Module:Text
This Lua module is used on approximately 1,570,000 pages, or roughly 22816% of all pages. To avoid major disruption and server load, any changes should be tested in the module's /sandbox or /testcases subpages, or in your own module sandbox. The tested changes can be added to this page in a single edit. Consider discussing changes on the talk page before implementing them. |
Text
– Module containing methods for the manipulation of text, wikimarkup and some HTML.
Template:Anchor Functions for templates
All methods have an unnamed parameter containing the text.
The return value is an empty string if the parameter does not meet the conditions. When the condition is matched or some result is successfully found, strings of at least one character are returned.
- char Template:Anchor
- Creates a string from a list of character codes.
- 1
- Space-separated list of character codes
- *
- Number of repetitions of the list in parameter 1; (Default 1).
- errors
0
– Silence errors
- concatParams Template:Anchor
- Combine any number of elements into a list, like
table.concat()
in Lua. - From a template:
- 1
- First element; missing and empty elements are ignored.
- 2 3 4 5 6 …
- Further list elements
- From Lua
- args
- table (sequence) of the elements
- apply
- Separator between elements; defaults to
|
- adapt
- optional formatting, which will be applied to each element; must contain
%s
.
- containsCJK Template:Anchor
- Returns whether the input string contains any CJK characters
- Returns nothing if there are no CJK characters
- removeDelimited
- Remove all text between delimiters, including the delimiters themselves.
- getPlainTemplate:Anchor
- Remove wikimarkup (except templates): comments, tags, bold, italic, nbsp
- isLatinRange Template:Anchor
- Returns some content, unless the string contains a character that would not normally be found in Latin text.
- Returns nothing if there is a non-Latin string.
- isQuote Template:Anchor
- Returns some content if the parameter passed is a single character, and that character is a quote, such as
'
.- Returns nothing for multiple characters, or if the character passed is not a quote.
- listToText Template:Anchor
- Formats list elements analogously to mw.text.listToText().
- The elements are separated by a comma and space ; the word "and" appears between the first and last.
- Unnamed parameters become the list items.
- Optional parameters for
#invoke
:format
– Every list element will first be formatted with this format string; see here for how to construct this string. The string must contain at least one%s
sequence.template=1
– List elements should be taken from the calling template.
- Returns the resulting string.
- quote Template:Anchor
- Wrap the string in quotes; quotes can be chosen for a specific language.
- 1
- Input text (will be automatically trimmed); may be empty.
- 2
- (optional) the ISO 639 language code for the quote marks; should be one of the supported languages Template:In lang
- 3
- (optional)
2
for second level quotes. This means the single quote marks in a statement such as: Jack said, “Jill said ‘fish’ last Tuesday.”
- quoteUnquoted Template:Anchor
- Wrap the string in quotes; quotes can be chosen for a specific language. Will not quote an empty string, and will not quote if there is a quote at the start or end of the (trimmed) string.
- 1
- Input text (will be automatically trimmed); may be empty.
- 2
- (optional) the ISO 639 language code for the quote marks; should be one of the supported languages Template:In lang
- 3
- (optional)
2
for second level quotes. This means the single quote marks in a statement such as: Jack said, “Jill said ‘fish’ last Tuesday.”
- removeDiacritics Template:Anchor
- Removes all diacritical marks from the input.
- 1
- Input text
- sentenceTerminated Template:Anchor
- Is this sentence terminated? Should work with CJK, and allows quotation marks to follow.
- Returns nothing if the sentence is unterminated.
- ucfirstAll Template:Anchor
- The first letter of every recognized word is converted to upper case. This contrasts with the parser function Template:Tlg which changes only the first character of the whole string passed.
- A few common HTML entities are protected; the implementation of this may mean that numerical entities passed (e.g.
&)
are converted to&
form - uprightNonlatin Template:Anchor
- Takes a string. Italicized non-Latin characters are un-italicized, unless they are a single Greek letter.
