------------------------------------------
 -- in the document. (Note that this is not cryptographic hash, the hash
 -- function is not one-way.) The hash procedure is used to protect parts
 -- of the document from further processing.
 local HASH = {
-	-- Has the hash been inited.
+-- Has the hash been inited.
-	inited = false,
+inited = false,
-	-- The unique string prepended to all hash values. This is to ensure
+-- The unique string prepended to all hash values. This is to ensure
-	-- that hash values do not accidently coincide with an actual existing
+-- that hash values do not accidently coincide with an actual existing
-	-- string in the document.
+-- string in the document.
-	identifier = "",
+identifier = "",
-	-- Counter that counts up for each new hash instance.
+-- Counter that counts up for each new hash instance.
-	counter = 0,
+counter = 0,
-	-- Hash table.
+-- Hash table.
-	table = {}
+table = {}
 }
 -- Inits hashing. Creates a hash_identifier that doesn't occur anywhere
 -- in the text.
-function init_hash(text)
+local function init_hash(text)
-	HASH.inited = true
+HASH.inited = true
-	HASH.identifier = ""
+HASH.identifier = ""
-	HASH.counter = 0
+HASH.counter = 0
-	HASH.table = {}
+HASH.table = {}
-	local s = "HASH"
+local s = "HASH"
-	local counter = 0
+local counter = 0
-	local id
+local id
-	while true do
+while true do
-		id  = s .. counter
+id  = s .. counter
-		if not text:find(id, 1, true) then break end
+if not text:find(id, 1, true) then break end
-		counter = counter + 1
+counter = counter + 1
-	end
+end
-	HASH.identifier = id
+HASH.identifier = id
 end
 -- Returns the hashed value for s.
-function hash(s)
+local function hash(s)
-	assert(HASH.inited)
+assert(HASH.inited)
-	if not HASH.table[s] then
+if not HASH.table[s] then
-		HASH.counter = HASH.counter + 1
+HASH.counter = HASH.counter + 1
-		local id = HASH.identifier .. HASH.counter .. "X"
+local id = HASH.identifier .. HASH.counter .. "X"
-		HASH.table[s] = id
+HASH.table[s] = id
-	end
+end
-	return HASH.table[s]
+return HASH.table[s]
 end
 ----------------------------------------------------------------------
 -- Protection
 ----------------------------------------------------------------------
 -- The protection module is used to "protect" parts of a document
 -- so that they are not modified by subsequent processing steps.
 -- Protected parts are saved in a table for later unprotection
 -- Protection data
 local PD = {
-	-- Saved blocks that have been converted
+-- Saved blocks that have been converted
-	blocks = {},
+blocks = {},
-	-- Block level tags that will be protected
+-- Block level tags that will be protected
-	tags = {"p", "div", "h1", "h2", "h3", "h4", "h5", "h6", "blockquote",
+tags = {"p", "div", "h1", "h2", "h3", "h4", "h5", "h6", "blockquote",
-	"pre", "table", "dl", "ol", "ul", "script", "noscript", "form", "fieldset",
+"pre", "table", "dl", "ol", "ul", "script", "noscript", "form", "fieldset",
-	"iframe", "math", "ins", "del"}
+"iframe", "math", "ins", "del"}
 }
 -- Pattern for matching a block tag that begins and ends in the leftmost
 -- column and may contain indented subtags, i.e.
 -- <div>
 --    A nested block.
 --    <div>
 --        Nested data.
 --     </div>
 -- </div>
-function block_pattern(tag)
+local function block_pattern(tag)
-	return "\n<" .. tag .. ".-\n</" .. tag .. ">[ \t]*\n"
+return "\n<" .. tag .. ".-\n</" .. tag .. ">[ \t]*\n"
 end
 -- Pattern for matching a block tag that begins and ends with a newline
-function line_pattern(tag)
+local function line_pattern(tag)
-	return "\n<" .. tag .. ".-</" .. tag .. ">[ \t]*\n"
+return "\n<" .. tag .. ".-</" .. tag .. ">[ \t]*\n"
 end
 -- Protects the range of characters from start to stop in the text and
 -- returns the protected string.
-function protect_range(text, start, stop)
+local function protect_range(text, start, stop)
-	local s = text:sub(start, stop)
+local s = text:sub(start, stop)
-	local h = hash(s)
+local h = hash(s)
-	PD.blocks[h] = s
+PD.blocks[h] = s
-	text = text:sub(1,start) .. h .. text:sub(stop)
+text = text:sub(1,start) .. h .. text:sub(stop)
-	return text
+return text
 end
 -- Protect every part of the text that matches any of the patterns. The first
 -- matching pattern is protected first, etc.
-function protect_matches(text, patterns)
+local function protect_matches(text, patterns)
-	while true do
+while true do
-		local start, stop = find_first(text, patterns)
+local start, stop = find_first(text, patterns)
-		if not start then break end
+if not start then break end
-		text = protect_range(text, start, stop)
+text = protect_range(text, start, stop)
-	end
+end
-	return text
+return text
 end
 -- Protects blocklevel tags in the specified text
-function protect(text)
+local function protect(text)
-	-- First protect potentially nested block tags
+-- First protect potentially nested block tags
-	text = protect_matches(text, map(PD.tags, block_pattern))
+text = protect_matches(text, map(PD.tags, block_pattern))
-	-- Then protect block tags at the line level.
+-- Then protect block tags at the line level.
-	text = protect_matches(text, map(PD.tags, line_pattern))
+text = protect_matches(text, map(PD.tags, line_pattern))
-	-- Protect <hr> and comment tags
+-- Protect <hr> and comment tags
-	text = protect_matches(text, {"\n<hr[^>]->[ \t]*\n"})
+text = protect_matches(text, {"\n<hr[^>]->[ \t]*\n"})
-	text = protect_matches(text, {"\n<!%-%-.-%-%->[ \t]*\n"})
+text = protect_matches(text, {"\n<!%-%-.-%-%->[ \t]*\n"})
-	return text
+return text
 end
 -- Returns true if the string s is a hash resulting from protection
-function is_protected(s)
+local function is_protected(s)
-	return PD.blocks[s]
+return PD.blocks[s]
 end
 -- Unprotects the specified text by expanding all the nonces
-function unprotect(text)
+local function unprotect(text)
-	for k,v in pairs(PD.blocks) do
+for k,v in pairs(PD.blocks) do
-		v = v:gsub("%%", "%%%%")
+v = v:gsub("%%", "%%%%")
-		text = text:gsub(k, v)
+text = text:gsub(k, v)
-	end
+end
-	return text
+return text
 end
 ----------------------------------------------------------------------
 -- Block transform
 -- characters.
 -- Returns true if the line is a ruler of (char) characters.
 -- The line must contain at least three char characters and contain only spaces and
 -- char characters.
