Module:etymon
Jump to navigation
Jump to search
- The following documentation is located at Module:etymon/documentation. [edit] Categories were auto-generated by Module:documentation. [edit]
- Useful links: subpage list • links • transclusions • testcases • sandbox
This module provides the backend for {{etymon}}
.
local export = {}
local concat = table.concat
local find = string.find
local findTemplates = require("Module:template parser").findTemplates
local get_current_title = mw.title.getCurrentTitle
local get_lang = require("Module:languages").getByCode
local insert = table.insert
local ipairs = ipairs
local max = math.max
local new_title = mw.title.new
local split = require("Module:string utilities").split
local sub = string.sub
local type = type
local unpack = unpack
local full_link = require("Module:links").full_link
--Given a full code and entry name, return the Wiktionary page title.
local function pageName(lang, entry)
--TODO: is there a better way?
if sub(entry, 1, 1) == "*" then
entry = lang:makeEntryName(entry)
return "Reconstruction:" .. lang:getCanonicalName() .. "/" .. sub(entry, 2)
else
return lang:makeEntryName(entry)
end
end
--Given an etymon param, return its parts.
local function getParts(lang, etymonParam, normalized)
local parts, etymonLang, etymonPage, etymonId = split(etymonParam, ">", true, true)
if #parts == 2 then
--Assume language is the same as the template call if none is provided.
etymonLang, etymonPage, etymonId = lang, unpack(parts)
else
etymonLang, etymonPage, etymonId = unpack(parts)
etymonLang = get_lang(etymonLang, true, true)
end
if normalized then
etymonLang = etymonLang:getFull()
etymonPage = pageName(etymonLang, etymonPage)
end
return etymonLang, etymonPage, etymonId
end
local paramsOf = {}
--Given an etymon, scrape the page and get its parameters.
--This function returns either: a table of the params, "missing", "redlink", or "nolink"
local function getParams(lang, etymonParam)
--Get normalized parts of the etymon parameter.
local etymonLang, etymonPage, etymonId = getParts(lang, etymonParam, true)
--"?" is a special value that unlinks the page.
if etymonId == "?" then
return "nolink"
end
--Find the parameters by scraping etymonPage.
--Store data in the paramsOf[] table to save time in case the same etymon is accessed again.
--The key is a normalized version of etymonParam.
local key = etymonLang:getFullCode() .. ">" .. etymonPage .. ">" .. etymonId
if paramsOf[key] == nil then
local pageContent = new_title(etymonPage):getContent()
if pageContent == nil then
paramsOf[key] = "redlink"
else
--Find the template on the page.
for name, templateArgs in findTemplates(pageContent) do
if name == "etymon" then
--Might as well store whatever we find along the way to potentially save time later.
paramsOf[templateArgs[1] .. ">" .. etymonPage .. ">" .. templateArgs["id"]] = templateArgs
if templateArgs[1] == etymonLang and templateArgs["id"] == etymonId then
--This "break" saves time only if etymonPage is only ever accessed once.
--Since this is usually true, it is probably beneficial overall.
break
end
end
end
--If scraping did not produce anything, the template must be missing from the page.
if paramsOf[key] == nil then
paramsOf[key] = "missing"
end
end
end
return paramsOf[key]
end
-- [tag]: {abbreviation, label glossary anchor, start text, start text plus, middle text}
-- Note: the keywords `afeq`, `conf`, and `unc` are also recognized, but do not use this dictionary.
local keywordDict = {
["from"] = {false, false, "From", "From", "from"},
["inh"] = {false, false, "From", "[[Appendix:Glossary#inherited|Inherited]] from", "from"},
["af"] = {false, false, "From", "From", "from"},
["bor"] = {"bor.", "loanword", "Borrowed from", "[[Appendix:Glossary#loanword|Borrowed]] from", "borrowed from"},
["lbor"] = {"lbor.", "learned_borrowing", "Learned borrowing from", "[[Appendix:Glossary#learned_borrowing|Learned borrowing]] from", "borrowed from"},
["der"] = {"der.", "derived_terms", "Derived from", "[[Appendix:Glossary#derived_terms|Derived]] from", "from"},
["calque"] = {"calq.", "calque", "Calque of", "[[Appendix:Glossary#calque|Calque]] of", "a calque of"},
["influence"] = {"influ.", "contamination", "", "", ""}
}
--Keep track of what's already been added to the tree and the depth reached.
--Use the normalized etymon parameters as the key.
local alreadySeen = {}
--This function takes an etymon and recursively builds a tree to display in an entry.
local function etyTree(currTitle, lang, args, isTopLevel, isUncertain, label)
local maxDepth = 0
local subtree, subtreeDepth, etymonLang, etymonPage, etymonParams
local subtrees = {}
local normLang = lang:getFullCode()
local normTitle = pageName(lang, currTitle)
local key = normLang .. ">" .. normTitle .. ">" .. args["id"]
local derType, confidence, ignoreEtymons = "from", "conf", false
--If the entry has already been included in the tree, then don't process it again.
if alreadySeen[key] == nil then
--Loop over each parameter in the current template.
for i, param in ipairs(args) do
if i > 1 and find(param, ">") and not ignoreEtymons then
--We want to display the unnormalized language and page title.
