మాడ్యూల్:Citation/CS1: కూర్పుల మధ్య తేడాలు
Content deleted Content added
చి change link_param_ok (value) to allow underscore; |
Synch from sandbox; |
||
పంక్తి 184:
--[=[
Does this thing that purports to be a domain name seem to be a valid domain name?
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
BNF defined here: https://tools.ietf.org/html/rfc4234
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
see also [[Single-letter second-level domain]]
list of tlds: https://www.iana.org/domains/root/db
rfc952 (modified by rfc 1123) requires the first and last character of a hostname to be a letter or a digit. Between
the first and last characters the name may use letters, digits, and the hyphen
Also allowed are IPv4 addresses. IPv6 not supported
domain is expected to be stripped of any path so that the last character in the last character of the tld. tld
is two or more alpha characters. Any preceding '//' (from splitting a url with a scheme) will be stripped
here. Perhaps not necessary but retained incase it is necessary for IPv4 dot decimal.
There are several tests:
the first character of the whole domain name including subdomains must be a letter or a digit
single-letter/digit second-level domains in the .org TLD
q, x, and z SL domains in the .com TLD
i and q SL domains in the .net TLD
single-letter SL domains in the ccTLDs (where the ccTLD is two letters)
two-character SL domains in gTLDs (where the gTLD is two or more letters)
three-plus-character SL domains in gTLDs (where the gTLD is two or more letters)
IPv4 dot-decimal address format; TLD not allowed
returns true if domain appears to be a proper name and tld or IPv4 address, else false
]=]
local function is_domain_name (domain)
Line 209 ⟶ 224:
domain = domain:gsub ('^//', ''); -- strip '//' from domain name if present; done here so we only have to do it once
if not domain:match ('^[%a%d]
return false;
end
if domain:match ('%f[%a%d][%a%d]%.org$') then -- one character .org hostname
return true;
elseif domain:match ('%f[%a][qxz]%.com$') then -- assigned one character .com hostname (x.com times out 2015-12-10)
return true;
elseif domain:match ('%f[%a][iq]%.net$') then -- assigned one character .net hostname (q.net registered but not active 2015-12-10)
return true;
elseif domain:match ('%f[%a%d][%a%d]%.%a%a$') then -- one character hostname and cctld (2 chars)
return true;
elseif domain:match ('%f[%a%d][%a%d][%a%d]%.%a%a+$') then -- two character hostname and tld
return true;
elseif domain:match ('
return true;
elseif domain:match ('^%d%d?%d?%.%d%d?%d?%.%d%d?%d?%.%d%d?%d?') then -- IPv4 address
Line 242 ⟶ 269:
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
Split a url into a scheme,
If protocol relative url, return nil scheme and domain else return nil for both scheme and domain.
When not protocol relative, get scheme, authority indicator, and domain. If there is an authority indicator (one
or more '/' characters following the scheme's colon), make sure that there are only 2.
]]
local function split_url (url_str)
local scheme, authority, domain;
domain = url_str:match ('^//(%S*)')
elseif url_str:match ('%S-:/*%S+') then -- if there is what appears to be a scheme, optional authority indicator, and domain name
scheme, authority, domain = url_str:match ('(%S-:)(/*)(%S+)'); -- extract the scheme, authority indicator, and domain portions
authority = authority:gsub ('//', '', 1); -- replace place 1 pair of '/' with nothing;
if is_set(authority) then -- if anything left (1 or 3+ '/' where authority should be) then
domain = nil; -- set to nil which will cause an error message
end
end
Line 266 ⟶ 302:
Link parameters are to hold the title of a wikipedia article so none of the WP:TITLESPECIALCHARACTERS are allowed:
# < > [ ] | { } _
except the underscore which is used as a space in wiki urls and # which is used for section links
returns false when the value contains any of these characters.
Line 277 ⟶ 313:
local function link_param_ok (value)
local scheme, domain;
if value:find ('[
return false;
end
Line 291 ⟶ 327:
First we test for space characters. If any are found, return false. Then split the url into scheme and domain
portions, or for protocol relative (//example.com) urls, just the domain. Use
]]
Line 310 ⟶ 346:
Return true if a parameter value has a string that begins and ends with square brackets [ and ] and the first
non-space characters following the opening bracket
The frontier pattern prevents a match on interwiki links which are similar to scheme:path urls. The tests that
find bracketed urls are required because the parameters that call this test (currently |title=, |chapter=,
and |
is possible as might be [[en://Hus]].
]=]
Line 323 ⟶ 359:
local scheme, domain;
scheme, domain = value:match ('%f[%[]%[
elseif value:match ('%f[%[]%[//%S*%.%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
domain = value:match ('%f[%[]%[//(%S*%.%S+).*%]');
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
scheme, domain = value:match ('(%a%S*:)(%S+)');
elseif value:match ('//%S*%.%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
domain = value:match ('//(%S*%.%S+)'); -- what is left should be the domain
else
end
Line 351 ⟶ 393:
end
end
if is_set (error_message) then
table.insert( z.message_tail, { set_error( 'param_has_ext_link', {error_message}, true ) } );
end
Line 620 ⟶ 662:
if value ~= nil and selected ~= alias then -- if we have already selected one of the aliases
local skip;
for _, v in ipairs(error_list) do
if v == alias then
skip = true;
break;
end
end
if not skip then
table.insert( error_list, alias );
end
else
Line 726 ⟶ 768:
end
--[[--------------------------< H A S _ I N V I S I B L E _ C H A R S >----------------------------------------
This function searches a parameter's value for nonprintable or invisible characters. The search stops at the
first match.
