Module:書名
 | This module is rated as beta, and is ready for widespread use. It is still new and should be used with some caution to ensure the results are as expected. |
Synopsis[edit]
{{#invoke:書名 |Syu1meng2 [ |mode=chapter ] }}
{{#invoke:書名 |Zyun1ming4 [ |mode=dotted ] }}
{{#invoke:書名 |Zoek6zung6}}
p .auto_build_citable( book title, chapter title );
p .cvs( value );
p .ref( value );
p .cjk_p( string );
p .array_p( value );
Description[edit]
This module implements the backend of several templates designed to typeset a number of Chinese-language punctuation marks that are difficult to typeset through normal means. It currently handles the ˈsyˌmiŋˍhou (book titles), ˈdzynˌmiŋˍhou (personal and geographical names) and ˍdzœkˍdzuŋˍhou (emphasis mark). Compatibility with text browsers and screen readers was a priority.
Typeset output aims to follow print. As a result titles marked with type 2 ˈsyˌmiŋˍhou (i.e., sinicized guillemets) are kerned, which is traditional for similar punctuation (for example quotation marks).[1] The module doesn’t actually do any actual formatting, but will add enough structuring and classes for formatting via CSS.
Parameters are normally retrieved directly from the frame’s parent, except when an entry point is shared between two templates, in which case the discriminating parameter is retrieved from the frame. These currently include mode=chapter (chapter title mode) for Syu1meng2 and mode=dotted (transliterated personal names mode) for Zyun1ming4.
Invocation from other modules[edit]
After importing this module, auto_build_citable can be directly invoked as a method. The method takes 1 to 2 strings (either can be nil) as input and returns a string as output, and is essentially the same as {{#invoke:書名|Syu1meng2}} except no CSS will be requested.
Auxiliaries[edit]
Other methods are available after importing this module. These include
- cvs: takes any value as input, returns string as output. Produces a stringified form of the input that’s more useful (similar to PostScript’s cvs operator)
- cjk_p: takes 1 string as input, returns boolean as output. Determines whether the input strings appears to consist purely of CJK characters.
- array_p: takes any value as input, returns boolean as output. Determines whether the input appears to be an array.
There is also ref, which takes any value as input, and returns string as output. Similar to type except that if the input is an array, it will try to guess whether it’s a structure returned by parse_title (an internal function), and return “parsed-title” if this is the case.
History[edit]
This module was created specifically for the Cantonese Wikipedia and was derived from Javascript and PHP code that was already in use at the author’s own site.
Notes[edit]
This is a mostly complete translation of the original documentation at the Cantonese Wikipedia (in Cantonese). Please refer to the original documentation if you need information that might be more up to date.
The name of the module means “book title”.
See also[edit]
References[edit]
- ↑ See, for example, . Translated by Ān, Píng. "Overview of the Printing and Allied Trades Association in London" 倫敦伯特拉印刷研究所概況. The Graphic Printer 1 (6): 40–47. June 1, 1931. Archived from the original on October 31, 2022. Retrieved October 11, 2022.
-- vi:set sw=4 ts=4 ai sm:
---- This module implements the backend logic for displaying the two types
---- of ˈsyˌmiŋˍhou (punctuation mark for indicating the title of a citable
---- work in Chinese languages)
---- It has also been modified to deal with ˍdzœkˍdzuŋˍhou (emphasis marks,
---- which unfortunately needs to be dealt with despite being a standard part
