Content deleted Content added
No edit summary |
No edit summary |
||
| (53 intermediate revisions by 8 users not shown) | |||
Line 1:
--[[
History of changes since last sync:
2025-08-01: fix 8-digit only medRxiv test; see Help_talk:Citation_Style_1#medRxiv_error_detection_flaw
]]
Line 34 ⟶ 36:
as an aid to internationalizing identifier-label wikilinks, gets identifier article names from Wikidata.
returns w:<lang code>:<article title> when <q> has an <article title> for <lang code>; nil else. 'w:<lang code>'
ensures that sister project (like wiktionary) will link to the <lang code>.wikipedia article.
for identifiers that do not have <q>, returns nil
for wikis that do not have mw.wikibase installed, returns nil
Line 53 ⟶ 56:
if wd_article then
wd_article = table.concat ({'w:', this_wiki_code, ':', wd_article});
end
Line 60 ⟶ 63:
--[[--------------------------< L
common function to create
returns the first available of:
1. redirect from local wiki's handler table (if enabled)
2. Wikidata sitelink to the local language wikipedia article (if there is a Wikidata entry for this identifier in the local
3.
]]
local function
local wd_article;
Line 108 ⟶ 111:
return table.concat ({
make_wikilink (
options.separator or ' ',
ext_link
Line 130 ⟶ 133:
return table.concat (
{
make_wikilink (
options.separator or ' ', -- add the separator
make_wikilink (
Line 173 ⟶ 176:
--[=[-------------------------< I S _ V A L I D _
for biorxiv, returns true if:
2019-12-11T00:00Z <= biorxiv_date < today + 2 days
for medrxiv, returns true if:
2020-01-01T00:00Z <= medrxiv_date < today + 2 days
The dated form of biorxiv identifier has a start date of 2019-12-11. The Unix timestamp for that date is {{#time:U|2019-12-11}} = 1576022400
The medrxiv identifier has a start date of 2020-01-01. The Unix timestamp for that date is {{#time:U|2020-01-01}} = 1577836800
<today> is the current date at time 00:00:00 UTC plus 48 hours
if today's date is
adding 24 hours gives
adding 24 hours gives
inputs:
<y>, <m>, <d> – year, month, day parts of the date from the birxiv or medrxiv identifier
<select> 'b' for biorxiv, 'm' for medrxiv; defaults to 'b'
]=]
local function
if 0 == tonumber (m) and 12 < tonumber (m) then -- <m> must be a number 1–12
return false;
end
if 0 == tonumber (d) and 31 < tonumber (d) then -- <d> must be a number 1–31; TODO: account for month length and leap yer?
return false;
end
local rxiv_date = table.concat ({y, m, d}, '-'); -- make ymd date string
local good1, good2;
local
local lang_object = mw.getContentLanguage();
good1,
good2, tomorrow_ts = pcall (lang_object.formatDate, lang_object, 'U', 'today + 2 days' ); -- today midnight + 2 days is one second more than all day tomorrow
if good1 and good2 then -- lang.formatDate() returns a timestamp in the local script which tonumber() may not understand
tomorrow_ts = tonumber (tomorrow_ts) or lang_object:parseFormattedNumber (tomorrow_ts);
else
Line 208 ⟶ 221:
end
local limit_ts = ((select and ('m' == select)) and 1577836800) or 1576022400; -- choose the appropriate limit timesatmp
return ((limit_ts <= rxiv_ts) and (rxiv_ts < tomorrow_ts)) -- limit_ts <= rxiv_date < tomorrow's date
end
Line 258 ⟶ 273:
--[[--------------------------< N O R M A L I Z E _ L C C N >--------------------------------------------------
LCCN normalization (
1. Remove all blanks.
2. If there is a forward slash (/) in the string, remove it, and remove all characters to the right of the forward slash.
Line 268 ⟶ 283:
Returns a normalized LCCN for lccn() to validate. There is no error checking (step 3.b.1) performed in this function.
]]
Line 287 ⟶ 303:
return lccn;
Line 294 ⟶ 310:
--[[--------------------------< A R X I V >--------------------------------------------------------------------
See:
format and error check arXiv identifier. There are three valid forms of the identifier:
Line 319 ⟶ 335:
<date code> and <version> are as defined for 0704-1412
<number> is a five-digit number
]]
Line 373 ⟶ 390:
if is_set (class) then
if id:match ('^%d+') then
text = table.concat ({text, ' [[https://arxiv.org/archive/', class, ' ', class, ']]'}); -- external link within square brackets, not wikilink
else
set_message ('err_class_ignored');
end
else -- class not set
if id:match ('^%d+') and options.CitationClass == 'arxiv' then -- new (post 2007) format; {{cite arxiv}} only
set_message ('maint_missing_class'); -- add maint cat
end
end
Line 387 ⟶ 408:
Validates (sort of) and formats a bibcode ID.