-function is_ruler_of(line, char)
+local function is_ruler_of(line, char)
-	if not line:match("^[ %" .. char .. "]*$") then return false end
+if not line:match("^[ %" .. char .. "]*$") then return false end
-	if not line:match("%" .. char .. ".*%" .. char .. ".*%" .. char) then return false end
+if not line:match("%" .. char .. ".*%" .. char .. ".*%" .. char) then return false end
-	return true
+return true
 end
 -- Identifies the block level formatting present in the line
-function classify(line)
+local function classify(line)
-	local info = {line = line, text = line}
+local info = {line = line, text = line}
-	if line:match("^    ") then
+if line:match("^    ") then
-		info.type = "indented"
+info.type = "indented"
-		info.outdented = line:sub(5)
+info.outdented = line:sub(5)
-		return info
+return info
-	end
+end
-	for _,c in ipairs({'*', '-', '_', '='}) do
+for _,c in ipairs({'*', '-', '_', '='}) do
-		if is_ruler_of(line, c) then
+if is_ruler_of(line, c) then
-			info.type = "ruler"
+info.type = "ruler"
-			info.ruler_char = c
+info.ruler_char = c
-			return info
+return info
-		end
+end
-	end
+end
-	if line == "" then
+if line == "" then
-		info.type = "blank"
+info.type = "blank"
-		return info
+return info
-	end
+end
-	if line:match("^(#+)[ \t]*(.-)[ \t]*#*[ \t]*$") then
+if line:match("^(#+)[ \t]*(.-)[ \t]*#*[ \t]*$") then
-		local m1, m2 = line:match("^(#+)[ \t]*(.-)[ \t]*#*[ \t]*$")
+local m1, m2 = line:match("^(#+)[ \t]*(.-)[ \t]*#*[ \t]*$")
-		info.type = "header"
+info.type = "header"
-		info.level = m1:len()
+info.level = m1:len()
-		info.text = m2
+info.text = m2
-		return info
+return info
-	end
+end
-	if line:match("^ ? ? ?(%d+)%.[ \t]+(.+)") then
+if line:match("^ ? ? ?(%d+)%.[ \t]+(.+)") then
-		local number, text = line:match("^ ? ? ?(%d+)%.[ \t]+(.+)")
+local number, text = line:match("^ ? ? ?(%d+)%.[ \t]+(.+)")
-		info.type = "list_item"
+info.type = "list_item"
-		info.list_type = "numeric"
+info.list_type = "numeric"
-		info.number = 0 + number
+info.number = 0 + number
-		info.text = text
+info.text = text
-		return info
+return info
-	end
+end
-	if line:match("^ ? ? ?([%*%+%-])[ \t]+(.+)") then
+if line:match("^ ? ? ?([%*%+%-])[ \t]+(.+)") then
-		local bullet, text = line:match("^ ? ? ?([%*%+%-])[ \t]+(.+)")
+local bullet, text = line:match("^ ? ? ?([%*%+%-])[ \t]+(.+)")
-		info.type = "list_item"
+info.type = "list_item"
-		info.list_type = "bullet"
+info.list_type = "bullet"
-		info.bullet = bullet
+info.bullet = bullet
-		info.text= text
+info.text= text
-		return info
+return info
-	end
+end
-	if line:match("^>[ \t]?(.*)") then
+if line:match("^>[ \t]?(.*)") then
-		info.type = "blockquote"
+info.type = "blockquote"
-		info.text = line:match("^>[ \t]?(.*)")
+info.text = line:match("^>[ \t]?(.*)")
-		return info
+return info
-	end
+end
-	if is_protected(line) then
+if is_protected(line) then
-		info.type = "raw"
+info.type = "raw"
-		info.html = unprotect(line)
+info.html = unprotect(line)
-		return info
+return info
-	end
+end
-	info.type = "normal"
+info.type = "normal"
-	return info
+return info
 end
 -- Find headers constisting of a normal line followed by a ruler and converts them to
 -- header entries.
-function headers(array)
+local function headers(array)
-	local i = 1
+local i = 1
-	while i <= #array - 1 do
+while i <= #array - 1 do
-		if array[i].type  == "normal" and array[i+1].type == "ruler" and
+if array[i].type  == "normal" and array[i+1].type == "ruler" and
-			(array[i+1].ruler_char == "-" or array[i+1].ruler_char == "=") then
+(array[i+1].ruler_char == "-" or array[i+1].ruler_char == "=") then
-			local info = {line = array[i].line}
+local info = {line = array[i].line}
-			info.text = info.line
+info.text = info.line
-			info.type = "header"
+info.type = "header"
-			info.level = iff(array[i+1].ruler_char == "=", 1, 2)
+info.level = iff(array[i+1].ruler_char == "=", 1, 2)
-			table.remove(array, i+1)
+table.remove(array, i+1)
-			array[i] = info
+array[i] = info
-		end
+end
-		i = i + 1
+i = i + 1
-	end
+end
-	return array
+return array
+end
+-- Forward declarations
+local block_transform, span_transform, encode_code
+-- Convert lines to html code
+local function blocks_to_html(lines, no_paragraphs)
+local out = {}
+local i = 1
+while i <= #lines do
+local line = lines[i]
+if line.type == "ruler" then
+table.insert(out, "<hr/>")
+elseif line.type == "raw" then
+table.insert(out, line.html)
+elseif line.type == "normal" then
+local s = line.line
+while i+1 <= #lines and lines[i+1].type == "normal" do
+i = i + 1
+s = s .. "\n" .. lines[i].line
+end
+if no_paragraphs then
+table.insert(out, span_transform(s))
+else
+table.insert(out, "<p>" .. span_transform(s) .. "</p>")
+end
+elseif line.type == "header" then
+local s = "<h" .. line.level .. ">" .. span_transform(line.text) .. "</h" .. line.level .. ">"
+table.insert(out, s)
+else
+table.insert(out, line.line)
+end
+i = i + 1
+end
+return out
 end
 -- Find list blocks and convert them to protected data blocks
-function lists(array, sublist)
+local function lists(array, sublist)
-	local function process_list(arr)
+local function process_list(arr)
-		local function any_blanks(arr)
+local function any_blanks(arr)
-			for i = 1, #arr do
+for i = 1, #arr do
-				if arr[i].type == "blank" then return true end
+if arr[i].type == "blank" then return true end
-			end
+end
-			return false
+return false
-		end
+end
-		local function split_list_items(arr)
+local function split_list_items(arr)
-			local acc = {arr[1]}
+local acc = {arr[1]}
-			local res = {}
+local res = {}
-			for i=2,#arr do
+for i=2,#arr do
-				if arr[i].type == "list_item" then
+if arr[i].type == "list_item" then
-					table.insert(res, acc)
+table.insert(res, acc)
-					acc = {arr[i]}
+acc = {arr[i]}
-				else
+else
-					table.insert(acc, arr[i])
+table.insert(acc, arr[i])
-				end
+end
-			end
+end
-			table.insert(res, acc)
+table.insert(res, acc)
-			return res
+return res
-		end
+end
-		local function process_list_item(lines, block)
+local function process_list_item(lines, block)
-			while lines[#lines].type == "blank" do
+while lines[#lines].type == "blank" do
-				table.remove(lines)
+table.remove(lines)
-			end
+end
-			local itemtext = lines[1].text
+local itemtext = lines[1].text
-			for i=2,#lines do
+for i=2,#lines do
-				itemtext = itemtext .. "\n" .. outdent(lines[i].line)
+itemtext = itemtext .. "\n" .. outdent(lines[i].line)
-			end
+end
-			if block then
+if block then
-				itemtext = block_transform(itemtext, true)
+itemtext = block_transform(itemtext, true)
-				if not itemtext:find("<pre>") then itemtext = indent(itemtext) end
+if not itemtext:find("<pre>") then itemtext = indent(itemtext) end
-				return "    <li>" .. itemtext .. "</li>"
+return "    <li>" .. itemtext .. "</li>"
-			else
+else
-				local lines = split(itemtext)
+local lines = split(itemtext)
-				lines = map(lines, classify)
+lines = map(lines, classify)
-				lines = lists(lines, true)
+lines = lists(lines, true)
-				lines = blocks_to_html(lines, true)
+lines = blocks_to_html(lines, true)
-				itemtext = table.concat(lines, "\n")
+itemtext = table.concat(lines, "\n")
-				if not itemtext:find("<pre>") then itemtext = indent(itemtext) end
+if not itemtext:find("<pre>") then itemtext = indent(itemtext) end
-				return "    <li>" .. itemtext .. "</li>"
+return "    <li>" .. itemtext .. "</li>"
-			end
+end
-		end
+end
-		local block_list = any_blanks(arr)
+local block_list = any_blanks(arr)
-		local items = split_list_items(arr)
+local items = split_list_items(arr)
-		local out = ""
+local out = ""
-		for _, item in ipairs(items) do
+for _, item in ipairs(items) do
-			out = out .. process_list_item(item, block_list) .. "\n"
+out = out .. process_list_item(item, block_list) .. "\n"
-		end
+end
-		if arr[1].list_type == "numeric" then
+if arr[1].list_type == "numeric" then
-			return "<ol>\n" .. out .. "</ol>"
+return "<ol>\n" .. out .. "</ol>"
-		else
+else
-			return "<ul>\n" .. out .. "</ul>"
+return "<ul>\n" .. out .. "</ul>"
-		end
+end
-	end
+end
-	-- Finds the range of lines composing the first list in the array. A list
+-- Finds the range of lines composing the first list in the array. A list
-	-- starts with (^ list_item) or (blank list_item) and ends with
+-- starts with (^ list_item) or (blank list_item) and ends with
-	-- (blank* $) or (blank normal).