---- of CSS) and also some experimental support for ˈdzynˌmiŋˍhou (punctuation
---- mark for indicating personal and geographical names)
require('strict');
local p = {};
local TYPE_1 = '甲';
local TYPE_2 = '乙';
--- Auxiliaries ---------------------------------------------------------------
-- Figure out if a thing is an array
local function array_p( thing )
local it = (type(thing) == 'table');
if it then
local i_min, i_max;
for i, v in pairs(thing) do
if type(i) ~= 'number' then
it = false;
else
if i_min == nil or i < i_min then
i_min = i;
end
if i_max == nil or i > i_max then
i_max = i;
end
end
end
-- in PostScript array indexes start at 0, but in Lua they start at 1
-- in addition, Lua "sequences" cannot contain nils (ignoring for now)
if it and not (i_min and i_min == 1 and i_max == #thing) then
it = false; -- index has holes, not a real array
end
end
return it;
end
-- Stringify something into a form suitable for debugging and error messages
-- (cvs is the name of the Postscript operator that does this)
local function cvs( s )
if s == nil or s == false or s == true then
s= tostring(s);
elseif type(s) == 'string' then
s = '(' .. mw.ustring.gsub(s, '([()])', '\\%1') .. ')';
elseif type(s) == 'function' then
s = 'FUNCTION'; -- sigh
elseif type(s) == 'table' then
local s2 = '';
if array_p(s) then
for i = 1, #s, 1 do
if #s2 > 0 then
s2 = s2 .. ' ';
end
s2 = s2 .. cvs(s[i]);
end
s2 = '[' .. s2 .. ']';
else
for i, v in pairs(s) do
if #s2 > 1 then
s2 = s2 .. ' ';
end
if type(i) == 'string' and i:match('^%w%w*$') then
s2 = s2 .. '/' .. i;
else
s2 = s2 .. cvs(i);
end
s2 = s2 .. ' ' .. cvs(v);
end
s2 = '<<' .. s2 .. '>>';
end
s = s2;
end
return s;
end
-- Return a sanitized version of an argument value
local function sanitize( arg )
if arg ~= nil and arg:match('^{{{%w-}}}$') then
arg = nil
end
return arg;
end
-- If the passed string is a null string, turn it into an actual null
local function nullify( s )
if s == '' then
s = nil
end
return s;
end
-- Check if the given string is (believed to be) CJK
local function cjk_p( s )
return mw.ustring.match(s, '^['
.. '—' -- 2014 (em dash)
.. '…' -- 2026
.. '─' -- 2500 (line drawing)
.. '○' -- 25CB (circle [not zero])
.. '⺀-䶿' -- 2E80-4DBF
.. '一-鿿' -- 4E00-9FFF
.. '가-' -- AC00-D7AF
.. '豈-' -- F900-FAFF
.. '︰-﹏' -- FE30-FE4F
.. '！-｠' -- FF01-FF60
.. '𠀀-' -- 20000-2FA1F
.. ']+$');
end
local function space_p( c )
return mw.ustring.match(c, '[ ]'); -- sp, CJK sp
end
local function kernable_left_punctuation_p( c )
return c and mw.ustring.match(c, '[《〈（【「『]');
end
local function kernable_right_punctuation_p( c )
return c and mw.ustring.match(c, '[》〉）】」』]');
end
local function left_kernable_narrow_punctuation_p( c )
return c and mw.ustring.match(c, '[，；：。！]');
end
local function right_kernable_narrow_punctuation_p( c )
return c and mw.ustring.match(c, '[、，；：。！]');
end
-- Canonicalize the type parameter
local function canonicalize_type_value( type )
if type == '1' or type == '甲' then
type = '甲'
elseif type == '2' or type == '乙' then
type = '乙'
end
return type;
end
-- Try to guess what REALLY is the thing, in the context of this module.
-- Like Lisp or Perl, Lua has no real concept of classes, but unlike Perl
-- there's no way to "bless" an array (table in Lua) into a class.
-- So everything is just a generic table - a most non-ideal situation.
-- (ref is the name of the Perl operator that does this)
local function ref( thing )
local it = type(thing);
if thing == nil then
it = nil
elseif it == 'table' then
-- Are we looking at a parsed title?
if thing.label and type(thing.label) == 'string' then
it = 'parsed-title';
-- Are we looking at an array?
