Module:Sandbox/Erutuon: Difference between revisions

Revision as of 00:26, 7 July 2018

Revision as of 01:44, 7 July 2018

Line 445:

})

-- An array of patterns for each subtag, and a "type" field for the name

-- of the subtag.

-- The patterns are checked in order, and any of the subtags can be skipped.

-- So, for example, the "language" subtag must precede the "script"

-- subtag, but a tag may contain a "language" subtag, no "script" subtag

-- and then a "region" subtag.

-- If the full list of subtags has been iterated over, the remaining subtags

-- must match the pattern for a private-use subtag, or the tag is invalid.

local subtag_info = { -- can be put in data module

{ "%a%a%a?", "1%a+", type = "language" }, -- ll or lll; special case

-- include extlang?

{ "%a%a%a%a", type = "script" }, -- Ssss

{ "%a%a", "%d%d%d", type = "region" }, -- rr, DDD

{

"%d%d%d%d", -- 4 digits

"%w%w%w%w%w%w?%w?%w?", -- 5-8 alnum characters

type = "variant",

repeatable = true, -- There can be multiple variants.

}

}

-- A previous draft, in [[Module:Lang/sandbox]]:

Line 481 ⟶ 501:

-- "invalid" is the portion of the tag after the last valid subtag (minus a

-- hyphen).

local potential_subtagssegments = mw.text.split(tag, "-")

local parsed_subtags = parsed_subtags_mt(potential_subtagssegments)

-- Language tags probably only contain ASCII alphabetic and numerical

Line 493 ⟶ 513:

return tag:find "[^A-Za-z0-9-]"

end,

potential_subtagssegments))

end

local subtag_i = 1 -- Index of current item in subtag_info.

-- An array of patterns for each subtag, and a "type" field for the name

local segment_i = 1 -- Index of current segment.

-- of the subtag.

while segments[segment_i] and subtag_info[subtag_i] do

-- The patterns are checked in order, and any of the subtags can be skipped.

local segment = segments[segment_i]

-- So, for example, the "language" subtag must precede the "script"

-- subtag, but a tag may contain a "language" subtag, no "script" subtag

-- and then a "region" subtag.

-- If the full list of subtags has been iterated over, the remaining subtags

-- must match the pattern for a private-use subtag, or the tag is invalid.

local subtag_info = { -- can be put in data module

{ "%a%a%a?", "1%a+", type = "language" }, -- ll or lll; special case

-- include extlang?

{ "%a%a%a%a", type = "script" }, -- Ssss

{ "%a%a", "%d%d%d", type = "region" }, -- rr, DDD

{

"%d%d%d%d", -- 4 digits

"%w%w%w%w%w%w?%w?%w?", -- 5-8 alnum characters

type = "variant"

}

}

local index = 1

local last_matched_subtag_i = 0

for subtag_i, subtag in ipairs(potential_subtags) do

local subtag_type

while not subtag_type and subtag_info[subtag_i] do

local matched = false

-- Check each pattern for the subtag type at "subtag_i" in "subtag_info".

while not matched do

local cur_subtag = subtag_info[subtag_i]

-- Check each pattern for the subtag type at "index" in "subtag_info".

for _, pattern in ipairs(subtag_info[index]cur_subtag) do

if subtagsegment:find("^" .. pattern .. "$") then

subtag_type = subtag_info[index]cur_subtag.type

matched = true

-- There can be multiple "variant" subtags (and "extension"

-- subtags, if those are added).

if subtag_typenot ~= "variant"cur_subtag.repeatable then

indexsubtag_i = indexsubtag_i + 1

end

break

end

end

if not matched then -- Go to next item in subtag_info.

if not subtag_type then -- No match; try next subtag.

index = index + 1

subtag_i = subtag_i + 1

if not subtag_info[index] then

break

end

end

end

-- If language subtag has not been found, or the current segment has not

if subtag_i == 1 and subtag_type ~= "language" then

-- been matched as a subtag, break the loop and check for

-- a private-use subtag.