+-- (blank* $) or (blank normal).
-	--
+--
-	-- A sublist can start with just (list_item) does not need a blank...
+-- A sublist can start with just (list_item) does not need a blank...
-	local function find_list(array, sublist)
+local function find_list(array, sublist)
-		local function find_list_start(array, sublist)
+local function find_list_start(array, sublist)
-			if array[1].type == "list_item" then return 1 end
+if array[1].type == "list_item" then return 1 end
-			if sublist then
+if sublist then
-				for i = 1,#array do
+for i = 1,#array do
-					if array[i].type == "list_item" then return i end
+if array[i].type == "list_item" then return i end
-				end
+end
-			else
+else
-				for i = 1, #array-1 do
+for i = 1, #array-1 do
-					if array[i].type == "blank" and array[i+1].type == "list_item" then
+if array[i].type == "blank" and array[i+1].type == "list_item" then
-						return i+1
+return i+1
-					end
+end
-				end
+end
-			end
+end
-			return nil
+return nil
-		end
+end
-		local function find_list_end(array, start)
+local function find_list_end(array, start)
-			local pos = #array
+local pos = #array
-			for i = start, #array-1 do
+for i = start, #array-1 do
-				if array[i].type == "blank" and array[i+1].type ~= "list_item"
+if array[i].type == "blank" and array[i+1].type ~= "list_item"
-					and array[i+1].type ~= "indented" and array[i+1].type ~= "blank" then
+and array[i+1].type ~= "indented" and array[i+1].type ~= "blank" then
-					pos = i-1
+pos = i-1
-					break
+break
-				end
+end
-			end
+end
-			while pos > start and array[pos].type == "blank" do
+while pos > start and array[pos].type == "blank" do
-				pos = pos - 1
+pos = pos - 1
-			end
+end
-			return pos
+return pos
-		end
+end
-		local start = find_list_start(array, sublist)
+local start = find_list_start(array, sublist)
-		if not start then return nil end
+if not start then return nil end
-		return start, find_list_end(array, start)
+return start, find_list_end(array, start)
-	end
+end
-	while true do
+while true do
-		local start, stop = find_list(array, sublist)
+local start, stop = find_list(array, sublist)
-		if not start then break end
+if not start then break end
-		local text = process_list(splice(array, start, stop))
+local text = process_list(splice(array, start, stop))
-		local info = {
+local info = {
-			line = text,
+line = text,
-			type = "raw",
+type = "raw",
-			html = text
+html = text
-		}
+}
-		array = splice(array, start, stop, {info})
+array = splice(array, start, stop, {info})
-	end
+end
-	-- Convert any remaining list items to normal
+-- Convert any remaining list items to normal
-	for _,line in ipairs(array) do
+for _,line in ipairs(array) do
-		if line.type == "list_item" then line.type = "normal" end
+if line.type == "list_item" then line.type = "normal" end
-	end
+end
-	return array
+return array
 end
 -- Find and convert blockquote markers.
-function blockquotes(lines)
+local function blockquotes(lines)
-	local function find_blockquote(lines)
+local function find_blockquote(lines)
-		local start
+local start
-		for i,line in ipairs(lines) do
+for i,line in ipairs(lines) do
-			if line.type == "blockquote" then
+if line.type == "blockquote" then
-				start = i
+start = i
-				break
+break
-			end
+end
-		end
+end
-		if not start then return nil end
+if not start then return nil end
-		local stop = #lines
+local stop = #lines
-		for i = start+1, #lines do
+for i = start+1, #lines do
-			if lines[i].type == "blank" or lines[i].type == "blockquote" then
+if lines[i].type == "blank" or lines[i].type == "blockquote" then
-			elseif lines[i].type == "normal" then
+elseif lines[i].type == "normal" then
-				if lines[i-1].type == "blank" then stop = i-1 break end
+if lines[i-1].type == "blank" then stop = i-1 break end
-			else
+else
-				stop = i-1 break
+stop = i-1 break
-			end
+end
-		end
+end
-		while lines[stop].type == "blank" do stop = stop - 1 end
+while lines[stop].type == "blank" do stop = stop - 1 end
-		return start, stop
+return start, stop
-	end
+end
-	local function process_blockquote(lines)
+local function process_blockquote(lines)
-		local raw = lines[1].text
+local raw = lines[1].text
-		for i = 2,#lines do
+for i = 2,#lines do
-			raw = raw .. "\n" .. lines[i].text
+raw = raw .. "\n" .. lines[i].text
-		end
+end
-		local bt = block_transform(raw)
+local bt = block_transform(raw)
-		if not bt:find("<pre>") then bt = indent(bt) end
+if not bt:find("<pre>") then bt = indent(bt) end
-		return "<blockquote>\n    " .. bt ..
+return "<blockquote>\n    " .. bt ..
-			"\n</blockquote>"
+"\n</blockquote>"
-	end
+end
-	while true do
+while true do
-		local start, stop = find_blockquote(lines)
+local start, stop = find_blockquote(lines)
-		if not start then break end
+if not start then break end
-		local text = process_blockquote(splice(lines, start, stop))
+local text = process_blockquote(splice(lines, start, stop))
-		local info = {
+local info = {
-			line = text,
+line = text,
-			type = "raw",
+type = "raw",
-			html = text
+html = text
-		}
+}
-		lines = splice(lines, start, stop, {info})
+lines = splice(lines, start, stop, {info})
-	end
+end
-	return lines
+return lines
 end
 -- Find and convert codeblocks.