Format for bibcodes is specified here:
But, this: 2015arXiv151206696F is apparently valid so apparently, the only things that really matter are length, 19 characters
Line 405 ⟶ 426:
local access = options.access;
local handler = options.handler;
local ignore_invalid = options.accept;
local err_type;
local err_msg = '';
Line 427 ⟶ 449:
if id:find('&%.') then
err_type = cfg.err_msg_supl.journal; -- journal abbreviation must not have '&.' (if it does it's missing a letter)
end
if id:match ('.........%.tmp%.') then -- temporary bibcodes when positions 10–14 are '.tmp.'
set_message ('maint_bibcode');
end
end
end
if is_set (err_type) and not ignore_invalid then
set_message ('err_bad_bibcode', {err_type});
options.coins_list_t['BIBCODE'] = nil; -- when error, unset so not included in COinS
end
Line 462 ⟶ 486:
local patterns = {
'^10%.1101/%d%d%d%d%d%d$', -- simple 6-digit identifier (before 2019-12-11)
'^10%.1101/(20
'^10%.1101/(20
}
Line 472 ⟶ 496:
if m then -- m is nil when id is the six-digit form
if not
break; -- date fail; break out early so we don't unset the error message
end
Line 479 ⟶ 503:
break; -- and done
end
end -- err_cat remains set here when no match
if err_msg then
Line 497 ⟶ 521:
The description of the structure of this identifier can be found at Help_talk:Citation_Style_1/Archive_26#CiteSeerX_id_structure
]]
Line 532 ⟶ 557:
and terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely
if ever used in DOI names.
https://www.doi.org/doi_handbook/2_Numbering.html -- 2.2 Syntax of a DOI name
https://www.doi.org/doi_handbook/2_Numbering.html#2.2.2 -- 2.2.2 DOI prefix
]]
Line 542 ⟶ 570:
local handler = options.handler;
local err_flag;
local function is_extended_free (registrant, suffix) -- local function to check those few registrants that are mixed; identifiable by the doi suffix <incipit>
if cfg.extended_registrants_t[registrant] then -- if this registrant has known free-to-read extentions
for _, incipit in ipairs (cfg.extended_registrants_t[registrant]) do -- loop through the registrant's incipits
if mw.ustring.find (suffix, '^' .. incipit) then -- if found
return true;
end
end
end
end
local text;
if is_set (inactive) then
local inactive_year = inactive:match("%d%d%d%d")
local inactive_month, good;
Line 556 ⟶ 594:
end
end
end -- otherwise, |doi-broken-date= has something but it isn't a date
if is_set (inactive_year) and is_set (inactive_month) then
Line 570 ⟶ 606:
end
local suffix;
local registrant, suffix = mw.ustring.match (id, '^10%.([^/]+)/([^%s–]-[^%.,])$'); -- registrant and suffix set when DOI has the proper basic form
local registrant_err_patterns = { -- these patterns are for code ranges that are not supported
'^[^1-3]%d%d%d%d%.%d
'^[^1-
'^[^1-9]%d%d%d%.%d
'^[^1-9]%d%d%d$', -- 4 digits without subcode (0xxx); accepts: 1000–9999
'^%d%d%d%d%d%d+', -- 6 or more digits
'^%d%d?%d?$', -- less than 4 digits without subcode (3 digits with subcode is legitimate)
'^%d%d?%.[%d%.]+', -- 1 or 2 digits with subcode
'^5555$', -- test registrant will never resolve
'[^%d%.]', -- any character that isn't a digit or a dot
Line 600 ⟶ 638:
if err_flag then
options.coins_list_t['DOI'] = nil; -- when error, unset so not included in COinS
else
if not access and (cfg.known_free_doi_registrants_t[registrant] or is_extended_free (registrant, suffix)) then -- |doi-access=free not set and <registrant> is known to be free
set_message ('maint_doi_unflagged_free'); -- set a maint cat
end
end
Line 626 ⟶ 668:
if ever used in HDLs.
Query string parameters are named here:
but since '?' is an allowed character in an HDL, '?' followed by one of the query parameters is the only way we
have to detect the query string so that it isn't URL-encoded with the rest of the identifier.
Line 636 ⟶ 678:
local access = options.access;
local handler = options.handler;
local query_params = { -- list of known query parameters from
'noredirect',
'ignore_aliases',
Line 684 ⟶ 726:
]]
local function isbn (
local isbn_str =
local ignore_invalid =
local handler =
local year = options_t.Year; -- when set, valid anchor_year; may have a disambiguator which must be removed
local function return_result (check, err_type) -- local function to handle the various returns
Line 696 ⟶ 739:
else -- here when not ignoring
if not check then -- and there is an error
set_message ('err_bad_isbn', err_type); -- set an error message
return ISBN; -- return id text
Line 702 ⟶ 745:
end
return ISBN; -- return id text
end
if year and not ignore_invalid then --
year = year:match ('%d%d%d%d?'); -- strip disambiguator if present
if 1965 > tonumber(year) then
set_message ('err_invalid_isbn_date'); -- set an error message
return internal_link_id ({link = handler.link, label = handler.label, redirect = handler.redirect,
prefix = handler.prefix, id = isbn_str, separator = handler.separator});
end
end
Line 805 ⟶ 857:
Determines whether an ISMN string is valid. Similar to ISBN-13, ISMN is 13 digits beginning 979-0-... and uses the
same check digit calculations. See
section 2, pages 9–12.