elseif array_p(thing) then
it = 'array'
end
end
return it;
end
-- Analyze the title and see if it's a link of some kind
local function parse_title( s )
local it = s;
local t = ref(s);
if t == 'string' then
local label, link, url;
local bold, italic, tmp;
bold, tmp = mw.ustring.match(s, "^(''')(.*)%1$");
if bold then
s = tmp;
end
italic, tmp = mw.ustring.match(s, "^('')(.*)%1$");
if italic then
s = tmp;
end
link, label = mw.ustring.match(s, '^%[%[%s*([^%[][^%[]*)%s*|%s*([^%[%]]*)%s*%]%]$');
if link then
if label and #label == 0 then
label = link;
end
else
link = mw.ustring.match(s, '^%[%[([^%[]+)%]%]$');
if link then
label = link;
else
url, label = mw.ustring.match(s, '^%[(%S+)%s([^%]]+)%]$');
if not url then
label = s;
end
end
end
-- Silent change a few things to make it easier to typeset
label = mw.ustring.gsub(label, '[─]', '—'); -- U+2500 line drawing
label = mw.ustring.gsub(label, '[○]', '〇'); -- U+25CB circle
label = mw.ustring.gsub(label, '⸺', '——'); -- 2-em dash -0> 2x em dash'
it = {
['label'] = label;
['link'] = link;
['url'] = url;
['bold'] = bold;
['italic'] = italic;
};
elseif t and t ~= 'parsed-title' then
error('parse-title got unexpected argument '..cvs(s) .. ' (t=' .. cvs(t) .. ')', 2);
end
return it;
end
-- Inverse of parse_title: Try to reconstruct a wikified link given its parse
-- This does not seem to actually work
local function linkify_title( label, link, url )
local it;
local bold, italic;
if type(label) == 'table' then
link = label.link;
url = label.url;
bold = label.bold;
italic = label.italic;
label = label.label;
end
if label == nil then
-- all is lost
elseif link ~= nil then
it = '[[' .. link .. '|' .. label .. ']]';
elseif url ~= nil then
it = '[' .. url .. ' ' .. label .. ']';
else
it = label;
end
if bold then
it = "'''" .. it .. "'''";
end
if italic then
it = "''" .. it .. "''";
end
return it;
end
-- Return a value that should be suitable for use as an alt or aria-label
-- and safe to use for double-quoting
local function build_aria_label_from( s )
if s == nil then
return s;
end
if type(s) ~= 'string' then
error('ERROR: build_aria_label got a non-string '..cvs(s), 2);
end
return mw.ustring.gsub(
mw.ustring.gsub(
mw.ustring.gsub(
mw.ustring.gsub(
mw.ustring.gsub(
mw.ustring.gsub(s,
'<[^<>]*>', ''), -- try to nuke tags
"'''''", ''), -- Wikitext bold italic
"'''", ''), -- Wikitext bold
"'", ''), -- Wikitext italic
'"', '”'),
"'", '’')
end
local function get_token( t )
local t_next;
local it = {['pre'] = ''; ['c'] = ''; ['post'] = '';};
for stage = 1, 3, 1 do
local formatting = '';
while true do
local t1, t2;
local done_p = false;
if stage == 1 then
t1, t2 = mw.ustring.match(t, '^(<[^/][^<>]*>)(.*)$');
elseif stage == 2 then
t1 = nil;
elseif stage == 3 then
t1, t2 = mw.ustring.match(t, '^(</[^<>]*>)(.*)$');
else
error('Internal error 1, stage='..cvs(stage), 1)
end
if t1 then
formatting = formatting .. t1;
t_next = t2;
elseif stage == 1 or stage == 3 then
t1, t2 = mw.ustring.match(t, "^(''''')(.*)$");
if not t1 then
t1, t2 = mw.ustring.match(t, "^(''')(.*)$")
if not t1 then
t1, t2 = mw.ustring.match(t, "^('')(.*)$");
end
end
if t1 then
formatting = formatting .. t1;
t_next = t2;
else
done_p = true;
end
elseif stage == 2 then
t1, t2 = mw.ustring.match(t, '^([^<])(.*)$');
if t1 then
it.c = it.c .. t1;
t_next = t2;
end
done_p = true;
else
error('Internal error 2, stage='..cvs(stage), 1)
end
if done_p then break end
t = t_next;
end
if formatting then
if stage == 1 then
it.pre = it.pre .. formatting;
else
it.post = it.post .. formatting;
end
end
end
return it, t_next or '';
end
-- Build a type 1 part (not necessarily the entire title)
-- Try to construct the final HTML so that it line wrap correctly
-- while also following kinsokushori rules
local function build_type_1_part( s )
local it;
local stage1 = {};
local opening_p = false;
s = parse_title(s);
assert(ref(s) == 'parsed-title')
it = '';
local t = s.label;
while #t > 0 do
local c, t_next = get_token(t);
local closing_p = mw.ustring.match('[】》〉）｣」』]', c.c);
if opening_p or closing_p then
table.insert(stage1[#stage1], c);
else
table.insert(stage1, {c});
end
opening_p = mw.ustring.match('[【《〈（｢「『]', c.c);
t = t_next;
end
for i = 1, #stage1, 1 do
local stage2 = stage1[i];
local class = 'zit3';
if i == #stage1 then
class = 'hai2zeoi3mei1ge3 '..class;
end
if i == 1 then
class = 'hai2tau4ge3 '..class;
end
it = it .. '<span class="'..class..'">' .. '<span class=zit3>';
for j = 1, #stage2, 1 do
local class = 'zi6';
if j == #stage2 then
class = 'hai2zeoi3mei1ge3 '..class;
end
if j == 1 then
class = 'hai2tau4ge3 '..class;
end
local s = stage2[j];
local c = table.concat({s.pre or '', s.c, s.post or ''});
it = it .. '<span class="'..class..'">' .. c .. '</span>';
end
it = it .. '</span>' .. '</span>';
end
--it = 'DEBUG: stage1 = ' .. cvs(stage1) .. '<br>→ it = ' .. cvs(it);
s.label = it;
it = linkify_title(s);
return it;
end
-- Try to kern the given string
-- The logic is a deterministic finite state machine. The DFA diagram is
-- currently on a piece of paper and will be added here later.
