Linguifex

Module:script utilities: Difference between revisions

Module Discussion

Module:script utilities (view source)

Revision as of 19:29, 30 November 2022

3,409 bytes added ,  30 November 2022
no edit summary
No edit summary
No edit summary
Line 9: Line 9:
[[Module:languages]]
[[Module:languages]]
[[Module:parameters]]
[[Module:parameters]]
[[Module:utilities]]
[[Module:utilities/format_categories]]
]=]
]=]


function export.is_Latin_script(sc)
function export.is_Latin_script(sc)
-- Latn, Latf, Latinx, pjt-Latn
-- Latn, Latf, Latinx, pjt-Latn
return nil and true or mw.ustring.find(require("Module:scripts").getByCode(sc)["_code"], "Lat") and true or false
return sc:getCode():find("Lat") and true or false
end
end


Line 43: Line 43:
end
end


-- Apply a function to `text`, but not to the target of wikilinks or to HTML tags.
-- Ustring turns on the codepoint-aware string matching. The basic string function
local function munge_text(text, fn)
-- should be used for simple sequences of characters, Ustring function for
local has_html = text:find("<")
-- sets – [].
local has_two_part_link = text:find("%[%[.*|")
local function trackPattern(text, pattern, tracking, ustring)
if not has_html and not has_two_part_link then
local find = ustring and mw.ustring.find or string.find
return fn(text)
end
end
 
local function track(text, lang, sc)
local U = mw.ustring.char
if lang and text then
local langCode = lang:getCode()
-- [[Special:WhatLinksHere/Template:tracking/script/ang/acute]]
if langCode == "ang" then
local decomposed = mw.ustring.toNFD(text)
local acute = U(0x301)
trackPattern(decomposed, acute, "ang/acute")
--[=[
[[Special:WhatLinksHere/Template:tracking/script/Greek/wrong-phi]]
[[Special:WhatLinksHere/Template:tracking/script/Greek/wrong-theta]]
[[Special:WhatLinksHere/Template:tracking/script/Greek/wrong-kappa]]
[[Special:WhatLinksHere/Template:tracking/script/Greek/wrong-rho]]
ϑ, ϰ, ϱ, ϕ should generally be replaced with θ, κ, ρ, φ.
]=]
elseif langCode == "el" or langCode == "grc" then
trackPattern(text, "ϑ", "Greek/wrong-theta")
trackPattern(text, "ϰ", "Greek/wrong-kappa")
trackPattern(text, "ϱ", "Greek/wrong-rho")
trackPattern(text, "ϕ", "Greek/wrong-phi")
--[=[
[[Special:WhatLinksHere/Template:tracking/script/Ancient Greek/spacing-coronis]]
[[Special:WhatLinksHere/Template:tracking/script/Ancient Greek/spacing-smooth-breathing]]
[[Special:WhatLinksHere/Template:tracking/script/Ancient Greek/wrong-apostrophe]]
When spacing coronis and spacing smooth breathing are used as apostrophes,
they should be replaced with right single quotation marks (’).
]=]
if langCode == "grc" then
trackPattern(text, U(0x1FBD), "Ancient Greek/spacing-coronis")
trackPattern(text, U(0x1FBF), "Ancient Greek/spacing-smooth-breathing")
trackPattern(text, "[" .. U(0x1FBD) .. U(0x1FBF) .. "]", "Ancient Greek/wrong-apostrophe", true)
end
-- [[Special:WhatLinksHere/Template:tracking/script/Russian/grave-accent]]
elseif langCode == "ru" then
local decomposed = mw.ustring.toNFD(text)
trackPattern(decomposed, U(0x300), "Russian/grave-accent")
-- [[Special:WhatLinksHere/Template:tracking/script/Tibetan/trailing-punctuation]]
elseif langCode == "bo" then
trackPattern(text, "[་།]$", "Tibetan/trailing-punctuation", true)
trackPattern(text, "[་།]%]%]$", "Tibetan/trailing-punctuation", true)


local strutils = require("Module:string utilities")
--[=[
[[Special:WhatLinksHere/Template:tracking/script/Thai/broken-ae]]
[[Special:WhatLinksHere/Template:tracking/script/Thai/broken-am]]
[[Special:WhatLinksHere/Template:tracking/script/Thai/wrong-rue-lue]]
]=]
elseif langCode == "th" then
trackPattern(text, "เ".."เ", "Thai/broken-ae")
trackPattern(text, "ํ[่้๊๋]?า", "Thai/broken-am", true)
trackPattern(text, "[ฤฦ]า", "Thai/wrong-rue-lue", true)


