Content deleted Content added
Copy changes by User:Tholme from Module:String2/sandbox per edit request. |
sync from sandbox; Tag: Reverted |
||
Line 1:
require ('strict');
local p = {}
Line 12 ⟶ 13:
p.ucfirst = function (frame )
local s =
if not s or '' == s or s:match ('^%s+$') then -- when <s> is nil, empty, or only whitespace
return s; -- abandon because nothing to do
end
s = mw.text.trim( frame.args[1] or "" )
local s1 = ""
local prefix_patterns_t = { -- sequence of prefix patterns
'\127[^\127]*UNIQ%-%-%a+%-%x+%-QINU[^\127]*\127', -- stripmarker
'^([%*;:#]+)', -- various list markup
'^(\'\'\'*)', -- bold / italic markup
'^(%b<>)', -- html-like tags because some templates render these
'^(&%a+;)', -- html character entities because some templates render these
'^(&#%d+;)', -- html numeric (decimal) entities because some templates render these
'^(&#x%x+;)', -- html numeric (hexadecimal) entities because some templates render these
'^(%s+)', -- any whitespace characters
'^([%(%)%-%+%?%.%%!~!@%$%^&_={}/`,‘’„“”ʻ|\"\'\\]+)', -- miscellaneous punctuation
}
local prefixes_t = {}; -- list, bold/italic, and html-like markup, & whitespace saved here
-- local function prefix_strip (s, prefixes_t) -- local function to strip prefixes from <s>
local function prefix_strip (s) -- local function to strip prefixes from <s>
for _, pattern in ipairs (prefix_patterns_t) do -- spin through <prefix_patterns_t>
if s:match (pattern) then -- when there is a match
local prefix = s:match (pattern); -- get a copy of the matched prefix
table.insert (prefixes_t, prefix); -- save it
s = s:sub (prefix:len() + 1); -- remove the prefix from <s>
return s, true; -- return <s> without prefix and flag; force restart at top of sequence because misc punct removal can break stripmarker
end
end
return s; -- no prefix found; return <s> with nil flag
end
-- local prefixes_t = {}; -- list, bold/italic, and html-like markup, & whitespace saved here
local prefix_removed; -- flag; boolean true as long as prefix_strip() finds and removes a prefix
repeat -- one by one remove list, bold/italic, html-like markup, whitespace, etc from start of <s>
-- s, prefix_removed = prefix_strip (s, prefixes_t);
s, prefix_removed = prefix_strip (s);
until (not prefix_removed); -- until <prefix_removed> is nil
s1 = table.concat (prefixes_t); -- recreate the prefix string for later reattachment
-- if it's a list chop off and (store as s1) everything up to the first <li>
-- why just list markup? Could be any sort of html markup:
-- <span title="Spanish-language text"><i lang="es">casa</i></span> (this from {{lang|es|casa}})
-- what about wikitext italic or bold markup? (''italic text'', '''bold text''') – both may be legitimate
--
-- if lipos then
-- s = mw.ustring.sub(s, lipos + 4) -- get everything after the <li> tag
-- end
-- -- s1 is either "" or the first part of the list markup, so we can continue
-- -- and prepend s1 to the returned string
local first_text = mw.ustring.match (s, '^%[%[[^%]]+%]%]'); -- extract wikilink at start of string if present; TODO: this can be string.match()?
local upcased;
if
if first_text:match ('^%[%[[^|]+|[^%]]+%]%]') then -- if <first_text> is a piped link
else -- here when <first_text> is a wikilink but not a piped link
upcased = mw.ustring.match (s, '^%[%[%W*%w'); -- get '[[' and first letter
upcased = mw.ustring.upper (upcased); -- upcase first letter character
s = mw.ustring.gsub (s, '^%[%[%W*%w', upcased); -- replace; no capture needed here
end
elseif s:match ('^%[%S+%s+[^%]]+%]') then -- if <s> is a ext link of some sort; must have label text
upcased = mw.ustring.match (s, '^%[%S+%s+%W*(%w)'); -- get first letter character
upcased = mw.ustring.upper (upcased); -- upcase first letter character
s = mw.ustring.gsub (s, '^(%[%S+%s+%W*)%w', '%1' .. upcased); -- replace
elseif s:match ('^%[%S+%s*%]') then -- if <s> is a ext link without label text; nothing to do
return s1 .. s; -- reattach prefix string (if present) and done
else -- <s> is not a wikilink or ext link; assume plain text
upcased = mw.ustring.match (s, '^%W*%w'); -- get the first letter character
upcased = mw.ustring.upper (upcased); -- upcase first letter character
s = mw.ustring.gsub (s, '^%W*%w', upcased); -- replace; no capture needed here
end
return s1 .. s; -- reattach prefix string (if present) and done
-- if mw.ustring.find(s, "^%[%[[^|]+|[^%]]+%]%]") then
-- -- this is a piped wikilink, so we capitalise the text, not the pipe
-- local _
-- _, letterpos = mw.ustring.find(s, "|%W*%w") -- find the first letter after the pipe
-- else
-- letterpos = mw.ustring.find(s, '%w')
-- end
-- if letterpos then
-- local first = mw.ustring.sub(s, 1, letterpos - 1)
-- local letter = mw.ustring.sub(s, letterpos, letterpos)
-- local rest = mw.ustring.sub(s, letterpos + 1)
-- return s1 .. first .. mw.ustring.upper(letter) .. rest
-- else
-- return s1 .. s
-- end
end
Line 204 ⟶ 282:
-- Escape Pattern helper function so that all characters are treated as plain text, as per Module:String
function p._escapePattern( pattern_str )
return mw.ustring.gsub( pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1" )
end
| |||