Module:Escape

Usage

This module is designed as an way to escape strings in a customized and efficient manner. It works by replacing characters that are preceded by your escape char (or phrase) There are two ways to call this module:

From another module:

Template:Codett
esc:char(Template:Green)
Template:Codett(Template:Green)
Template:Green
Template:Codett

From a template:

{{#invoke:Escape|main|mode=Template:Green|char=Template:Green|Template:Green}}

In a template, the most useful function is <syntaxhighlight lang="text" class="" style="" inline="1">kill</syntaxhighlight>.

This module is primarily intended to be used by other modules. However all functions can be called in template space using |mode=the function you want to call followed by arguments.

All module functions (i.e. any func. other than <syntaxhighlight lang="text" class="" style="" inline="1">main()</syntaxhighlight>) should be called using a colon (Template:Char), e.g. <syntaxhighlight lang="lua" class="" style="" inline="1">esc:char('%')</syntaxhighlight> or <syntaxhighlight lang="lua" class="" style="" inline="1">esc:kill{'{{example|\\}}}', '}'} == '{{example|}'</syntaxhighlight>

Template:TOC tab	This function takes only one argument: A string. All characters in this string which are preceded by the sequence set by <syntaxhighlight lang="text" class="" style="" inline="1">escape:char()</syntaxhighlight> will be replaced with placeholders that can be converted back into that char by escape:undo()
Template:TOC tab	Takes two arguments: The string that may contain placeholders set by <syntaxhighlight lang="text" class="" style="" inline="1">escape:text()</syntaxhighlight> Optional, a char to be placed in front of any characters that have been de-escaped. (i.e. if you need to re-escape those string with a different char)
Template:TOC tab	This is basically equivalent to calling <syntaxhighlight lang="text" class="" style="" inline="1">string.gsub()</syntaxhighlight> on the string returned by <syntaxhighlight lang="text" class="" style="" inline="1">escape:text()</syntaxhighlight> and feeding that result into <syntaxhighlight lang="text" class="" style="" inline="1">escape:undo()</syntaxhighlight> in a single step. Takes three arguments: A string A sequence of characters to be removed from that string. (May use a <syntaxhighlight lang="text" class="" style="" inline="1">string.gsub()</syntaxhighlight> pattern) Optional, a char to be placed in front of any characters that have been de-escaped.
Template:TOC tab	This function's primary use is to initialize the patterns to scan a string for an escape/escaped sequence. It takes two arguments, the first being the escape character and the second being a table of arguments (optional). By default, this module will escape the <syntaxhighlight lang="text" class="" style="" inline="1">\</syntaxhighlight> char. To escape the <syntaxhighlight lang="text" class="" style="" inline="1">{</syntaxhighlight> char instead, you can do <syntaxhighlight lang="lua" class="" style="" inline="1">require('Module:Escape'):char('{')</syntaxhighlight> (or <syntaxhighlight lang="lua" class="" style="" inline="1">esc:char('{')</syntaxhighlight> (presuming you stored the table returned by this module in the local variable <syntaxhighlight lang="text" class="" style="" inline="1">esc</syntaxhighlight>). When called without the second argument, char() will return a table containing the functions. This allows, for example, <syntaxhighlight lang="lua" class="" style="" inline="1">escape:char('*'):kill('1*23', '%d')</syntaxhighlight> which would return 'Template:Samp'. For the most part, there is very little reason to set |mode= in template space since the patterns it stores are not shared with other invokations of this module. Templates should instead use the |char= if a new escape sequence is desired. Shortcut If provided a second argument that is a table containing a {key = value} pair, such that the key is <syntaxhighlight lang="text" class="" style="" inline="1">text</syntaxhighlight>, <syntaxhighlight lang="text" class="" style="" inline="1">undo</syntaxhighlight>, or <syntaxhighlight lang="text" class="" style="" inline="1">kill</syntaxhighlight> and the value is a table containing the arguments that would have been passed to those functions. For example, <syntaxhighlight lang="lua" class="" style="" inline="1">escape:char('\\', {text = 'string'})</syntaxhighlight> is equivalent to <syntaxhighlight lang="lua" class="" style="" inline="1">escape:char('\\'):text('string')</syntaxhighlight>. Note that if multiple key-value pairs are provided, only one may execute. <syntaxhighlight lang="text" class="" style="" inline="1">kill</syntaxhighlight> is ignored if either <syntaxhighlight lang="text" class="" style="" inline="1">text</syntaxhighlight> or <syntaxhighlight lang="text" class="" style="" inline="1">undo</syntaxhighlight> are present. <syntaxhighlight lang="text" class="" style="" inline="1">undo</syntaxhighlight> is ignored if <syntaxhighlight lang="text" class="" style="" inline="1">text</syntaxhighlight> is present.