--We can't use args[1] here because that would give the normalized language code.
etymonLang, etymonPage = getParts(lang, param, false)
--Scrape the page and get the parameters.
etymonParams = getParams(lang, param)
--Check if the function returned an error code.
if type(etymonParams) == "table" then
--Recurse into the etymon and append its tree to the list of subtrees.
subtree, subtreeDepth = etyTree(etymonPage, etymonLang, etymonParams, false, confidence == "unc", derType)
insert(subtrees, subtree)
maxDepth = max(maxDepth, subtreeDepth)
end
elseif i > 1 then
--Reached a keyword.
if param == "conf" or param == "unc" then
confidence = param
elseif keywordDict[param] ~= nil then
ignoreEtymons = false
confidence = "conf"
derType = param
elseif param == "afeq" then
ignoreEtymons = true
else
error("Received unknown keyword: " .. param)
end
end
end
--Add the page to alreadySeen.
alreadySeen[key] = true
end
--Create term block.
local link
if isTopLevel then
link = lang:getCanonicalName() .. " " .. full_link({lang=lang, alt="'''" .. currTitle .. "'''", tr=lang:transliterate(currTitle)}, "term")
else
link = lang:getCanonicalName() .. " " .. full_link({lang=lang, term=currTitle, id=args["id"]}, "term")
end
--Create tree.
local tree = "<div style=\"position:relative;display:inline;z-index:1;vertical-align:bottom;margin:0 4px 0\"><p style=\"position:relative;margin:8px 0;display:inline-block;padding:5px 10px;background-color:#fffbf2;border:1px solid #ccc;border-radius:4px\">" .. link
--Add a short top connector if multiple subtrees exist.
if #subtrees >= 2 then
tree = tree .. "<span style=\"position:absolute;z-index:-1;inset:-10px 50% 0 auto;border-right:2px solid #9e9e9e\"></span>"
end
tree = tree .. "</p>"
--Add derivation and uncertainty labels.
if (label ~= "" and keywordDict[label][1] ~= false) or isUncertain then
tree = tree .. "<span style=\"position:absolute;z-index:2;transform:translate(-50%);top:calc(100% + 8.5px);left:50%;border-radius:2px;background-color:rgba(234,255,255,0.85);font-size:12px;height:10px;line-height:10px\">"
if label ~= "" and keywordDict[label][1] ~= false then
tree = tree .. "[[Appendix:Glossary#" .. keywordDict[label][2] .. "|<span style=\"color:black;font-style:italic\">" .. keywordDict[label][1] .. "</span>]]"
if isUncertain then
--Add uncertainty label next to the derivation label.
tree = tree .. "<span class=\"desc-arr\" title=\"uncertain\" style=\"position:absolute;top:50%;transform:translate(0,-48%);left:calc(100% + 2px);font-size:10px;border-radius:2px;background-color:rgba(255,224,240,0.85);padding:1px 2px;font-weight:bold\">?</span>"
end
elseif isUncertain then
--Add uncertainty label in the middle.
tree = tree .. "<span class=\"desc-arr\" title=\"uncertain\" style=\"position:absolute;top:50%;left:50%;transform:translate(calc(-50% - 1px),-48%);font-size:10px;border-radius:2px;background-color:rgba(255,224,240,0.85);padding:1px 2px;font-weight:bold\">?</span>"
end
tree = tree .. "</span>"
end
tree = tree .. "</div>"
--Add line break if parents exist.
if #subtrees >= 1 then
tree = "<br>" .. tree
end
--Append subtrees. For cleaner HTML, only add text-align:center div when necessary.
local subtreeString = ""
if #subtrees == 1 then
--Add a long bottom connector to the subtree.
--Use sub() to insert it right before the </div>, which is a little hacky.
subtreeString = sub(subtrees[1], 1, -7) .. "<span style=\"position:absolute;z-index:-1;inset:0 50% 0 auto;height:50px;border-right:2px solid #9e9e9e\"></span></div>"
elseif #subtrees >= 2 then
for i,v in ipairs(subtrees) do
if i == 1 then
--Add left connector.
v = v .. "<span style=\"position:absolute;inset:calc(100% - 9px) -20px 0 calc(50% - 2px);border-bottom:2px solid #9e9e9e;border-left:2px solid #9e9e9e;border-bottom-left-radius:4px\"></span>"
elseif i == #subtrees then
--Add right connector.
v = v .. "<span style=\"position:absolute;inset:calc(100% - 9px) 50% 0 -20px;border-bottom:2px solid #9e9e9e;border-right:2px solid #9e9e9e;border-bottom-right-radius:4px\"></span>"
else
--Add middle connector.
v = v .. "<span style=\"position:absolute;inset:calc(100% - 9px) -20px 0;border-bottom:2px solid #9e9e9e\"></span>"
end
subtreeString = subtreeString .. "<div style=\"position:relative;display:inline-block\">" .. v .. "</div>"
end
end
tree = subtreeString .. tree
--Add outer divs.
if isTopLevel then
tree = "<div style=\"font-size:14px;white-space:nowrap;line-height:1.1;display:inline-block;text-align:center\">" .. tree .. "</div>"
tree = "<div class=\"NavFrame\"><div class=\"NavHead\" style=\"background:#eee\">Etymology tree</div><div class=\"NavContent\" style=\"overflow:auto;text-align:left;padding:0.5em\">" .. tree .. "</div></div>"
end
return tree, maxDepth + 1
end
--This function takes an etymon and generates some text to display in an entry.