local function kern( s0 )
local it;
local s = parse_title(s0);
assert(ref(s) == 'parsed-title');
local t = s.label;
local segment;
local STATE = { ['INITIAL'] = 'INITIAL';
['OPENING'] = 'OPENING';
['CLOSING'] = 'CLOSING';
};
local state = STATE.INITIAL;
local function tag_p( s )
return s and s:sub(1, 1) == '<';
end
local function remember_kerned( c, class, segment, state )
if segment == nil then
segment = {};
end
table.insert(segment, { ['c'] = c;
['class'] = class;
['state'] = state; });
return segment;
end
local function remember_unkerned(c, segment, state)
if segment == nil then
segment = {};
end
if #segment == 0
or segment[#segment].class
or segment[#segment].post
or c.pre then
table.insert(segment, { ['c'] = c;
['state'] = state; } );
else
segment[#segment].c.c = table.concat({segment[#segment].c.c, c.c});
segment[#segment].c.post = c.post;
end
return segment;
end
local function change_class_of_last_remembered( segment, class, state )
if segment ~= nil and #segment > 0 then
segment[#segment].class = class;
end
return segment;
end
local function remove_last_character_from_memory( segment )
local it = {};
if segment and #segment > 0
and segment[#segment].c.post then -- last segment has a tag. sigh.
it.c = '';
elseif segment and #segment > 0
and #(segment[#segment].c.c) > 0 then
local s1, s2 = mw.ustring.match(segment[#segment].c.c, '^(.*)(.)$');
it.c = s2;
it.post = segment[#segment].c.post;
segment[#segment].c.c = s1;
segment[#segment].c.post = nil;
if #s1 == 0 then
it.pre = segment[#segment].c.pre;
segment[#segment].c.pre = nil;
end
else
it.c = '';
end
return it, segment;
end
local function flush_segment( segment, it, state )
local s = '';
if segment then
for i, v in pairs(segment) do
s = s .. (v.c.pre or '');
if v.class then
s = s .. '<span class=' .. v.class .. '>' .. v.c.c .. '</span>'
else
s = s .. v.c.c;
end
s = s .. (v.c.post or '');
end
if #s > 0 then
s = '<span class=zit3 ' .. '>' .. s .. '</span>';
--s=mw.ustring.gsub(s,'<','<');--DEBUG
it = it .. s;
end
end
segment = {};
return segment, it;
end
it = '';
while #t > 0 do
-- We might encounter a tag. try to not break it
-- XXX Actually, if we see a tag maybe we should stop and give up
local c, t_next = get_token(t);
if state == STATE.INITIAL then
if kernable_left_punctuation_p(c.c) then
state = STATE.OPENING;
if #it == 0 then
segment = remember_kerned(c, 'koen1zo2', segment, state);
else
segment = remember_kerned(c, 'koen1zo2siu2siu2', segment, state);
end
elseif kernable_right_punctuation_p(c.c) then
state = STATE.CLOSING;
local c_last;
c_last, segment = remove_last_character_from_memory(segment);
segment, it = flush_segment(segment, it, state);
segment = remember_unkerned(c_last, segment, state);
segment = remember_kerned(c, 'koen1jau6', segment, state);
else
segment = remember_unkerned(c, segment, state);
end
elseif state == STATE.OPENING then
if kernable_left_punctuation_p(c.c) then
segment = remember_kerned(c, 'koen1zo2', segment, state);
elseif kernable_right_punctuation_p(c.c) then
state = STATE.CLOSING;
segment = remember_kerned(c, 'koen1jau6', segment, state);
else
state = STATE.INITIAL;
segment = remember_unkerned(c, segment, state);
segment, it = flush_segment(segment, it, state);
end
elseif state == STATE.CLOSING then
if kernable_left_punctuation_p(c.c) then
state = STATE.OPENING;
segment, it = flush_segment(segment, it, state);
segment = remember_unkerned(c, segment, state);
elseif kernable_right_punctuation_p(c.c) then
segment = remember_kerned(c, 'koen1jau6', segment, state);
else
state = STATE.INITIAL;