if segment_i == 1 and subtag_type ~= "language" or not subtag_type then

break

else

elseif subtag_type then

if parsed_subtags[subtag_type] then -- Create an array.

if type(parsed_subtags[subtag_type]) == "string" then

parsed_subtags[subtag_type] = { parsed_subtags[subtag_type] }

end -- else table

table.insert(parsed_subtags[subtag_type], subtagsegment)

else

parsed_subtags[subtag_type] = subtagsegment

end

last_matched_segment_i = segment_i

last_matched_subtag_i = subtag_i

elseif not subtag_info[index] then

break

end

segment_i = segment_i + 1

end

if segments[segment_i] then -- More segments to scan?

if last_matched_subtag_i < #potential_subtags then

-- Not all potential subtags were matched. Check for private-use subtags.

-- https://tools.ietf.org/html/bcp47#section-2.2.7

Line 567:

-- alphanumeric characters preceded by "x-".

-- Alphanumericity has already been checked.

if potential_subtags[last_matched_subtag_i + 1]

-- A tag must start with either a language subtag or a private-use subtag.

and potential_subtags[last_matched_subtag_i + 1]:lower() ~= "x" then

-- If next segment is not "x", introducing a private-use subtag, there

-- is no private-use subtag.

if segments[segment_i] and segments[segment_i]:lower() ~= "x" then

if not parsed_subtags.language then

return parsed_subtags:throw("no_language", 1)

else

return parsed_subtags:throw("invalid_subtag",

segment_i)

last_matched_subtag_i + 1)

end

elseif not segments[segment_i + 1] then

end

-- Check length of all following subtags.

if not potential_subtags[last_matched_subtag_i + 2] then

return parsed_subtags:throw("empty_private_use",

segment_i)

last_matched_subtag_i + 1)

end

-- Check length of all segments after "x".

for i = last_matched_subtag_i + 2, #potential_subtags do

for i = segment_i + 1, #segments do

local length = #potential_subtags[i]

local length = #segments[i]

if not (1 <= length and length <= 8) then

return parsed_subtags

:throw("invalid_private_use", segment_i)

last_matched_subtag_i + 1)

end

end

if not potential_subtagssegments[last_matched_subtag_ilast_matched_segment_i + 3] then -- There is only one private-use subtag.

parsed_subtags.private_use = potential_subtagssegments[last_matched_subtag_isegment_i + 21]

else

parsed_subtags.private_use = {}

for i = last_matched_subtag_isegment_i + 21, #potential_subtagssegments do

table.insert(parsed_subtags.private_use, potential_subtagssegments[i])

end

end

Line 616:

-- everything else is lowercase.

-- Check existence of language tag.

if parsed_subtags.language and

not (lang_data.override[parsed_subtags.language]

Line 622 ⟶ 623:

end

-- Check existence of script tag.

if parsed_subtags.script then

iflocal notlower_script = lang_name_table.script[parsed_subtags.script:lower()] then

if not lang_name_table.script[lower_script] then

mw.log("Invalid script code", parsed_subtags.script, "in", parsed_subtags:get_tag())

end

-- Check that script tag is not marked as superfluous (because the

local lower_script = parsed_subtags.script:lower()

-- it is considered the default one for the language).

if lang_name_table.suppressed[lower_script]

and parsed_subtags.language

and m_table.inArray(

lang_name_table.suppressed[lower_script],

Line 636 ⟶ 641:

end

-- Check existence of region code..

if parsed_subtags.region and not lang_name_table.region[parsed_subtags.region:lower()] then

mw.log("Invalid region code", parsed_subtags.region, "in", parsed_subtags:get_tag())

end

-- Check that variant code is valid, and that it can validly be used with the

-- given combination of language, script, region, and variant.

-- Check for duplicate variant subtags?

if parsed_subtags.variant then

local lower_tag = parsed_subtags:get_tag():lower()

Line 660 ⟶ 669:

end

-- Check that the private-use subtag is actually used by Wikipedia.

if parsed_subtags.private_use and not (type(parsed_subtags.private_use) == "string"

and lang_data.override[parsed_subtags.tag]) then