-function codeblocks(lines)
+local function codeblocks(lines)
-	local function find_codeblock(lines)
+local function find_codeblock(lines)
-		local start
+local start
-		for i,line in ipairs(lines) do
+for i,line in ipairs(lines) do
-			if line.type == "indented" then start = i break end
+if line.type == "indented" then start = i break end
-		end
+end
-		if not start then return nil end
+if not start then return nil end
-		local stop = #lines
+local stop = #lines
-		for i = start+1, #lines do
+for i = start+1, #lines do
-			if lines[i].type ~= "indented" and lines[i].type ~= "blank" then
+if lines[i].type ~= "indented" and lines[i].type ~= "blank" then
-				stop = i-1
+stop = i-1
-				break
+break
-			end
+end
-		end
+end
-		while lines[stop].type == "blank" do stop = stop - 1 end
+while lines[stop].type == "blank" do stop = stop - 1 end
-		return start, stop
+return start, stop
-	end
+end
-	local function process_codeblock(lines)
+local function process_codeblock(lines)
-		local raw = detab(encode_code(outdent(lines[1].line)))
+local raw = detab(encode_code(outdent(lines[1].line)))
-		for i = 2,#lines do
+for i = 2,#lines do
-			raw = raw .. "\n" .. detab(encode_code(outdent(lines[i].line)))
+raw = raw .. "\n" .. detab(encode_code(outdent(lines[i].line)))
-		end
+end
-		return "<pre><code>" .. raw .. "\n</code></pre>"
+return "<pre><code>" .. raw .. "\n</code></pre>"
-	end
+end
-	while true do
+while true do
-		local start, stop = find_codeblock(lines)
+local start, stop = find_codeblock(lines)
-		if not start then break end
+if not start then break end
-		local text = process_codeblock(splice(lines, start, stop))
+local text = process_codeblock(splice(lines, start, stop))
-		local info = {
+local info = {
-			line = text,
+line = text,
-			type = "raw",
+type = "raw",
-			html = text
+html = text
-		}
+}
-		lines = splice(lines, start, stop, {info})
+lines = splice(lines, start, stop, {info})
-	end
+end
-	return lines
+return lines
-end
--- Convert lines to html code
-function blocks_to_html(lines, no_paragraphs)
-	local out = {}
-	local i = 1
-	while i <= #lines do
-		local line = lines[i]
-		if line.type == "ruler" then
-			table.insert(out, "<hr/>")
-		elseif line.type == "raw" then
-			table.insert(out, line.html)
-		elseif line.type == "normal" then
-			local s = line.line
-			while i+1 <= #lines and lines[i+1].type == "normal" do
-				i = i + 1
-				s = s .. "\n" .. lines[i].line
-			end
-			if no_paragraphs then
-				table.insert(out, span_transform(s))
-			else
-				table.insert(out, "<p>" .. span_transform(s) .. "</p>")
-			end
-		elseif line.type == "header" then
-			local s = "<h" .. line.level .. ">" .. span_transform(line.text) .. "</h" .. line.level .. ">"
-			table.insert(out, s)
-		else
-			table.insert(out, line.line)
-		end
-		i = i + 1
-	end
-	return out
 end
 -- Perform all the block level transforms
 function block_transform(text, sublist)
-	local lines = split(text)
+local lines = split(text)
-	lines = map(lines, classify)
+lines = map(lines, classify)
-	lines = headers(lines)
+lines = headers(lines)
-	lines = lists(lines, sublist)
+lines = lists(lines, sublist)
-	lines = codeblocks(lines)
+lines = codeblocks(lines)
-	lines = blockquotes(lines)
+lines = blockquotes(lines)
-	lines = blocks_to_html(lines)
+lines = blocks_to_html(lines)
-	local text = table.concat(lines, "\n")
+local text = table.concat(lines, "\n")
-	return text
+return text
-end
--- Debug function for printing a line array to see the result
--- of partial transforms.
-function print_lines(lines)
-	for i, line in ipairs(lines) do
-		print(i, line.type, line.text or line.line)
-	end
 end
 ----------------------------------------------------------------------
 -- Span transform
 ----------------------------------------------------------------------
 -- Functions for transforming the text at the span level.
 -- These characters may need to be escaped because they have a special
 -- meaning in markdown.
-escape_chars = "'\\`*_{}[]()>#+-.!'"
+local escape_chars = "'\\`*_{}[]()>#+-.!'"
-escape_table = {}
+local escape_table = {}
-function init_escape_table()
+local function init_escape_table()
-	escape_table = {}
+escape_table = {}
-	for i = 1,#escape_chars do
+for i = 1,#escape_chars do
-		local c = escape_chars:sub(i,i)
+local c = escape_chars:sub(i,i)
-		escape_table[c] = hash(c)
+escape_table[c] = hash(c)
-	end
+end
 end
 -- Adds a new escape to the escape table.
-function add_escape(text)
+local function add_escape(text)
-	if not escape_table[text] then
+if not escape_table[text] then
-		escape_table[text] = hash(text)
+escape_table[text] = hash(text)
-	end
+end
-	return escape_table[text]
+return escape_table[text]
 end
+-- Encode backspace-escaped characters in the markdown source.
+local function encode_backslash_escapes(t)
+for i=1,escape_chars:len() do
+local c = escape_chars:sub(i,i)
+t = t:gsub("\\%" .. c, escape_table[c])
+end
+return t
+end
 -- Escape characters that should not be disturbed by markdown.
-function escape_special_chars(text)
+local function escape_special_chars(text)
-	local tokens = tokenize_html(text)
+local tokens = tokenize_html(text)
-	local out = ""
+local out = ""
-	for _, token in ipairs(tokens) do
+for _, token in ipairs(tokens) do
-		local t = token.text
+local t = token.text
-		if token.type == "tag" then
+if token.type == "tag" then
-			-- In tags, encode * and _ so they don't conflict with their use in markdown.
+-- In tags, encode * and _ so they don't conflict with their use in markdown.
-			t = t:gsub("%*", escape_table["*"])
+t = t:gsub("%*", escape_table["*"])
-			t = t:gsub("%_", escape_table["_"])
+t = t:gsub("%_", escape_table["_"])
-		else
+else
-			t = encode_backslash_escapes(t)
+t = encode_backslash_escapes(t)
-		end
+end
-		out = out .. t
+out = out .. t
-	end
+end
-	return out
+return out
-end
--- Encode backspace-escaped characters in the markdown source.
-function encode_backslash_escapes(t)
-	for i=1,escape_chars:len() do
-		local c = escape_chars:sub(i,i)
-		t = t:gsub("\\%" .. c, escape_table[c])
-	end
-	return t
 end
 -- Unescape characters that have been encoded.
-function unescape_special_chars(t)
+local function unescape_special_chars(t)
-	local tin = t
+local tin = t
-	for k,v in pairs(escape_table) do
+for k,v in pairs(escape_table) do
-		k = k:gsub("%%", "%%%%")
+k = k:gsub("%%", "%%%%")
-		t = t:gsub(v,k)
+t = t:gsub(v,k)
-	end
+end
-	if t ~= tin then t = unescape_special_chars(t) end
+if t ~= tin then t = unescape_special_chars(t) end
-	return t
+return t
 end
 -- Encode/escape certain characters inside Markdown code runs.
 -- The point is that in code, these characters are literals,
 -- and lose their special Markdown meanings.
 function encode_code(s)
-	s = s:gsub("%&", "&amp;")
+s = s:gsub("%&", "&amp;")
-	s = s:gsub("<", "&lt;")
+s = s:gsub("<", "&lt;")
-	s = s:gsub(">", "&gt;")
+s = s:gsub(">", "&gt;")
-	for k,v in pairs(escape_table) do
+for k,v in pairs(escape_table) do
-		s = s:gsub("%"..k, v)
+s = s:gsub("%"..k, v)
-	end
+end
-	return s
+return s
 end
 -- Handle backtick blocks.
-function code_spans(s)
+local function code_spans(s)
-	s = s:gsub("\\\\", escape_table["\\"])
+s = s:gsub("\\\\", escape_table["\\"])
-	s = s:gsub("\\`", escape_table["`"])
+s = s:gsub("\\`", escape_table["`"])
-	local pos = 1
+local pos = 1
-	while true do
+while true do
-		local start, stop = s:find("`+", pos)
+local start, stop = s:find("`+", pos)
-		if not start then return s end
+if not start then return s end
-		local count = stop - start + 1
+local count = stop - start + 1
-		-- Find a matching numbert of backticks
+-- Find a matching numbert of backticks
-		local estart, estop = s:find(string.rep("`", count), stop+1)
+local estart, estop = s:find(string.rep("`", count), stop+1)
-		local brstart = s:find("\n", stop+1)
+local brstart = s:find("\n", stop+1)
-		if estart and (not brstart or estart < brstart) then
+if estart and (not brstart or estart < brstart) then
-			local code = s:sub(stop+1, estart-1)
+local code = s:sub(stop+1, estart-1)
-			code = code:gsub("^[ \t]+", "")
+code = code:gsub("^[ \t]+", "")
-			code = code:gsub("[ \t]+$", "")
+code = code:gsub("[ \t]+$", "")
-			code = code:gsub(escape_table["\\"], escape_table["\\"] .. escape_table["\\"])
+code = code:gsub(escape_table["\\"], escape_table["\\"] .. escape_table["\\"])
-			code = code:gsub(escape_table["`"], escape_table["\\"] .. escape_table["`"])
+code = code:gsub(escape_table["`"], escape_table["\\"] .. escape_table["`"])
-			code = "<code>" .. encode_code(code) .. "</code>"
+code = "<code>" .. encode_code(code) .. "</code>"
-			code = add_escape(code)
+code = add_escape(code)
-			s = s:sub(1, start-1) .. code .. s:sub(estop+1)
+s = s:sub(1, start-1) .. code .. s:sub(estop+1)
-			pos = start + code:len()
+pos = start + code:len()
-		else
+else
-			pos = stop + 1
+pos = stop + 1
-		end
+end
-	end
+end
-	return s
+return s
 end
 -- Encode alt text... enodes &, and ".