local function munge_text_with_html(txt)
--[=[
local parts = strutils.capturing_split(txt, "(<[^>]->)")
[[Special:WhatLinksHere/Template:tracking/script/Lao/broken-ae]]
for i = 1, #parts, 2 do
[[Special:WhatLinksHere/Template:tracking/script/Lao/broken-am]]
parts[i] = fn(parts[i])
[[Special:WhatLinksHere/Template:tracking/script/Lao/possible-broken-ho-no]]
end
[[Special:WhatLinksHere/Template:tracking/script/Lao/possible-broken-ho-mo]]
return table.concat(parts)
[[Special:WhatLinksHere/Template:tracking/script/Lao/possible-broken-ho-lo]]
end
]=]
elseif langCode == "lo" then
trackPattern(text, "ເ".."ເ", "Lao/broken-ae")
trackPattern(text, "ໍ[່້໊໋]?າ", "Lao/broken-am", true)
trackPattern(text, "ຫນ", "Lao/possible-broken-ho-no")
trackPattern(text, "ຫມ", "Lao/possible-broken-ho-mo")
trackPattern(text, "ຫລ", "Lao/possible-broken-ho-lo")


if has_two_part_link then
--[=[
-- The hard case is when both two-part links and HTML tags occur, because crippled Lua patterns
[[Special:WhatLinksHere/Template:tracking/script/Lü/broken-ae]]
-- don't support alternation. We need to first split on two-part links (which seem more likely
[[Special:WhatLinksHere/Template:tracking/script/Lü/possible-wrong-sequence]]
-- to occur), then split odd-numbered fragments on HTML tags, then apply the function to
]=]
-- odd-numbered subfragments. This is unlikely to be very efficient, but should occur rarely.
elseif langCode == "khb" then
local parts = strutils.capturing_split(text, "(%[%[[^%[%]|]-|)")
trackPattern(text, "ᦵ".."ᦵ", "Lü/broken-ae")
for i = 1, #parts, 2 do
trackPattern(text, "[ᦀ-ᦫ][ᦵᦶᦷᦺ]", "Lü/possible-wrong-sequence", true)
if has_html then
parts[i] = munge_text_with_html(parts[i])
else
parts[i] = fn(parts[i])
end
end
end
return table.concat(parts)
else -- HTML tags only
return munge_text_with_html(text)
end
end
end
end
Line 84: Line 139:
if not sc then
if not sc then
sc = require("Module:scripts").findBestScript(text, lang)
sc = require("Module:scripts").findBestScript(text, lang)
end
track(text, lang, sc)
-- Replace space characters with newlines in Mongolian-script text, which is written top-to-bottom.
if sc:getDirection() == "down" and text:find(" ") then
text = require("Module:munge_text")(text, function(txt)
-- having extra parentheses makes sure only the first return value gets through
return (txt:gsub(" +", "<br>"))
end)
end
end


-- Hack Korean text to remove hyphens. This should be handled in a more general fashion, but needs to
-- Hack Korean script text to remove hyphens.
-- XXX: This should be handled in a more general fashion, but needs to
-- be efficient by not doing anything if no hyphens are present, and currently this is the only
-- be efficient by not doing anything if no hyphens are present, and currently this is the only
-- language needing such processing.
-- language needing such processing.
if lang:getCode() == "ko" and text:find("%-") then
-- 20220221: Also convert 漢字(한자) to ruby, instead of needing [[Template:Ruby]].
text = munge_text(text, function(txt)
if sc:getCode() == "Kore" and (text:find("%-") or text:find("[()]")) then
-- having extra parentheses makes sure only the first return value gets through
text = require("Module:munge_text")(text, function(txt)
return (txt:gsub("%-", ""))
-- Hani/Hang regex is a reasonable subset of Hani/Hang from [[Module:scripts/data]],
-- last checked on 20220221
txt = txt:gsub("%-", "")
txt = mw.ustring.gsub(txt, "([一-鿿㐀-䶿𠀀-𮯯𰀀-𱍏]+)%(([가-힣ᄀ-ᇿꥠ-ꥼힰ-ퟻ]+)%)", "<ruby>%1<rp>(</rp><rt>%2</rt><rp>)</rp></ruby>")
return txt
end)
end)
end
end
Line 249: Line 319:
return "<small>[" .. disp_script .. " needed]</small>" ..
return "<small>[" .. disp_script .. " needed]</small>" ..
(nocat and "" or require("Module:utilities").format_categories({category}, lang, sort_key))
(nocat and "" or require("Module:utilities/format_categories")({category}, lang, sort_key))
end
end


Retrieved from "https://linguifex.com/wiki/Special:MobileDiff/292228"