- zip Template:Anchor
- Combines a tuple of lists by convolution. This is easiest to explain by example: given two lists, list1 = "a b c" and list2 = "1 2 3", then
zip(liste1, liste2, sep = " ", isep = "-", osep = "/")
outputsa-1/b-2/c-3
- 1, 2, 3, … – Lists to be combined
sep
– A separator (in Lua regex form) used to split the lists. If empty, the lists are split into individual characters.sep1
,sep2
,sep3
, … – Allows a different separator to be used for each list.isep
– Output separator; placed between elements which were at the same index in their lists.osep
– Output separator; placed between elements which had different original indices; i.e. between the groups joined withisep
Examples and test page
There are tests available Template:In lang to illustrate this in practice.
Template:Anchor Use in another Lua module
All of the above functions can be called from other Lua modules. Use require()
; the below code checks for errors loading it:
<syntaxhighlight lang="lua">
local lucky, Text = pcall( require, "Module:Text" )
if type( Text ) == "table" then
Text = Text.Text()
else
-- In the event of errors, Text is an error message.
return "" .. Text .. ""
end </syntaxhighlight> You may then call:
- Text.char( apply, again, accept )
- Text.concatParams( args, separator, format )
- Text.containsCJK( s )
- Text.removeDelimited( s )
- Text.getPlain( s )
- Text.isLatinRange( s )
- Text.isQuote( c )
- Text.listToText( table, format )
- Text.quote( s, lang, mode )
- Text.quoteUnquoted( s, lang, mode )
- Text.removeDiacritics( s )
- Text.sentenceTerminated( s )
- Text.ucfirstAll( s )
- Text.uprightNonlatin( s )
Text.zip(…)- Text.test( s )
Usage
This is a general library; use it anywhere.
Dependencies
None.
See also
Template:String-handling templates
local yesNo = require("Module:Yesno") local Text = { serial = "2022-07-21", suite = "Text" } --[=[ Text utilities ]=] -- local globals local PatternCJK = false local PatternCombined = false local PatternLatin = false local PatternTerminated = false local QuoteLang = false local QuoteType = false local RangesLatin = false local SeekQuote = false local function initLatinData() if not RangesLatin then RangesLatin = { { 7, 687 }, { 7531, 7578 }, { 7680, 7935 }, { 8194, 8250 } } end if not PatternLatin then local range PatternLatin = "^[" for i = 1, #RangesLatin do range = RangesLatin[ i ] PatternLatin = PatternLatin .. mw.ustring.char( range[ 1 ], 45, range[ 2 ] ) end -- for i PatternLatin = PatternLatin .. "]*$" end end local function initQuoteData() -- Create quote definitions if not QuoteLang then QuoteLang = { af = "bd", ar = "la", be = "labd", bg = "bd", ca = "la", cs = "bd", da = "bd", de = "bd", dsb = "bd", et = "bd", el = "lald", en = "ld", es = "la", eu = "la", -- fa = "la", fi = "rd", fr = "laSPC", ga = "ld", he = "ldla", hr = "bd", hsb = "bd", hu = "bd", hy = "labd", id = "rd", is = "bd", it = "ld", ja = "x300C", ka = "bd", ko = "ld", lt = "bd", lv = "bd", nl = "ld", nn = "la", no = "la", pl = "bdla", pt = "lald", ro = "bdla", ru = "labd", sk = "bd", sl = "bd", sq = "la", sr = "bx", sv = "rd", th = "ld", tr = "ld", uk = "la", zh = "ld", ["de-ch"] = "la", ["en-gb"] = "lsld", ["en-us"] = "ld", ["fr-ch"] = "la", ["it-ch"] = "la", ["pt-br"] = "ldla", ["zh-tw"] = "x300C", ["zh-cn"] = "ld" } end if not QuoteType then QuoteType = { bd = { { 8222, 8220 }, { 8218, 8217 } }, bdla = { { 8222, 8220 }, { 171, 187 } }, bx = { { 8222, 8221 }, { 8218, 8217 } }, la = { { 171, 187 }, { 8249, 8250 } }, laSPC = { { 171, 187 }, { 