if not space_p(c.c) and not left_kernable_narrow_punctuation_p(c.c) then
segment = change_class_of_last_remembered(segment, 'koen1jau6siu2siu2', state)
end
segment = remember_unkerned(c, segment, state);
segment, it = flush_segment(segment, it, state);
end
else
error('kern() encountered unexpeected state '.. cvs(state));
end
t = t_next;
end
segment, it = flush_segment(segment, it, state);
if ref(s0) == 'parsed-title' then -- try to return a result that's the same type
s.label = it;
it = s;
end
return it;
end
-- Build a type 1 citable with type 2 as a fallback
local function build_type_1_citable( work, part )
local it;
local part1, part2;
local class1, class2;
local prefix, suffix;
local infix = '';
work = parse_title(work);
part = parse_title(part);
if part ~= nil then
prefix = '〈';
suffix = '〉';
if work ~= nil then
infix = '・'; -- U+30FB
part1 = build_type_1_part(work);
part2 = build_type_1_part(part);
class1 = 'syu1ming4';
class2 = 'pin1ming4';
else
part1 = build_type_1_part(part);
class1 = 'pin1ming4';
end
elseif work ~= nil then
part1 = build_type_1_part(work);
class1 = 'syu1ming4';
prefix = '《';
suffix = '》';
end
local alt;
-- build HTML tag with fallback and aria-label
it = '';
if part1 ~= nil then
it = it .. '<span class=' .. class1 .. '>'
.. '<span class=hoi1>' .. prefix .. '</span>'
.. linkify_title(part1)
.. '</span>';
end
if part2 ~= nil then
it = it .. '<span class=' .. class2 .. '>'
.. '<span class=fan1gaak3>' .. infix .. '</span>'
.. linkify_title(part2)
.. '</span>';
if part1 ~= nil then
alt = prefix .. build_aria_label_from(part1
..infix
..part2) .. suffix;
else
alt = prefix .. build_aria_label_from(part2) .. suffix;
end
elseif part1 ~= nil then
alt = prefix .. build_aria_label_from(part1) .. suffix;
end
if it ~= nil then
it = mw.ustring.gsub(it,
'(</span>)$',
'<span class=saan1>' ..suffix .. '</span>%1');
it = '<span aria-label="' .. alt .. '">' .. it .. '</span>';
end
return it;
end
-- Build a type 2 citable using the correct quotation marks
-- and, if needed, inner separator
local function build_type_2_citable( work, part )
local it;
local class;
local prefix;
local suffix;
local root;
local alt;
work = parse_title(work);
part = parse_title(part);
if part ~= nil then
class = 'pin1ming4';
if work ~= nil then
prefix = '《';
suffix = '》';
work.label = prefix .. work.label;
part.label = part.label .. suffix;
root = linkify_title(kern(work))
.. '・' -- dot = U+30FB
.. linkify_title(kern(part));
alt = work.label .. '・' .. part.label;
else
prefix = '〈';
suffix = '〉';
part.label = prefix .. part.label .. suffix;
root = linkify_title(kern(part));
alt = part.label;
end
elseif work ~= nil then
class = 'syu1ming4';
prefix = '《';
suffix = '》';
work.label = prefix .. work.label .. suffix;
root = linkify_title(kern(work));
alt = work.label;
end
if root ~= nil then
alt = build_aria_label_from(alt);
--prefix = '<span class=hoi1-adj>' .. prefix .. '</span>';
--suffix = '<span class=saan1-adj>'.. suffix .. '</span>';
it = '<span class = "' .. class .. '-b" aria-label="' .. alt .. '">'
.. root
.. '</span>';
end
return it;
end
-- Build a non-citable proper noun using CSS underlining or borders
local function build_noncitable_proper_simple( parts, use_dot_p )
local it;
local class = 'zyun1ming4';
local zwsp = ''; -- U+200B
local zwnj = ''; -- U+200C
local infix = '・' -- dot = U+30FB
local root;
local alt;
if not use_dot_p then
infix = zwnj;
end
if parts then
for k, v in pairs(parts) do
local part = parse_title(parts[k]);
if not root then
root = '';
alt = '';
else
root = root .. zwnj;
alt = alt .. infix;
end