Caveats

• When using a multi-character escape sequence, this module only marks it using the byte value of the first character. Thus, <syntaxhighlight lang="lua" class="" style="" inline="1">escape:undo()</syntaxhighlight> will unescape, for example, all characters escaped with <syntaxhighlight lang="text" class="" style="" inline="1">'e'</syntaxhighlight> and <syntaxhighlight lang="text" class="" style="" inline="1">'esc'</syntaxhighlight> if both were used. In practice however this shouldn't be a problem as multiple escape sequences are pretty rare unless you're transitioning between multiple code languages. (Multiple multi-char escape sequences beginning with the same character are simply bad practice anyhow.)
• Since byte values are stored as numbers, it is not recommended for you to use a number as an escape sequence (though it may work just fine).
• Placeholder byte values separated with return (<syntaxhighlight lang="text" class="" style="" inline="1">'\r'</syntaxhighlight>) characters--chosen because they are seldom used at all, and virtually never used unpaired with <syntaxhighlight lang="text" class="" style="" inline="1">'\n'</syntaxhighlight>; moreover, it is distinct from the markers generated by <nowiki>...</nowiki> or <syntaxhighlight lang="text" class="" style="" inline="1">mw.text.nowiki()</syntaxhighlight> (which use the delete char). To set a different separator char, include the key-value pair {safeChr = Template:Green} in the table that you pass to escape:char().

Speed

The following are benchmarks...

when executing the following module function: <syntaxhighlight lang="lua">

function p.test_kill500(frame)
 local esc = require('Module:Escape')
 for k = 1, 500 do
  local v = esc:kill(p.test_string2(), 'test')
 end
 return os.clock(esc)
end

</syntaxhighlight> 0.02384

when repeating the following line 500 times in a template:

<syntaxhighlight lang="wikitext" class="" style="" inline="1">{{#invoke:Escape|main|mode=kill|{{#invoke:Escape/testcases|test_string2}}|test}}</syntaxhighlight>

0.767

All times in seconds. The module time x500 was calculated when you loaded this doc page (normally between 0.02 and 0.07). The template time x500 was recorded on Jan 15, 2015.

Examples

Module

Here's some sample output from the debug console below the module editor:

Template:Blue

test, \test, \7b

044
7btest\\ \\ \\\\

Template:Blue

test, 5c

0116
5cest, 5c

055
5cb

044
7btest5c

092
5c 5c

092
5c 5c

092
5c5c

092
5c

Template:Blue

test, 5c

0116
5cest, 5c

055
5cb

044
7btest5c

092
5c 5c

092
5c 5c

092
5c5c

092
5c

Template:Blue

test, test, 7b

044
7btest\ \ \\

Template:Blue

true

Template:BlueTemplate:Green

test, 5c

0116
5cest, 5c

055
5cb

044
7btest5c

092
5c 5c

092
5c 5c

092
5c5c

092
5c

Template:Blue

test, \test, \,test\\ \\ \\\\

Template:Blue

test, test, 7b

044
7btest\ \ \\

Template:Blue

test, test, {,test\ \ \\

Template:Blue

false

Template:Blue

true

Template:Blue

test { test {\{ test, \test, \{,test\ \ \ {\

Template:Blue

test  test { test, test, {,test   \

Template:Blue

true

---

local escape = {
	char = function(self, chr, args)
		args = args or {}
		local safe = args.safeChr or string.char(13)
		chr = tostring(chr or '\\')
		self[1] = ('%s0%%s%s'):format(
			('%x%s%s'):format(chr:byte(), safe, safe),
			('%s%x'):format(safe, chr:byte())
		)
		if not self[self[1]] then
			self[self[1]] = {
				char = chr,
				text = ('%s(.)'):format(chr),
				undo = self[1]:format'(%d+)'
			}
		end
		return args.text and self:text(args.text)
			or args.undo and self:undo(args.undo, chr)
			or args.kill and self:kill(args.kill)
			or self
	end,
	exec = function(self, text, mode, newEscape)
		local target = self[self[1] or self:char() and self[1]]
		for v in text:gfind(target[mode]) do
			text = text:gsub(
				mode == 'text' and
					('%s%s'):format(target.char, v:gsub('%W', '%%%1'))
					or self[1]:format(v),
				mode == 'text' and
					self[1]:format(v:byte())
					or (newEscape or '') .. v:char()
			)
		end
		return text
	end,
	text = function(self, text)
		return self:exec(type(text) == 'table' and text[1] or text, 'text')
	end,
	undo = function(self, text, newEscape)
		if type(text) == 'table' then
			text, newEscape = unpack(text)
		end
		return self:exec(text, 'undo', newEscape)
	end,
	kill = function(self, text, chars, newEscape)
		if type(text) == 'table' then
			text, chars, newEscape = unpack(text)
		end
		return self:undo(self:text(text):gsub(chars or '', ''), newEscape)
	end
}

function escape.main(frame)
	local args, family = {}, {frame:getParent(), frame}
	for f = 1, 2 do
		for k, v in pairs(family[f] and family[f].args or {}) do
			args[k] = args[k] or v:match('^%s*(.-)%s*$')
		end
	end
	if args.mode == 'char' then
		return escape:char(args.char or args[2], args)
	end
	return escape[args.mode](escape:char(args.char), args)
end

return escape