8249, 8250 }, true }, labd = { { 171, 187 }, { 8222, 8220 } }, lald = { { 171, 187 }, { 8220, 8221 } }, ld = { { 8220, 8221 }, { 8216, 8217 } }, ldla = { { 8220, 8221 }, { 171, 187 } }, lsld = { { 8216, 8217 }, { 8220, 8221 } }, rd = { { 8221, 8221 }, { 8217, 8217 } }, x300C = { { 0x300C, 0x300D }, { 0x300E, 0x300F } } } end end -- initQuoteData() local function fiatQuote( apply, alien, advance ) -- Quote text -- Parameter: -- apply -- string, with text -- alien -- string, with language code -- advance -- number, with level 1 or 2 local r = apply and tostring(apply) or "" alien = alien or "en" advance = tonumber(advance) or 0 local suite initQuoteData() local slang = alien:match( "^(%l+)-" ) suite = QuoteLang[alien] or slang and QuoteLang[slang] or QuoteLang["en"] if suite then local quotes = QuoteType[ suite ] if quotes then local space if quotes[ 3 ] then space = " " else space = "" end quotes = quotes[ advance ] if quotes then r = mw.ustring.format( "%s%s%s%s%s", mw.ustring.char( quotes[ 1 ] ), space, apply, space, mw.ustring.char( quotes[ 2 ] ) ) end else mw.log( "fiatQuote() " .. suite ) end end return r end -- fiatQuote() Text.char = function ( apply, again, accept ) -- Create string from codepoints -- Parameter: -- apply -- table (sequence) with numerical codepoints, or nil -- again -- number of repetitions, or nil -- accept -- true, if no error messages to be appended -- Returns: string local r = "" apply = type(apply) == "table" and apply or {} again = math.floor(tonumber(again) or 1) if again < 1 then return "" end local bad = { } local codes = { } for _, v in ipairs( apply ) do local n = tonumber(v) if not n or (n < 32 and n ~= 9 and n ~= 10) then table.insert(bad, tostring(v)) else table.insert(codes, math.floor(n)) end end if #bad > 0 then if not accept then r = tostring( mw.html.create( "span" ) :addClass( "error" ) :wikitext( "bad codepoints: " .. table.concat( bad, " " )) ) end return r end if #codes > 0 then r = mw.ustring.char( unpack( codes ) ) if again > 1 then r = r:rep(again) end end return r end -- Text.char() local function trimAndFormat(args, fmt) local result = {} if type(args) ~= 'table' then args = {args} end for _, v in ipairs(args) do v = mw.text.trim(tostring(v)) if v ~= "" then table.insert(result,fmt and mw.ustring.format(fmt, v) or v) end end return result end Text.concatParams = function ( args, apply, adapt ) -- Concat list items into one string -- Parameter: -- args -- table (sequence) with numKey=string -- apply -- string (optional); separator (default: "|") -- adapt -- string (optional); format including "%s" -- Returns: string local collect = { } return table.concat(trimAndFormat(args,adapt), apply or "|") end -- Text.concatParams() Text.containsCJK = function ( s ) -- Is any CJK code within? -- Parameter: -- s -- string -- Returns: true, if CJK detected s = s and tostring(s) or "" if not patternCJK then patternCJK = mw.ustring.char( 91, 4352, 45, 4607, 11904, 45, 42191, 43072, 45, 43135, 44032, 45, 55215, 63744, 45, 64255, 65072, 45, 65103, 65381, 45, 65500, 131072, 45, 196607, 93 ) end return mw.ustring.find( s, patternCJK ) ~= nil end -- Text.containsCJK() Text.removeDelimited = function (s, prefix, suffix) -- Remove all text in s delimited by prefix and suffix (inclusive) -- Arguments: -- s = string to process -- prefix = initial delimiter -- suffix = ending delimiter -- Returns: stripped string s = s and tostring(s) or "" prefix = prefix and tostring(prefix) or "" suffix = suffix and tostring(suffix) or "" local