-function encode_alt(s)
+local function encode_alt(s)
-	if not s then return s end
+if not s then return s end
-	s = s:gsub('&', '&amp;')
+s = s:gsub('&', '&amp;')
-	s = s:gsub('"', '&quot;')
+s = s:gsub('"', '&quot;')
-	s = s:gsub('<', '&lt;')
+s = s:gsub('<', '&lt;')
-	return s
+return s
 end
+-- Forward declaration for link_db as returned by strip_link_definitions.
+local link_database
 -- Handle image references
-function images(text)
+local function images(text)
-	local function reference_link(alt, id)
+local function reference_link(alt, id)
-		alt = encode_alt(alt:match("%b[]"):sub(2,-2))
+alt = encode_alt(alt:match("%b[]"):sub(2,-2))
-		id = id:match("%[(.*)%]"):lower()
+id = id:match("%[(.*)%]"):lower()
-		if id == "" then id = text:lower() end
+if id == "" then id = text:lower() end
-		link_database[id] = link_database[id] or {}
+link_database[id] = link_database[id] or {}
-		if not link_database[id].url then return nil end
+if not link_database[id].url then return nil end
-		local url = link_database[id].url or id
+local url = link_database[id].url or id
-		url = encode_alt(url)
+url = encode_alt(url)
-		local title = encode_alt(link_database[id].title)
+local title = encode_alt(link_database[id].title)
-		if title then title = " title=\"" .. title .. "\"" else title = "" end
+if title then title = " title=\"" .. title .. "\"" else title = "" end
-		return add_escape ('<img src="' .. url .. '" alt="' .. alt .. '"' .. title .. "/>")
+return add_escape ('<img src="' .. url .. '" alt="' .. alt .. '"' .. title .. "/>")
-	end
+end
-	local function inline_link(alt, link)
+local function inline_link(alt, link)
-		alt = encode_alt(alt:match("%b[]"):sub(2,-2))
+alt = encode_alt(alt:match("%b[]"):sub(2,-2))
-		local url, title = link:match("%(<?(.-)>?[ \t]*['\"](.+)['\"]")
+local url, title = link:match("%(<?(.-)>?[ \t]*['\"](.+)['\"]")
-		url = url or link:match("%(<?(.-)>?%)")
+url = url or link:match("%(<?(.-)>?%)")
-		url = encode_alt(url)
+url = encode_alt(url)
-		title = encode_alt(title)
+title = encode_alt(title)
-		if title then
+if title then
-			return add_escape('<img src="' .. url .. '" alt="' .. alt .. '" title="' .. title .. '"/>')
+return add_escape('<img src="' .. url .. '" alt="' .. alt .. '" title="' .. title .. '"/>')
-		else
+else
-			return add_escape('<img src="' .. url .. '" alt="' .. alt .. '"/>')
+return add_escape('<img src="' .. url .. '" alt="' .. alt .. '"/>')
-		end
+end
-	end
+end
-	text = text:gsub("!(%b[])[ \t]*\n?[ \t]*(%b[])", reference_link)
+text = text:gsub("!(%b[])[ \t]*\n?[ \t]*(%b[])", reference_link)
-	text = text:gsub("!(%b[])(%b())", inline_link)
+text = text:gsub("!(%b[])(%b())", inline_link)
-	return text
+return text
 end
 -- Handle anchor references
-function anchors(text)
+local function anchors(text)
-	local function reference_link(text, id)
+local function reference_link(text, id)
-		text = text:match("%b[]"):sub(2,-2)
+text = text:match("%b[]"):sub(2,-2)
-		id = id:match("%b[]"):sub(2,-2):lower()
+id = id:match("%b[]"):sub(2,-2):lower()
-		if id == "" then id = text:lower() end
+if id == "" then id = text:lower() end
-		link_database[id] = link_database[id] or {}
+link_database[id] = link_database[id] or {}
-		if not link_database[id].url then return nil end
+if not link_database[id].url then return nil end
-		local url = link_database[id].url or id
+local url = link_database[id].url or id
-		url = encode_alt(url)
+url = encode_alt(url)
-		local title = encode_alt(link_database[id].title)
+local title = encode_alt(link_database[id].title)
-		if title then title = " title=\"" .. title .. "\"" else title = "" end
+if title then title = " title=\"" .. title .. "\"" else title = "" end
-		return add_escape("<a href=\"" .. url .. "\"" .. title .. ">") .. text .. add_escape("</a>")
+return add_escape("<a href=\"" .. url .. "\"" .. title .. ">") .. text .. add_escape("</a>")
-	end
+end
-	local function inline_link(text, link)
+local function inline_link(text, link)
-		text = text:match("%b[]"):sub(2,-2)
+text = text:match("%b[]"):sub(2,-2)
-		local url, title = link:match("%(<?(.-)>?[ \t]*['\"](.+)['\"]")
+local url, title = link:match("%(<?(.-)>?[ \t]*['\"](.+)['\"]")
-		title = encode_alt(title)
+title = encode_alt(title)
-		url  = url or  link:match("%(<?(.-)>?%)") or ""
+url  = url or  link:match("%(<?(.-)>?%)") or ""
-		url = encode_alt(url)
+url = encode_alt(url)
-		if title then
+if title then
-			return add_escape("<a href=\"" .. url .. "\" title=\"" .. title .. "\">") .. text .. "</a>"
+return add_escape("<a href=\"" .. url .. "\" title=\"" .. title .. "\">") .. text .. "</a>"
-		else
+else
-			return add_escape("<a href=\"" .. url .. "\">") .. text .. add_escape("</a>")
+return add_escape("<a href=\"" .. url .. "\">") .. text .. add_escape("</a>")
-		end
+end
-	end
+end
-	text = text:gsub("(%b[])[ \t]*\n?[ \t]*(%b[])", reference_link)
+text = text:gsub("(%b[])[ \t]*\n?[ \t]*(%b[])", reference_link)
-	text = text:gsub("(%b[])(%b())", inline_link)
+text = text:gsub("(%b[])(%b())", inline_link)
-	return text
+return text
 end
 -- Handle auto links, i.e. <http://www.google.com/>.
-function auto_links(text)
+local function auto_links(text)
-	local function link(s)
+local function link(s)
-		return add_escape("<a href=\"" .. s .. "\">") .. s .. "</a>"
+return add_escape("<a href=\"" .. s .. "\">") .. s .. "</a>"
-	end
+end
-	-- Encode chars as a mix of dec and hex entitites to (perhaps) fool
+-- Encode chars as a mix of dec and hex entitites to (perhaps) fool
-	-- spambots.
+-- spambots.
-	local function encode_email_address(s)
+local function encode_email_address(s)
-		-- Use a deterministic encoding to make unit testing possible.
+-- Use a deterministic encoding to make unit testing possible.
-		-- Code 45% hex, 45% dec, 10% plain.
+-- Code 45% hex, 45% dec, 10% plain.