local segment = linkify_title(part);
alt = build_aria_label_from(part.label);
segment = '<span class = "' .. class .. '-b" aria-label="' .. alt .. '">'
.. segment
.. '</span>';
root = root .. segment;
end
end
it = '<span class=zyun1ming4-b>'
.. root
.. '</span>';
return it;
end
-- Attempt to build a non-citable proper noun using Unicode
local function build_noncitable_proper_alternate( parts, use_dot_p )
local it;
local class = 'zyun1ming4';
local zwsp = ''; -- U+200B
local zwnj = ''; -- U+200C
local infix = '・' -- dot = U+30FB
local root;
local alt;
if not use_dot_p then
infix = zwnj;
end
if parts then
for k, v in pairs(parts) do
local part = parse_title(parts[k]);
if not root then
root = '';
alt = '';
else
root = root .. zwnj;
alt = alt .. infix;
end
local segment = build_type_1_part(part);
alt = build_aria_label_from(part.label);
segment = '<span class = "' .. class .. '" aria-label="' .. alt .. '">'
.. segment
.. '</span>';
root = root .. segment;
end
end
it = '<span class=zyun1ming4>'
.. root
.. '</span>';
return it;
end
-- Analyze the given title(s) and decide if type 1 is safe to use
local function determine_which_type_to_use( work, part )
local it;
local det;
local t = ref(work);
if t == 'parsed-title' then
work = work.label;
elseif t == 'array' then
work = table.concat(work);
end
if type(part) == 'table' then
part = part.label;
end
if work ~= nil and part ~= nil then
det = work .. part;
elseif work ~= nil then
det = work;
elseif part ~= nil then
det = part;
else
det = '';
end
if cjk_p(det) then
it = TYPE_1;
else
it = TYPE_2;
end
return it;
end
-- Automatically select whether to build either a type 1 or type 2 citable
local function auto_build_citable( work, part )
local type = determine_which_type_to_use(work, part);
local it;
if type == TYPE_1 then
it = build_type_1_citable(work, part);
else
it = build_type_2_citable(work, part);
end
return it;
end
-- Automatically select whether to build noncitable with Unicode or underlining
local function auto_build_noncitable( parts, use_dots_p )
local type = determine_which_type_to_use(parts);
local it;
if type == TYPE_1 then
it = build_noncitable_proper_alternate(parts, use_dots_p);
else
it = build_noncitable_proper_simple(parts, use_dots_p);
end
return it;
end
--- Exported, invocable functions ---------------------------------------------
-- Entry point for Template:書名 and Template:篇名
p.Syu1meng2 = function( frame )
local parent = frame:getParent();
local chapter_mode_p = (frame.args.mode and frame.args.mode == 'chapter') == true;
local it;
local alt = '';
local styles = 'Module:書名/styles.css';
local work, part;
local type;
-- figure out what is actually being marked up as a citable
local parts = {};
local name = parent:getTitle();
if chapter_mode_p then
table.insert(parts, '');
name = '篇名模';
end
for k, v in pairs(parent.args) do
local ps = ' (參數明細：' .. cvs(parent.args) .. ')';
v = sanitize(v);
if ref(k) == 'number' then
assert(v ~= nil);
table.insert(parts, v);
elseif k == 'type' or k == '式' then
if type then
error(name..'遇到 '..k..' 參數，但係已經指咗 ｢'..type..'｣ 式'..ps);
else
type = canonicalize_type_value(v);
end
elseif (not chapter_mode_p and k == 'title')
or (chapter_mode_p and k == 'work') then
if nullify(parts[1]) then
error(name..'遇到 '..k..' 參數，但係已經有 ｢'..parts[1]..'｣'..ps);
else
parts[1] = v;
end
elseif k == 'chapter'
or (chapter_mode_p and k == 'title') then
if nullify(parts[2]) then
error(name..'遇到 '..k..' 參數，但係已經有 ｢'..parts[2]..'｣'..ps);
else
if not parts[1] then
parts[1] = '';
end
parts[2] = v;
end
else
error(name..'遇到不明參數 ｢' .. k .. '｣'..ps);
end
end