prefixLen = mw.ustring.len(prefix) local suffixLen = mw.ustring.len(suffix) if prefixLen == 0 or suffixLen == 0 then return s end local i = s:find(prefix, 1, true) local r = s local j while i do j = r:find(suffix, i + prefixLen) if j then r = r:sub(1, i - 1)..r:sub(j+suffixLen) else r = r:sub(1, i - 1) end i = r:find(prefix, 1, true) end return r end Text.getPlain = function ( adjust ) -- Remove wikisyntax from string, except templates -- Parameter: -- adjust -- string -- Returns: string local r = Text.removeDelimited(adjust,"<!--","-->") r = r:gsub( "(</?%l[^>]*>)", "" ) :gsub( "'''", "" ) :gsub( "''", "" ) :gsub( " ", " " ) return r end -- Text.getPlain() Text.isLatinRange = function (s) -- Are characters expected to be latin or symbols within latin texts? -- Arguments: -- s = string to analyze -- Returns: true, if valid for latin only s = s and tostring(s) or "" --- ensure input is always string initLatinData() return mw.ustring.match(s, PatternLatin) ~= nil end -- Text.isLatinRange() Text.isQuote = function ( s ) -- Is this character any quotation mark? -- Parameter: -- s = single character to analyze -- Returns: true, if s is quotation mark s = s and tostring(s) or "" if s == "" then return false end if not SeekQuote then SeekQuote = mw.ustring.char( 34, -- " 39, -- ' 171, -- laquo 187, -- raquo 8216, -- lsquo 8217, -- rsquo 8218, -- sbquo 8220, -- ldquo 8221, -- rdquo 8222, -- bdquo 8249, -- lsaquo 8250, -- rsaquo 0x300C, -- CJK 0x300D, -- CJK 0x300E, -- CJK 0x300F ) -- CJK end return mw.ustring.find( SeekQuote, s, 1, true ) ~= nil end -- Text.isQuote() Text.listToText = function ( args, adapt ) -- Format list items similar to mw.text.listToText() -- Parameter: -- args -- table (sequence) with numKey=string -- adapt -- string (optional); format including "%s" -- Returns: string return mw.text.listToText(trimAndFormat(args, adapt)) end -- Text.listToText() Text.quote = function ( apply, alien, advance ) -- Quote text -- Parameter: -- apply -- string, with text -- alien -- string, with language code, or nil -- advance -- number, with level 1 or 2, or nil -- Returns: quoted string apply = apply and tostring(apply) or "" local mode, slang if type( alien ) == "string" then slang = mw.text.trim( alien ):lower() else slang = mw.title.getCurrentTitle().pageLanguage if not slang then -- TODO FIXME: Introduction expected 2017-04 slang = mw.language.getContentLanguage():getCode() end end if advance == 2 then mode = 2 else mode = 1 end return fiatQuote( mw.text.trim( apply ), slang, mode ) end -- Text.quote() Text.quoteUnquoted = function ( apply, alien, advance ) -- Quote text, if not yet quoted and not empty -- Parameter: -- apply -- string, with text -- alien -- string, with language code, or nil -- advance -- number, with level 1 or 2, or nil -- Returns: string; possibly quoted local r = mw.text.trim( apply and tostring(apply) or "" ) local s = mw.ustring.sub( r, 1, 1 ) if s ~= "" and not Text.isQuote( s, advance ) then s = mw.ustring.sub( r, -1, 1 ) if not Text.isQuote( s ) then r = Text.quote( r, alien, advance ) end end return r end -- Text.quoteUnquoted() Text.removeDiacritics = function ( adjust ) -- Remove all diacritics -- Parameter: -- adjust -- string -- Returns: string; all latin letters should be ASCII -- or basic greek or cyrillic or symbols etc. local cleanup, decomposed if not PatternCombined then PatternCombined = mw.ustring.char( 91, 0x0300, 45, 0x036F, 0x1AB0, 45, 0x1AFF, 0x1DC0, 45, 0x1DFF, 0xFE20, 45, 0xFE2F, 93 ) end decomposed = mw.ustring.toNFD( adjust and tostring(adjust) or "" ) cleanup = mw.ustring.gsub( decomposed, PatternCombined, "" ) return mw.ustring.toNFC( cleanup ) end -- Text.removeDiacritics() Text.sentenceTerminated = function ( analyse ) -- Is string terminated by dot, question or exclamation mark? -- Quotation, link termination and so on granted -- Parameter: -- analyse -- string -- Returns: true, if sentence terminated local r if not PatternTerminated then PatternTerminated = mw.ustring.char( 91, 12290, 65281, 65294, 65311 ) .. "!