-		local hex = {code = function(c) return "&#x" .. string.format("%x", c:byte()) .. ";" end, count = 1, rate = 0.45}
+local hex = {code = function(c) return "&#x" .. string.format("%x", c:byte()) .. ";" end, count = 1, rate = 0.45}
-		local dec = {code = function(c) return "&#" .. c:byte() .. ";" end, count = 0, rate = 0.45}
+local dec = {code = function(c) return "&#" .. c:byte() .. ";" end, count = 0, rate = 0.45}
-		local plain = {code = function(c) return c end, count = 0, rate = 0.1}
+local plain = {code = function(c) return c end, count = 0, rate = 0.1}
-		local codes = {hex, dec, plain}
+local codes = {hex, dec, plain}
-		local function swap(t,k1,k2) local temp = t[k2] t[k2] = t[k1] t[k1] = temp end
+local function swap(t,k1,k2) local temp = t[k2] t[k2] = t[k1] t[k1] = temp end
-		local out = ""
+local out = ""
-		for i = 1,s:len() do
+for i = 1,s:len() do
-			for _,code in ipairs(codes) do code.count = code.count + code.rate end
+for _,code in ipairs(codes) do code.count = code.count + code.rate end
-			if codes[1].count < codes[2].count then swap(codes,1,2) end
+if codes[1].count < codes[2].count then swap(codes,1,2) end
-			if codes[2].count < codes[3].count then swap(codes,2,3) end
+if codes[2].count < codes[3].count then swap(codes,2,3) end
-			if codes[1].count < codes[2].count then swap(codes,1,2) end
+if codes[1].count < codes[2].count then swap(codes,1,2) end
-			local code = codes[1]
+local code = codes[1]
-			local c = s:sub(i,i)
+local c = s:sub(i,i)
-			-- Force encoding of "@" to make email address more invisible.
+-- Force encoding of "@" to make email address more invisible.
-			if c == "@" and code == plain then code = codes[2] end
+if c == "@" and code == plain then code = codes[2] end
-			out = out .. code.code(c)
+out = out .. code.code(c)
-			code.count = code.count - 1
+code.count = code.count - 1
-		end
+end
-		return out
+return out
-	end
+end
-	local function mail(s)
+local function mail(s)
-		s = unescape_special_chars(s)
+s = unescape_special_chars(s)
-		local address = encode_email_address("mailto:" .. s)
+local address = encode_email_address("mailto:" .. s)
-		local text = encode_email_address(s)
+local text = encode_email_address(s)
-		return add_escape("<a href=\"" .. address .. "\">") .. text .. "</a>"
+return add_escape("<a href=\"" .. address .. "\">") .. text .. "</a>"
-	end
+end
-	-- links
+-- links
-	text = text:gsub("<(https?:[^'\">%s]+)>", link)
+text = text:gsub("<(https?:[^'\">%s]+)>", link)
-	text = text:gsub("<(ftp:[^'\">%s]+)>", link)
+text = text:gsub("<(ftp:[^'\">%s]+)>", link)
-	-- mail
+-- mail
-	text = text:gsub("<mailto:([^'\">%s]+)>", mail)
+text = text:gsub("<mailto:([^'\">%s]+)>", mail)
-	text = text:gsub("<([-.%w]+%@[-.%w]+)>", mail)
+text = text:gsub("<([-.%w]+%@[-.%w]+)>", mail)
-	return text
+return text
 end
 -- Encode free standing amps (&) and angles (<)... note that this does not
 -- encode free >.
-function amps_and_angles(s)
+local function amps_and_angles(s)
-	-- encode amps not part of &..; expression
+-- encode amps not part of &..; expression
-	local pos = 1
+local pos = 1
-	while true do
+while true do
-		local amp = s:find("&", pos)
+local amp = s:find("&", pos)
-		if not amp then break end
+if not amp then break end
-		local semi = s:find(";", amp+1)
+local semi = s:find(";", amp+1)
-		local stop = s:find("[ \t\n&]", amp+1)
+local stop = s:find("[ \t\n&]", amp+1)
-		if not semi or (stop and stop < semi) or (semi - amp) > 15 then
+if not semi or (stop and stop < semi) or (semi - amp) > 15 then
-			s = s:sub(1,amp-1) .. "&amp;" .. s:sub(amp+1)
+s = s:sub(1,amp-1) .. "&amp;" .. s:sub(amp+1)
-			pos = amp+1
+pos = amp+1
-		else
+else
-			pos = amp+1
+pos = amp+1
-		end
+end
-	end
+end
-	-- encode naked <'s
+-- encode naked <'s
-	s = s:gsub("<([^a-zA-Z/?$!])", "&lt;%1")
+s = s:gsub("<([^a-zA-Z/?$!])", "&lt;%1")
-	s = s:gsub("<$", "&lt;")
+s = s:gsub("<$", "&lt;")
-	-- what about >, nothing done in the original markdown source to handle them
+-- what about >, nothing done in the original markdown source to handle them
-	return s
+return s
 end
 -- Handles emphasis markers (* and _) in the text.
-function emphasis(text)
+local function emphasis(text)
-	for _, s in ipairs {"%*%*", "%_%_"} do
+for _, s in ipairs {"%*%*", "%_%_"} do
-		text = text:gsub(s .. "([^%s][%*%_]?)" .. s, "<strong>%1</strong>")
+text = text:gsub(s .. "([^%s][%*%_]?)" .. s, "<strong>%1</strong>")
-		text = text:gsub(s .. "([^%s][^<>]-[^%s][%*%_]?)" .. s, "<strong>%1</strong>")
+text = text:gsub(s .. "([^%s][^<>]-[^%s][%*%_]?)" .. s, "<strong>%1</strong>")
-	end
+end
-	for _, s in ipairs {"%*", "%_"} do
+for _, s in ipairs {"%*", "%_"} do
-		text = text:gsub(s .. "([^%s_])" .. s, "<em>%1</em>")
+text = text:gsub(s .. "([^%s_])" .. s, "<em>%1</em>")
-		text = text:gsub(s .. "(<strong>[^%s_]</strong>)" .. s, "<em>%1</em>")
+text = text:gsub(s .. "(<strong>[^%s_]</strong>)" .. s, "<em>%1</em>")
-		text = text:gsub(s .. "([^%s_][^<>_]-[^%s_])" .. s, "<em>%1</em>")
+text = text:gsub(s .. "([^%s_][^<>_]-[^%s_])" .. s, "<em>%1</em>")
-		text = text:gsub(s .. "([^<>_]-<strong>[^<>_]-</strong>[^<>_]-)" .. s, "<em>%1</em>")
+text = text:gsub(s .. "([^<>_]-<strong>[^<>_]-</strong>[^<>_]-)" .. s, "<em>%1</em>")
-	end
+end
-	return text
+return text
 end
 -- Handles line break markers in the text.
-function line_breaks(text)
+local function line_breaks(text)
-	return text:gsub("  +\n", " <br/>\n")
+return text:gsub("  +\n", " <br/>\n")
 end
 -- Perform all span level transforms.
 function span_transform(text)
-	text = code_spans(text)
+text = code_spans(text)
-	text = escape_special_chars(text)
+text = escape_special_chars(text)
-	text = images(text)
+text = images(text)
-	text = anchors(text)
+text = anchors(text)
-	text = auto_links(text)
+text = auto_links(text)
-	text = amps_and_angles(text)
+text = amps_and_angles(text)
-	text = emphasis(text)
+text = emphasis(text)
-	text = line_breaks(text)
+text = line_breaks(text)
-	return text
+return text
 end
 ----------------------------------------------------------------------
 -- Markdown
 ----------------------------------------------------------------------
 -- Cleanup the text by normalizing some possible variations to make further
 -- processing easier.