--Currently, it is only able to handle simple combinations of parameters.
local function etyText(title, lang, args, usePlusTemplates, maxDepth)
local text = ""
local depth = 1
local alreadyWritten = {}
local key, currLang, group, groupType, groupConfidence, confidence, derType, foundGroup, complexParams, ignoreEtymons, etymonLang, etymonTitle, etymonId
local params = args
--Loop and continuously expand the sentence until we reach the end of the chain.
while not maxDepth or depth <= maxDepth do
group, groupType, groupConfidence, confidence, derType, foundGroup, complexParams, ignoreEtymons, currLang = {}, "from", "conf", "conf", "from", false, false, false, lang
key = lang:getFullCode() .. ">" .. pageName(lang, title) .. ">" .. args["id"]
--Stop if we encounter an already-seen term.
if alreadyWritten[key] ~= nil then
break
end
alreadyWritten[key] = true
for i, param in ipairs(params) do
if i > 1 and find(param, ">") and not ignoreEtymons then
--The text should only continue if `args` is either (not including `influence` or `afeq` etymons):
--A single etymon, or single `af` group. Otherwise the parameters are too "complex" and are rejected.
--TODO: add smarter handling for complex parameters.
if foundGroup or (#group == 1 and derType ~= "af") then
complexParams = true
break
end
groupType = derType
if confidence == "unc" then
groupConfidence = "unc"
end
insert(group, param)
elseif i > 1 then
--Reached a keyword.
--Check if af was only given one parameter, which doesn't make sense.
if #group == 2 and derType == "af" then
complexParams = true
break
end
if param == "unc" then
confidence = param
elseif param == "afeq" or param == "influence" then
ignoreEtymons = true
if #group == 1 then
foundGroup = true
end
elseif keywordDict[param] ~= nil then
ignoreEtymons = false
confidence = "conf"
derType = param
if #group == 1 then
foundGroup = true
end
else
error("Received unknown keyword: " .. param)
end
end
end
if complexParams or #group == 0 then
break
end
if #group == 1 then
params = getParams(lang, group[1])
end
if text == "" then
--Start the sentence.
if groupConfidence == "conf" and not usePlusTemplates then
text = keywordDict[groupType][3]
elseif groupConfidence == "conf" and usePlusTemplates then
text = keywordDict[groupType][4]
else
text = "Possibly " .. keywordDict[groupType][5]
end
else
--Add a phrase onto the sentence.
if groupConfidence == "conf" then
text = text .. ", " .. keywordDict[groupType][5]
else
text = text .. ", possibly " .. keywordDict[groupType][5]
end
end
--Add the links.
for i = 1,#group do
etymonLang, etymonTitle, etymonId = getParts(lang, group[i], false)
if etymonLang ~= currLang then
group[i] = etymonLang:makeWikipediaLink() .. " " .. full_link({lang=etymonLang, term=etymonTitle, id=etymonId}, "term")
currLang = etymonLang
else
group[i] = full_link({lang=etymonLang, term=etymonTitle, id=etymonId}, "term")
end
end
text = text .. " " .. concat(group, " + ")
depth = depth + 1
if #group == 2 then
break
end
lang = etymonLang
title = etymonTitle
if type(params) ~= "table" then
break
end
end
--Add period at the end of the sentence.
if text ~= "" then
text = text .. "."
end
return text
end
function export.main(frame)
local args = frame:getParent().args
if args["id"] == nil then
error("You need to provide an `id` parameter!")
end
local lang = get_lang(args[1], true, true)
--The `title` parameter is used for overriding the page title.
local title = args["title"]
if args["title"] == nil then
title = get_current_title().prefixedText
end
--Add anchor to output.
local output = {"<ul id=\"" .. lang:getCanonicalName() .. ":_" .. args["id"] .. "\"></ul>"}
local currNode = args[1] .. ">" .. title .. ">" .. args["id"]
paramsOf[currNode] = args
--Insert tree.
if args["tree"] ~= nil then
local tree, _ = etyTree(title, lang, args, true, false, "")
insert(output, tree)
end
--Insert text.
if args["text"] == "++" then
insert(output, etyText(title, lang, args, true, false))
elseif args["text"] == "+" then
insert(output, etyText(title, lang, args, true, 1))
elseif args["text"] == "-" then
insert(output, etyText(title, lang, args, false, 1))
elseif args["text"] ~= nil then
insert(output, etyText(title, lang, args, false, false))
end
return concat(output)
end
return export