if #parts == 0 then
error('冇指定書名或者篇名');
elseif #parts > 2 then
error('指定咗太多章名，暫時處理唔到（parts='..cvs(parts)..'）');
end
work = parse_title(nullify(parts[1]));
part = parse_title(nullify(parts[2]));
-- fixup default type
if type == nil then
type = 'auto';
end
-- build it
if type == 'auto' then
it = auto_build_citable(work, part)
elseif type == TYPE_1 then -- 甲式 (type 1)
it = build_type_1_citable(work, part)
elseif type == TYPE_2 then -- 乙式 (type 2)
it = build_type_2_citable(work, part)
else
error('唔明'..cvs(type)..'式書名號係乜');
end
-- request our style sheet
it = table.concat ({
frame:extensionTag ('templatestyles', '', {src=styles}),
it
});
return it;
end
-- Entry point for Template:專名
p.Zyun1ming4 = function( frame )
local parent = frame:getParent();
local use_dots_p = (frame.args.mode and frame.args.mode == 'dotted') == true;
local it;
local alt = '';
local styles = 'Module:書名/styles.css';
local error;
local parts = {};
for k, v in pairs(parent.args) do
if type(k) == 'number' then
table.insert(parts, v);
elseif not error then
error = '專名模遇到不明參數 ｢' .. k .. '｣';
else
error = error .. '、｢' .. k .. '｣';
end
end
if #parts == 0 then
parts = nil;
end
-- build it
if parts ~= nil then
it = auto_build_noncitable(parts, use_dots_p);
else
if error == nil then
error = '專名模出錯，搵唔到有乜嘢名';
end
end
if it == nil and error ~= nil then
it = '<span class=error>' .. error .. '</span>'
.. '（parts=' .. cvs(parts)
.. '）'
end
-- request our style sheet
it = table.concat ({
frame:extensionTag ('templatestyles', '', {src=styles}),
it
});
return it;
end
-- Entry point for Template:着重
p.Zoek6zung6 = function( frame )
local parent = frame:getParent();
local it;
local alt = '';
local styles = 'Module:書名/styles.css';
local parts = {};
for k, v in pairs(parent.args) do
if type(k) == 'number' then
v = parse_title(v);
v.alt = build_aria_label_from(v.label);
table.insert(parts, v);
else
error('着重模遇到不明參數 ｢' .. k .. '｣');
end
end
if #parts == 0 then
error('着重模出錯，搵唔到着重乜嘢');
end
-- build it
it = '';
for i, v in pairs(parts) do
v = parse_title(v);
local segment = build_type_1_part(v.label);
if it ~= nil then
it = it
.. '<span class=zoek6zung6 aria-label="' .. v.alt .. '">'
.. segment
.. '</span>';
end
end
-- request our style sheet
it = table.concat ({
frame:extensionTag ('templatestyles', '', {src=styles}),
it
});
return it;
end
-- Entry point for Template:Kern
p.Kern = function( frame )
local parent = frame:getParent();
local it;
local alt = '';
local styles = 'Module:書名/styles.css';
local parts = {};
for k, v in pairs(parent.args) do
if type(k) == 'number' then
v = parse_title(v);
v.alt = build_aria_label_from(v.label);
table.insert(parts, v);
else
error('Kern 模遇到不明參數 ｢' .. k .. '｣');
end
end
if #parts == 0 then
error('Kern 模出錯，搵唔到 kern 乜嘢');
end
-- build it
it = '';
for i, v in pairs(parts) do
v = parse_title(v);
local segment = kern(v.label);
if it ~= nil then
it = it
.. '<span class=koen1 aria-label="' .. v.alt .. '">'
.. segment
.. '</span>';
end
end
-- request our style sheet
it = table.concat ({
frame:extensionTag ('templatestyles', '', {src=styles}),
it
});
return it;
end
--- Non-invocable internal functions exported for other modules to use --------
p.cvs = cvs;
p.ref = ref;
p.cjk_p = cjk_p;
p.kernable_left_punctuation_p = kernable_left_punctuation_p;
p.kernable_right_punctuation_p = kernable_right_punctuation_p;
p.array_p = array_p;
p.determine_which_type_to_use = determine_which_type_to_use;
p.build_type_1_citable = build_type_1_citable;
p.build_type_2_citable = build_type_2_citable;
p.auto_build_citable = auto_build_citable;
return p;