-function cleanup(text)
+local function cleanup(text)
-	-- Standardize line endings
+-- Standardize line endings
-	text = text:gsub("\r\n", "\n")  -- DOS to UNIX
+text = text:gsub("\r\n", "\n")  -- DOS to UNIX
-	text = text:gsub("\r", "\n")    -- Mac to UNIX
+text = text:gsub("\r", "\n")    -- Mac to UNIX
-	-- Convert all tabs to spaces
+-- Convert all tabs to spaces
-	text = detab(text)
+text = detab(text)
-	-- Strip lines with only spaces and tabs
+-- Strip lines with only spaces and tabs
-	while true do
+while true do
-		local subs
+local subs
-		text, subs = text:gsub("\n[ \t]+\n", "\n\n")
+text, subs = text:gsub("\n[ \t]+\n", "\n\n")
-		if subs == 0 then break end
+if subs == 0 then break end
-	end
+end
-	return "\n" .. text .. "\n"
+return "\n" .. text .. "\n"
 end
 -- Strips link definitions from the text and stores the data in a lookup table.
-function strip_link_definitions(text)
+local function strip_link_definitions(text)
-	local linkdb = {}
+local linkdb = {}
-	local function link_def(id, url, title)
+local function link_def(id, url, title)
-		id = id:match("%[(.+)%]"):lower()
+id = id:match("%[(.+)%]"):lower()
-		linkdb[id] = linkdb[id] or {}
+linkdb[id] = linkdb[id] or {}
-		linkdb[id].url = url or linkdb[id].url
+linkdb[id].url = url or linkdb[id].url
-		linkdb[id].title = title or linkdb[id].title
+linkdb[id].title = title or linkdb[id].title
-		return ""
+return ""
-	end
+end
-	local def_no_title = "\n ? ? ?(%b[]):[ \t]*\n?[ \t]*<?([^%s>]+)>?[ \t]*"
+local def_no_title = "\n ? ? ?(%b[]):[ \t]*\n?[ \t]*<?([^%s>]+)>?[ \t]*"
-	local def_title1 = def_no_title .. "[ \t]+\n?[ \t]*[\"'(]([^\n]+)[\"')][ \t]*"
+local def_title1 = def_no_title .. "[ \t]+\n?[ \t]*[\"'(]([^\n]+)[\"')][ \t]*"
-	local def_title2 = def_no_title .. "[ \t]*\n[ \t]*[\"'(]([^\n]+)[\"')][ \t]*"
+local def_title2 = def_no_title .. "[ \t]*\n[ \t]*[\"'(]([^\n]+)[\"')][ \t]*"
-	local def_title3 = def_no_title .. "[ \t]*\n?[ \t]+[\"'(]([^\n]+)[\"')][ \t]*"
+local def_title3 = def_no_title .. "[ \t]*\n?[ \t]+[\"'(]([^\n]+)[\"')][ \t]*"
-	text = text:gsub(def_title1, link_def)
+text = text:gsub(def_title1, link_def)
-	text = text:gsub(def_title2, link_def)
+text = text:gsub(def_title2, link_def)
-	text = text:gsub(def_title3, link_def)
+text = text:gsub(def_title3, link_def)
-	text = text:gsub(def_no_title, link_def)
+text = text:gsub(def_no_title, link_def)
-	return text, linkdb
+return text, linkdb
 end
-link_database = {}
 -- Main markdown processing function
-function markdown(text)
+local function markdown(text)
-	init_hash(text)
+init_hash(text)
-	init_escape_table()
+init_escape_table()
-	text = cleanup(text)
+text = cleanup(text)
-	text = protect(text)
+text = protect(text)
-	text, link_database = strip_link_definitions(text)
+text, link_database = strip_link_definitions(text)
-	text = block_transform(text)
+text = block_transform(text)
-	text = unescape_special_chars(text)
+text = unescape_special_chars(text)
-	return text
+return text
 end
 ----------------------------------------------------------------------
 -- End of module
 ----------------------------------------------------------------------
-setfenv(1, _G)
+-- For compatibility, set markdown function as a global
-M.lock(M)
+_G.markdown = markdown
--- Expose markdown function to the world
-markdown = M.markdown
 -- Class for parsing command-line options
 local OptionParser = {}
 OptionParser.__index = OptionParser
 -- Creates a new option parser
 function OptionParser:new()
-	local o = {short = {}, long = {}}
+local o = {short = {}, long = {}}
-	setmetatable(o, self)
+setmetatable(o, self)
-	return o
+return o
 end
 -- Calls f() whenever a flag with specified short and long name is encountered
 function OptionParser:flag(short, long, f)
-	local info = {type = "flag", f = f}
+local info = {type = "flag", f = f}
-	if short then self.short[short] = info end
+if short then self.short[short] = info end
-	if long then self.long[long] = info end
+if long then self.long[long] = info end
 end
 -- Calls f(param) whenever a parameter flag with specified short and long name is encountered
 function OptionParser:param(short, long, f)
-	local info = {type = "param", f = f}
+local info = {type = "param", f = f}
-	if short then self.short[short] = info end
+if short then self.short[short] = info end
-	if long then self.long[long] = info end
+if long then self.long[long] = info end
 end
 -- Calls f(v) for each non-flag argument
 function OptionParser:arg(f)
-	self.arg = f
+self.arg = f
 end
 -- Runs the option parser for the specified set of arguments. Returns true if all arguments
 -- where successfully parsed and false otherwise.
 function OptionParser:run(args)
-	local pos = 1
+local pos = 1
-	while pos <= #args do
+while pos <= #args do
-		local arg = args[pos]
+local arg = args[pos]
-		if arg == "--" then
+if arg == "--" then
-			for i=pos+1,#args do
+for i=pos+1,#args do
-				if self.arg then self.arg(args[i]) end
+if self.arg then self.arg(args[i]) end
-				return true
+return true
-			end
+end
-		end
+end
-		if arg:match("^%-%-") then
+if arg:match("^%-%-") then
-			local info = self.long[arg:sub(3)]
+local info = self.long[arg:sub(3)]
-			if not info then print("Unknown flag: " .. arg) return false end
+if not info then print("Unknown flag: " .. arg) return false end
-			if info.type == "flag" then
+if info.type == "flag" then
-				info.f()
+info.f()
-				pos = pos + 1
+pos = pos + 1
-			else
+else
-				param = args[pos+1]
+local param = args[pos+1]
-				if not param then print("No parameter for flag: " .. arg) return false end
+if not param then print("No parameter for flag: " .. arg) return false end
-				info.f(param)
+info.f(param)
-				pos = pos+2
+pos = pos+2
-			end
+end
-		elseif arg:match("^%-") then
+elseif arg:match("^%-") then
-			for i=2,arg:len() do
+for i=2,arg:len() do
-				local c = arg:sub(i,i)
+local c = arg:sub(i,i)
-				local info = self.short[c]
+local info = self.short[c]
-				if not info then print("Unknown flag: -" .. c) return false end
+if not info then print("Unknown flag: -" .. c) return false end
-				if info.type == "flag" then
+if info.type == "flag" then
-					info.f()
+info.f()
-				else
+else
-					if i == arg:len() then
+if i == arg:len() then
-						param = args[pos+1]
+local param = args[pos+1]
-						if not param then print("No parameter for flag: -" .. c) return false end
+if not param then print("No parameter for flag: -" .. c) return false end
-						info.f(param)
+info.f(param)
-						pos = pos + 1
+pos = pos + 1
-					else
+else
-						param = arg:sub(i+1)
+local param = arg:sub(i+1)
-						info.f(param)
+info.f(param)
-					end
+end
-					break
+break
-				end
+end
-			end
+end
-			pos = pos + 1
+pos = pos + 1
-		else
+else
-			if self.arg then self.arg(arg) end
+if self.arg then self.arg(arg) end
-			pos = pos + 1
+pos = pos + 1
-		end
+end
-	end
+end
-	return true
+return true
+end
+local function read_file(path, descr)
+local file = io.open(path) or error("Could not open " .. descr .. " file: " .. path)
+local contents = file:read("*a") or error("Could not read " .. descr .. " from " .. path)
+file:close()
+return contents
 end
 -- Handles the case when markdown is run from the command line
 local function run_command_line(arg)
-	-- Generate output for input s given options
+-- Generate output for input s given options
-	local function run(s, options)
+local function run(s, options)