%.%?…][\"'%]‹›«»‘’“”]*$" end if mw.ustring.find( analyse, PatternTerminated ) then r = true else r = false end return r end -- Text.sentenceTerminated() Text.ucfirstAll = function ( adjust) -- Capitalize all words -- Arguments: -- adjust = string to adjust -- Returns: string with all first letters in upper case adjust = adjust and tostring(adjust) or "" local r = mw.text.decode(adjust,true) local i = 1 local c, j, m m = (r ~= adjust) r = " "..r while i do i = mw.ustring.find( r, "%W%l", i ) if i then j = i + 1 c = mw.ustring.upper( mw.ustring.sub( r, j, j ) ) r = string.format( "%s%s%s", mw.ustring.sub( r, 1, i ), c, mw.ustring.sub( r, i + 2 ) ) i = j end end -- while i r = r:sub( 2 ) if m then r = mw.text.encode(r) end return r end -- Text.ucfirstAll() Text.uprightNonlatin = function ( adjust ) -- Ensure non-italics for non-latin text parts -- One single greek letter might be granted -- Precondition: -- adjust -- string -- Returns: string with non-latin parts enclosed in <span> local r initLatinData() if mw.ustring.match( adjust, PatternLatin ) then -- latin only, horizontal dashes, quotes r = adjust else local c local j = false local k = 1 local m = false local n = mw.ustring.len( adjust ) local span = "%s%s<span dir='auto' style='font-style:normal'>%s</span>" local flat = function ( a ) -- isLatin local range for i = 1, #RangesLatin do range = RangesLatin[ i ] if a >= range[ 1 ] and a <= range[ 2 ] then return true end end -- for i end -- flat() local focus = function ( a ) -- char is not ambivalent local r = ( a > 64 ) if r then r = ( a < 8192 or a > 8212 ) else r = ( a == 38 or a == 60 ) -- '&' '<' end return r end -- focus() local form = function ( a ) return string.format( span, r, mw.ustring.sub( adjust, k, j - 1 ), mw.ustring.sub( adjust, j, a ) ) end -- form() r = "" for i = 1, n do c = mw.ustring.codepoint( adjust, i, i ) if focus( c ) then if flat( c ) then if j then if m then if i == m then -- single greek letter. j = false end m = false end if j then local nx = i - 1 local s = "" for ix = nx, 1, -1 do c = mw.ustring.sub( adjust, ix, ix ) if c == " " or c == "(" then nx = nx - 1 s = c .. s else break -- for ix end end -- for ix r = form( nx ) .. s j = false k = i end end elseif not j then j = i if c >= 880 and c <= 1023 then -- single greek letter? m = i + 1 else m = false end end elseif m then m = m + 1 end end -- for i if j and ( not m or m < n ) then r = form( n ) else r = r .. mw.ustring.sub( adjust, k ) end end return r end -- Text.uprightNonlatin() Text.test = function ( about ) local r if about == "quote" then initQuoteData() r = { } r.QuoteLang = QuoteLang r.QuoteType = QuoteType end return r end -- Text.test() -- Export local p = { } for _, func in ipairs({'containsCJK','isLatinRange','isQuote','sentenceTerminated'}) do p[func] = function (frame) return Text[func]( frame.args[ 1 ] or "" ) and "1" or "" end end for _, func in ipairs({'getPlain','removeDiacritics','ucfirstAll','uprightNonlatin'}) do p[func] = function (frame) return Text[func]( frame.args[ 1 ] or "" ) end end function p.char( frame ) local params = frame:getParent().args local story = params[ 1 ] local codes, lenient, multiple if not story then params = frame.args story = params[ 1 ] end if story then local items = mw.text.split( mw.text.trim(story), "%s+" ) if #items > 0 then local j lenient = (yesNo(params.errors) == false) codes = { } multiple = tonumber( params[ "*" ] ) for _, v in ipairs( items ) do j = tonumber((v:sub( 1, 1 ) == "x" and "0" or "") .. v) table.insert( codes, j or v ) end end end return Text.char( codes, multiple, lenient ) end function p.concatParams( frame ) local args local template = frame.args.template if type( template ) == "string" then template = mw.text.trim( template ) template = ( template == "1" ) end if template then args = frame:getParent().args else args = frame.args end return Text.concatParams( args, frame.args.separator, frame.args.format ) end function p.listToFormat(frame) local lists = {} local pformat = frame.args["format"] local sep = frame.args["sep"] or ";" -- Parameter parsen: Listen for k, v in pairs(frame.args) do local knum = tonumber(k) if knum then lists[knum] = v end end -- Listen splitten local maxListLen = 0 for i = 1, #lists do lists[i] = mw.text.split(lists[i], sep) if #lists[i] > maxListLen then maxListLen = #lists[i] end end -- Ergebnisstring generieren local result = "" local result_line = "" for i = 1, maxListLen do result_line = pformat for j = 1, #lists do result_line = mw.ustring.gsub(result_line, "%%s", lists[j][i], 1) end result = result .. result_line end return result end function p.listToText( frame ) local args local template = frame.args.template if type( template ) == "string" then template = mw.text.trim( template ) template = ( template == "1" ) end if template then args = frame:getParent().args else args = frame.args end return Text.listToText( args, frame.args.format ) end function p.quote( frame ) local slang = frame.args[2] if type( slang ) == "string" then slang = mw.text.trim( slang ) if slang == "" then slang = false end end return Text.quote( frame.args[ 1 ] or "", slang, tonumber( frame.args[3] ) ) end function p.quoteUnquoted( frame ) local slang = frame.args[2] if type( slang ) == "string" then slang = mw.text.trim( slang ) if slang == "" then slang = false end end return Text.quoteUnquoted( frame.args[ 1 ] or "", slang, tonumber( frame.args[3] ) ) end function p.zip(frame) local lists = {} local seps = {} local defaultsep = frame.args["sep"] or "" local innersep = frame.args["isep"] or "" local outersep = frame.args["osep"] or "" -- Parameter parsen for k, v in pairs(frame.args) do local knum = tonumber(k) if knum then lists[knum] = v else if string.sub(k, 1, 3) == "sep" then local sepnum = tonumber(string.sub(k, 4)) if sepnum then seps[sepnum] = v end end end end -- sofern keine expliziten Separatoren angegeben sind, den Standardseparator verwenden for i = 1, math.max(#seps, #lists) do if not seps[i] then seps[i] = defaultsep end end -- Listen splitten local maxListLen = 0 for i = 1, #lists do lists[i] = mw.text.split(lists[i], seps[i]) if #lists[i] > maxListLen then maxListLen = #lists[i] end end local result = "" for i = 1, maxListLen do if i ~= 1 then result = result .. outersep end for j = 1, #lists do if j ~= 1 then result = result .. innersep end result = result .. (lists[j][i] or "") end end return result end function p.failsafe() return Text.serial end p.Text = function () return Text end -- p.Text return p