This function will detect the visible replacement character when it is part of the wikisource.
Detects but ignores nowiki and math stripmarkers. Also detects other named stripmarkers (gallery, math, pre, ref)
and identifies them with a slightly different error message. See also coins_cleanup().
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker
that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the
parameter value.
]]
local function has_invisible_chars (param, v)
local position = ''; -- position of invisible char or starting position of stripmarker
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
local capture; -- used by stripmarker detection to hold name of the stripmarker
local i=1;
local stripmarker, apostrophe;
while cfg.invisible_chars[i] do
local char=cfg.invisible_chars[i][1] -- the character or group name
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
position, dummy, capture = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
if position then
-- if 'nowiki' == capture or 'math' == capture or ('ref' == capture and 'quote' == param) then -- nowiki, math, or quote param and ref stripmarker (not an error condition)
if 'nowiki' == capture or 'math' == capture then -- nowiki, math stripmarker (not an error condition)
stripmarker = true; -- set a flag
elseif true == stripmarker and 'delete' == char then -- because stripmakers begin and end with the delete char, assume that we've found one end of a stripmarker
position = nil; -- unset
elseif 'apostrophe' == char then -- apostrophe template uses ‍, hair space and zero-width space
apostrophe = true;
elseif true == apostrophe and in_array (char, {'zero width joiner', 'zero width space', 'hair space'}) then
position = nil; -- unset
else
local err_msg;
if capture then
err_msg = capture .. ' ' .. char;
else
err_msg = char .. ' ' .. 'character';
end
table.insert( z.message_tail, { set_error( 'invisible_char', {err_msg, wrap_style ('parameter', param), position}, true ) } ); -- add error message
return; -- and done with this parameter
end
end
i=i+1; -- bump our index
end
end
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
Argument wrapper. This function provides support for argument mapping defined in the configuration file so that
multiple names can be transparently aliased to single internal variable.
]]
Line 775 ⟶ 874:
end
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
Looks for a parameter's name in the whitelist.
Line 782 ⟶ 881:
false - deprecated, supported parameters
nil - unsupported parameters
]]
Line 1,904 ⟶ 2,004:
end
--[[--------------------------< C O I N S _ C L E A N U P >----------------------------------------------------
Cleanup parameter values for the metadata by removing or replacing invisible characters and certain html entities.
2015-12-10: there is a bug in mw.text.unstripNoWiki (). It replaced math stripmarkers with the appropriate content
when it shouldn't. See https://phabricator.wikimedia.org/T121085 and Wikipedia_talk:Lua#stripmarkers_and_mw.text.unstripNoWiki.28.29
TODO: move the replacement patterns and replacement values into a table in /Configuration similar to the invisible
characters table?
]]
local function coins_cleanup (value)
value = mw.text.unstripNoWiki (value); -- replace nowiki stripmarkers with their content
value = value:gsub ('<span class="nowrap" style="padding%-left:0%.1em;">'s</span>', "'s"); -- replace {{'s}} template with simple apostrophe-s
value = value:gsub ('‍\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe
value = value:gsub ('\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe (as of 2015-12-11)
value = value:gsub (' ', ' '); -- replace entity with plain space
value = value:gsub ('\226\128\138', ' '); -- replace hair space with plain space
value = value:gsub ('‍', ''); -- remove ‍ entities
value = value:gsub ('[\226\128\141\226\128\139]', '') -- remove zero-width joiner, zero-width space
value = value:gsub ('[\194\173\009\010\013]', ' '); -- replace soft hyphen, horizontal tab, line feed, carriage return with plain space
return value;
end
--[[--------------------------< C O I N S >--------------------------------------------------------------------
Line 1,915 ⟶ 2,040:
return '';
end
for k, v in pairs (data) do -- spin through all of the metadata parameter values
if 'ID_list' ~= k and 'Authors' ~= k then -- except the ID_list and Author tables (author nowiki stripmarker done when Author table processed)
data[k] = coins_cleanup (v);
end
end
local ctx_ver = "Z39.88-2004";
Line 2,013 ⟶ 2,144:
local last, first;
for k, v in ipairs( data.Authors ) do
last, first = coins_cleanup (v.last), coins_cleanup (v.first or ''); -- replace any nowiki strip markers, non-printing or invisible characers
if k == 1 then -- for the first author name only
if is_set(last) and is_set(first) then -- set these COinS values if |first= and |last= specify the first author name
Line 2,858 ⟶ 2,989:
ID = A['Number']; -- yes, use it
else -- ID has a value so emit error message
table.insert( z.message_tail, { set_error('redundant_parameters', {wrap_style ('parameter', 'id') .. ' and ' .. wrap_style ('parameter', 'number')}, true )});
end
Line 3,872 ⟶ 4,002:
]]
--[[
local function has_invisible_chars (param, v)
local position = '';
Line 3,880 ⟶ 4,010:
local char=cfg.invisible_chars[i][1] -- the character or group name
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
v = mw.text.unstripNoWiki( v ); -- remove nowiki stripmarkers
position = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
if position then
Line 3,888 ⟶ 4,019:
end
end
]]
--[[--------------------------< Z . C I T A T I O N >----------------------------------------------------------
Line 3,970 ⟶ 4,101:
for k, v in pairs( args ) do
if 'string' == type (k) then -- don't evaluate positional parameters
has_invisible_chars (k, v);
end
end
return citation0( config, args)
| |||