-		s = markdown(s)
+s = markdown(s)
-		if not options.wrap_header then return s end
+if not options.wrap_header then return s end
-		local header = ""
+local header
-		if options.header then
+if options.header then
-			local f = io.open(options.header) or error("Could not open file: " .. options.header)
+header = read_file(options.header, "header")
-			header = f:read("*a")
+else
-			f:close()
+header = [[
-		else
-			header = [[
 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
 <html>
 <head>
-	<meta http-equiv="content-type" content="text/html; charset=CHARSET" />
+<meta http-equiv="content-type" content="text/html; charset=CHARSET" />
-	<title>TITLE</title>
+<title>TITLE</title>
-	<link rel="stylesheet" type="text/css" href="STYLESHEET" />
+<link rel="stylesheet" type="text/css" href="STYLESHEET" />
 </head>
 <body>
 ]]
-			local title = options.title or s:match("<h1>(.-)</h1>") or s:match("<h2>(.-)</h2>") or
+local title = options.title or s:match("<h1>(.-)</h1>") or s:match("<h2>(.-)</h2>") or
-				s:match("<h3>(.-)</h3>") or "Untitled"
+s:match("<h3>(.-)</h3>") or "Untitled"
-			header = header:gsub("TITLE", title)
+header = header:gsub("TITLE", title)
-			if options.inline_style then
+if options.inline_style then
-				local style = ""
+local style = read_file(options.stylesheet, "style sheet")
-				local f = io.open(options.stylesheet)
+header = header:gsub('<link rel="stylesheet" type="text/css" href="STYLESHEET" />',
-				if f then
+"<style type=\"text/css\"><!--\n" .. style .. "\n--></style>")
-					style = f:read("*a") f:close()
+else
-				else
+header = header:gsub("STYLESHEET", options.stylesheet)
-					error("Could not include style sheet " .. options.stylesheet .. ": File not found")
+end
-				end
+header = header:gsub("CHARSET", options.charset)
-				header = header:gsub('<link rel="stylesheet" type="text/css" href="STYLESHEET" />',
+end
-					"<style type=\"text/css\"><!--\n" .. style .. "\n--></style>")
+local footer = "</body></html>"
-			else
+if options.footer then
-				header = header:gsub("STYLESHEET", options.stylesheet)
+footer = read_file(options.footer, "footer")
-			end
+end
-			header = header:gsub("CHARSET", options.charset)
+return header .. s .. footer
-		end
+end
-		local footer = "</body></html>"
-		if options.footer then
+-- Generate output path name from input path name given options.
-			local f = io.open(options.footer) or error("Could not open file: " .. options.footer)
+local function outpath(path, options)
-			footer = f:read("*a")
+if options.append then return path .. ".html" end
-			f:close()
+local m = path:match("^(.+%.html)[^/\\]+$") if m then return m end
-		end
+m = path:match("^(.+%.)[^/\\]*$") if m and path ~= m .. "html" then return m .. "html" end
-		return header .. s .. footer
+return path .. ".html"
-	end
+end
-	-- Generate output path name from input path name given options.
+-- Default commandline options
-	local function outpath(path, options)
+local options = {
-		if options.append then return path .. ".html" end
+wrap_header = true,
-		local m = path:match("^(.+%.html)[^/\\]+$") if m then return m end
+header = nil,
-		m = path:match("^(.+%.)[^/\\]*$") if m and path ~= m .. "html" then return m .. "html" end
+footer = nil,
-		return path .. ".html"
+charset = "utf-8",
-	end
+title = nil,
+stylesheet = "default.css",
-	-- Default commandline options
+inline_style = false
-	local options = {
+}
-		wrap_header = true,
+local help = [[
-		header = nil,
-		footer = nil,
-		charset = "utf-8",
-		title = nil,
-		stylesheet = "default.css",
-		inline_style = false
-	}
-	local help = [[
 Usage: markdown.lua [OPTION] [FILE]
 Runs the markdown text markup to HTML converter on each file specified on the
 command line. If no files are specified, runs on standard input.
 No header:
 -f, --footer FILE    Use content of FILE for footer.
 Generated header:
 -c, --charset SET    Specifies charset (default utf-8).
 -i, --title TITLE    Specifies title (default from first <h1> tag).
 -s, --style STYLE    Specifies style sheet file (default default.css).
-	-l, --inline-style   Include the style sheet file inline in the header.
+-l, --inline-style   Include the style sheet file inline in the header.
 Generated files:
 -a, --append         Append .html extension (instead of replacing).
 Other options:
 -h, --help           Print this help text.
 -t, --test           Run the unit tests.
 ]]
-	local run_stdin = true
+local run_stdin = true
-	local op = OptionParser:new()
+local op = OptionParser:new()
-	op:flag("n", "no-wrap", function () options.wrap_header = false end)
+op:flag("n", "no-wrap", function () options.wrap_header = false end)
-	op:param("e", "header", function (x) options.header = x end)
+op:param("e", "header", function (x) options.header = x end)
-	op:param("f", "footer", function (x) options.footer = x end)
+op:param("f", "footer", function (x) options.footer = x end)
-	op:param("c", "charset", function (x) options.charset = x end)
+op:param("c", "charset", function (x) options.charset = x end)
-	op:param("i", "title", function(x) options.title = x end)
+op:param("i", "title", function(x) options.title = x end)
-	op:param("s", "style", function(x) options.stylesheet = x end)
+op:param("s", "style", function(x) options.stylesheet = x end)
-	op:flag("l", "inline-style", function(x) options.inline_style = true end)
+op:flag("l", "inline-style", function() options.inline_style = true end)
-	op:flag("a", "append", function() options.append = true end)
+op:flag("a", "append", function() options.append = true end)
-	op:flag("t", "test", function()
+op:flag("t", "test", function()
-		local n = arg[0]:gsub("markdown.lua", "markdown-tests.lua")
+local n = arg[0]:gsub("markdown%.lua", "markdown-tests.lua")
-		local f = io.open(n)
+local f = io.open(n)
-		if f then
+if f then
-			f:close() dofile(n)
+f:close()
-		else
+package.loaded.markdown = markdown
-			error("Cannot find markdown-tests.lua")
+dofile(n)
-		end
+else
-		run_stdin = false
+error("Cannot find markdown-tests.lua")
-	end)
+end
-	op:flag("h", "help", function() print(help) run_stdin = false end)
+run_stdin = false
-	op:arg(function(path)
+end)
-			local file = io.open(path) or error("Could not open file: " .. path)
+op:flag("h", "help", function() print(help) run_stdin = false end)
-			local s = file:read("*a")
+op:arg(function(path)
-			file:close()
+local s = read_file(path, "input")
-			s = run(s, options)
+s = run(s, options)
-			file = io.open(outpath(path, options), "w") or error("Could not open output file: " .. outpath(path, options))
+local file = io.open(outpath(path, options), "w") or error("Could not open output file: " .. outpath(path, options))
-			file:write(s)
+file:write(s)
-			file:close()
+file:close()
-			run_stdin = false
+run_stdin = false
-		end
+end
-	)
+)
-	if not op:run(arg) then
+if not op:run(arg) then
-		print(help)
+print(help)
-		run_stdin = false
+run_stdin = false
-	end
+end
-	if run_stdin then
+if run_stdin then
-		local s = io.read("*a")
+local s = io.read("*a")
-		s = run(s, options)
+s = run(s, options)
-		io.write(s)
+io.write(s)
-	end
+end
 end
 -- If we are being run from the command-line, act accordingly
 if arg and arg[0]:find("markdown%.lua$") then
-	run_command_line(arg)
+run_command_line(arg)
 else
-	return markdown
+return markdown
 end

Mercurial > repos > lgen / file comparison

comparison: markdown.lua

markdown.lua