Line 834 ⟶ 886:
text = table.concat ( -- because no place to link to yet
{
make_wikilink (
handler.separator,
id_copy
Line 854 ⟶ 906:
like this:
|issn=0819 4327 gives: [
This code now prevents that by inserting a hyphen at the ISSN midpoint. It also validates the ISSN for length
Line 958 ⟶ 1,010:
Format LCCN link and do simple error checking. LCCN is a character string 8-12 characters long. The length of
the LCCN dictates the character type of the first 1-3 characters; the rightmost eight are always digits.
https://oclc-research.github.io/infoURI-Frozen/info-uri.info/info:lccn/reg.html
length = 8 then all digits
Line 1,013 ⟶ 1,065:
return external_link_id ({link = handler.link, label = handler.label, q = handler.q, redirect = handler.redirect,
prefix = handler.prefix, id = lccn, separator = handler.separator, encode = handler.encode});
end
--[[--------------------------< M E D R X I V >-----------------------------------------------------------------
Format medRxiv ID and do simple error checking. Similar to later bioRxiv IDs, medRxiv IDs are prefixed with a
yyyy.mm.dd. date and suffixed with an optional version identifier. Ealiest date accepted is 2020.01.01
The medRxiv ID is a date followed by an eight-digit number followed by an optional version indicator 'v' and one or more digits:
https://www.medrxiv.org/content/10.1101/2020.11.16.20232009v2 -> 10.1101/2020.11.16.20232009v2
]]
local function medrxiv (options)
local id = options.id;
local handler = options.handler;
local err_msg_flag = true; -- flag; assume that there will be an error
local patterns = {
'^%d%d%d%d%d%d%d%d$', -- simple 8-digit identifier; these should be relatively rare
'^10%.1101/(20%d%d)%.(%d%d)%.(%d%d)%.%d%d%d%d%d%d%d%dv%d+$', -- y.m.d. date + 8-digit identifier + version (2020-01-01 and later)
'^10%.1101/(20%d%d)%.(%d%d)%.(%d%d)%.%d%d%d%d%d%d%d%d$', -- y.m.d. date + 8-digit identifier (2020-01-01 and later)
}
for _, pattern in ipairs (patterns) do -- spin through the patterns looking for a match
if id:match (pattern) then
local y, m, d = id:match (pattern); -- found a match, attempt to get year, month and date from the identifier
if m then -- m is nil when id is the 8-digit form
if not is_valid_rxiv_date (y, m, d, 'b') then -- validate the encoded date; 'b' for medrxiv limit
break; -- date fail; break out early so we don't unset the error message
end
end
err_msg_flag = nil; -- we found a match so unset the error message
break; -- and done
end
end -- <err_msg_flag> remains set here when no match
if err_msg_flag then
options.coins_list_t['MEDRXIV'] = nil; -- when error, unset so not included in COinS
set_message ('err_bad_medrxiv'); -- and set the error message
end
return external_link_id ({link = handler.link, label = handler.label, q = handler.q, redirect = handler.redirect,
prefix = handler.prefix, id = id, separator = handler.separator,
encode = handler.encode, access = handler.access});
end
Line 1,074 ⟶ 1,172:
elseif id:match('^%d+$') then -- no prefix
number = id; -- get the number
if
number = nil; --
end
end
Line 1,207 ⟶ 1,305:
text = table.concat ( -- still embargoed so no external link
{
make_wikilink (
handler.separator,
id,
Line 1,536 ⟶ 1,634:
['JSTOR'] = jstor,
['LCCN'] = lccn,
['MEDRXIV'] = medrxiv,
['MR'] = mr,
['OCLC'] = oclc,
Line 1,559 ⟶ 1,658:
options_t.handler = cfg.id_handlers[hkey];
options_t.coins_list_t = ID_list_coins_t; -- pointer to ID_list_coins_t; for |asin= and |ol=; also to keep erroneous values out of the citation's metadata
options_t.coins_list_t[hkey] = v; -- id value without accept-as-written markup for metadata
if options_t.handler.access and not in_array (options_t.handler.access, cfg.keywords_lists['id-access']) then
Line 1,617 ⟶ 1,717:
]]
local function identifier_lists_get (
local ID_list_coins_t = extract_ids (
options_check (ID_list_coins_t, ID_support_t); -- ID support parameters must have matching identifier parameters
local ID_access_levels_t = extract_id_access_levels (
local ID_list_t = build_id_list (ID_list_coins_t, options_t, ID_access_levels_t); -- get a sequence table of rendered identifier strings
Line 1,654 ⟶ 1,754:
auto_link_urls = auto_link_urls, -- table of identifier URLs to be used when auto-linking |title=
identifier_lists_get = identifier_lists_get, -- experiment to replace individual calls to build_id_list(), extract_ids, extract_id_access_levels
is_embargoed = is_embargoed;
set_selected_modules = set_selected_modules;
| |||