require('strict');
local p = {};
--[=[------------------------< I S O _ S Y N O N Y M _ E X T R A C T >-----------------------------------------
{{#invoke:Language/name/data/ISO 639 synonym extraction tool|ISO_synonym_extract|file-date=2013-01-11}}
reads a local copy of data from the table at http://www.loc.gov/standards/iso639-2/php/English_list.php, extracts
the ISO 639-2 (or 639-2T) codes that have equivalent ISO 639-1 codes and creates a table to translate 639-2 to 639-1.
ISO-639-3 uses 639-2T codes
useful lines in the source table have the form:
<English name>\t<all English names>\t<all French names>\t<639-2 code>\t<639-1 code>\n
where:
<English name> is primary English name (not used here); one of <all English names> so duplicates code listing
<all English names> is all of the English names (not used here)
<all French names> is all of the French names (not used here)
<639-2 code> is the three-character ISO 639-2 or 639-2B/639-2T language code; when 639-2T present, use that code
<639-1 code> is the two-character ISO 639-1 language code synonym of the -2 code (if one is defined)
like this (with synonym):
Abkhazian Abkhazian abkhaze abk ab
or (without synonym):
Achinese Achinese aceh ace
for the file date use the date listed at the bottom of the source page in yyyymmdd numeric format without hyphens or spaces
]=]
function p.ISO_synonym_extract (frame)
local page = mw.title.getCurrentTitle(); -- get a page object for this page
local content = page:getContent(); -- get unparsed content
local content_table = {}; -- table of text lines from source
local split_table = {}; -- table of lines split at the tabs
local skip_table = {}; -- table of 636-2/639-2T codes that have been handled; used to prevent duplication
local out_table = {}; -- output table
local file_date = 'File-Date: ' .. frame.args["file-date"]; -- set the file date line from |file-date= (from the bottom of the source page)
content_table = mw.text.split (content, '[\r\n]'); -- make a table of text lines
for _, line in ipairs (content_table) do -- for each line
split_table = mw.text.split (line, '\t'); -- split at the table
if split_table[5] and (' ' ~= split_table[5]) then -- if there is a 639-1 code
local code = split_table[4]:match ('%a+/(%a+)') or split_table[4]; -- when 639-2B/639-2T use 639-2T else use 639-2
if not skip_table[code] then -- skip if code already in the skip table because more than one language name
skip_table[code] = true; -- remember that we've handled this 636-2/639-2T code
table.insert (out_table, "[\"" .. code .. "\"] = \"" .. split_table[5] .. "\""); -- make new table entry
end
end
end
table.sort (out_table);
return "<br /><pre>-- " .. file_date .. "<br />return {<br />	" .. table.concat (out_table, ',<br />	') .. "<br />	}<br />" .. "</pre>";
end
return p;
