| 1 |
--[[ Copyright (c) 2010 Peter "Corsix" Cawley |
| 2 |
|
| 3 |
Permission is hereby granted, free of charge, to any person obtaining a copy of |
| 4 |
this software and associated documentation files (the "Software"), to deal in |
| 5 |
the Software without restriction, including without limitation the rights to |
| 6 |
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies |
| 7 |
of the Software, and to permit persons to whom the Software is furnished to do |
| 8 |
so, subject to the following conditions: |
| 9 |
|
| 10 |
The above copyright notice and this permission notice shall be included in all |
| 11 |
copies or substantial portions of the Software. |
| 12 |
|
| 13 |
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 14 |
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 15 |
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 16 |
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 17 |
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 18 |
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE |
| 19 |
SOFTWARE. --]] |
| 20 |
|
| 21 |
local lfs = require("lfs") |
| 22 |
local TH = require("TH") |
| 23 |
|
| 24 |
--! Layer which handles the loading of localised text. |
| 25 |
class "Strings" |
| 26 |
|
| 27 |
---@type Strings |
| 28 |
local Strings = _G["Strings"] |
| 29 |
|
| 30 |
function Strings:Strings(app) |
| 31 |
self.app = app |
| 32 |
end |
| 33 |
|
| 34 |
local utf8conv |
| 35 |
local cp437conv |
| 36 |
local function id(...) return ... end |
| 37 |
|
| 38 |
function Strings:init() |
| 39 |
-- Load (but do not execute) everything from the language directory |
| 40 |
-- Note that files are loaded with loadfile_envcall |
| 41 |
self.language_chunks = {} |
| 42 |
local pathsep = package.config:sub(1,1) |
| 43 |
local path = self.app:getFullPath({"Lua", "languages"}, true) |
| 44 |
for file in lfs.dir(path) do |
| 45 |
if file:match("%.lua$") then |
| 46 |
local result, err = loadfile_envcall(path .. file) |
| 47 |
if not result then |
| 48 |
print("Error loading languages" .. pathsep .. file .. ":\n" .. tostring(err)) |
| 49 |
else |
| 50 |
self.language_chunks[result] = "languages" .. pathsep .. file |
| 51 |
end |
| 52 |
end |
| 53 |
end |
| 54 |
|
| 55 |
-- Build the language table from Language() calls |
| 56 |
-- Every file in the languages folder should have a call to Language() near |
| 57 |
-- the start of the file which gives the names for the language. These names |
| 58 |
-- are used to link the user's choice of language to a file, and to link the |
| 59 |
-- names given to Inherit() to a file. The first name is used as the official |
| 60 |
-- name for this language, the others may be abbreviations or such. |
| 61 |
self.languages = {} |
| 62 |
self.languages_english = {} |
| 63 |
self.language_to_chunk = {} |
| 64 |
self.chunk_to_font = {} |
| 65 |
self.chunk_to_names = {} |
| 66 |
self.language_to_lang_code = {} |
| 67 |
self.languages_with_arabic_numerals = {} |
| 68 |
for chunk, filename in pairs(self.language_chunks) do |
| 69 |
-- To allow the file to set global variables without causing an error, it |
| 70 |
-- is given an infinite table as an environment. Reading a non-existent |
| 71 |
-- key from an infinite table returns another infinite table, rather than |
| 72 |
-- the default value of nil. |
| 73 |
local infinite_table_mt |
| 74 |
infinite_table_mt = { |
| 75 |
__index = function(_, _) |
| 76 |
return setmetatable({}, infinite_table_mt) |
| 77 |
end |
| 78 |
} |
| 79 |
-- To abort evaluation of the chunk after Language() is called, a unique |
| 80 |
-- marker is used as an error message. The other alternative would be to |
| 81 |
-- do complex co-routine calls. A freshly made empty table is a suitable |
| 82 |
-- unique marker. |
| 83 |
local good_error_marker = {} |
| 84 |
local env = setmetatable({ |
| 85 |
utf8 = id, |
| 86 |
cp437 = cp437conv, |
| 87 |
pairs = pairs, |
| 88 |
Language = function(...) |
| 89 |
local names = {...} |
| 90 |
-- Use the first name for display purposes (case-dependent!). |
| 91 |
if names[1] ~= "original_strings" then |
| 92 |
self.languages[#self.languages + 1] = names[1] |
| 93 |
-- Also save the second name for tooltips and internal purposes. |
| 94 |
assert(names[2], filename .. " does not have an English name.") |
| 95 |
self.languages_english[names[1]] = names[2] |
| 96 |
end |
| 97 |
-- Associate every passed name with this file, case-independently |
| 98 |
for _, name in pairs(names) do |
| 99 |
self.language_to_chunk[name:lower()] = chunk |
| 100 |
self.language_to_lang_code[name:lower()] = names[3] |
| 101 |
end |
| 102 |
self.chunk_to_names[chunk] = names |
| 103 |
error(good_error_marker) |
| 104 |
end, |
| 105 |
Font = function(...) |
| 106 |
self.chunk_to_font[chunk] = ... |
| 107 |
end, |
| 108 |
-- Set Inherit and SetSpeechFile to do nothing |
| 109 |
Inherit = function() end, |
| 110 |
SetSpeechFile = function() end, |
| 111 |
Encoding = function() end, |
| 112 |
IsArabicNumerals = function() end, |
| 113 |
-- Set LoadStrings to return an infinite table |
| 114 |
LoadStrings = infinite_table_mt.__index, |
| 115 |
}, infinite_table_mt) |
| 116 |
-- Actually run the language file |
| 117 |
local status, err = pcall(chunk, env) |
| 118 |
if not status and err ~= good_error_marker and TheApp.good_install_folder then |
| 119 |
print("Error evaluating " .. filename .. ":\n" .. tostring(err)) |
| 120 |
end |
| 121 |
end |
| 122 |
table.sort(self.languages) |
| 123 |
end |
| 124 |
|
| 125 |
-- String tables are shadowed so that all access to a string table goes |
| 126 |
-- through a metamethod (which is desirable to catch invalid reads, provide |
| 127 |
-- "__random", and prevent writes). This means that every string table is in |
| 128 |
-- fact empty, and has it's keys and values stored in a separate table, called |
| 129 |
-- the shadow table. The shadows table is used to associate a shadow table |
| 130 |
-- with a string table. |
| 131 |
local shadows = setmetatable({}, {__mode = "k"}) |
| 132 |
|
| 133 |
-- Metatable which is used for all tables returned by Strings:load() |
| 134 |
-- The end effect is to raise errors on accesses to non-existent strings |
| 135 |
-- (unless no_restriction is set to true), to add a special string called |
| 136 |
-- "__random" to each table (which always resolves to a random string from |
| 137 |
-- the table), and to prevent editing or adding to a string table. |
| 138 |
local strings_metatable = function(no_restriction) return { |
| 139 |
__index = function(t, key) |
| 140 |
t = shadows[t] |
| 141 |
local val = t[key] |
| 142 |
if val ~= nil then |
| 143 |
return val |
| 144 |
end |
| 145 |
if key ~= "__random" then |
| 146 |
if no_restriction then return nil end |
| 147 |
error("Non-existent string: " .. tostring(key), 2) |
| 148 |
end |
| 149 |
local candidates = {} |
| 150 |
for _, v in pairs(t) do |
| 151 |
candidates[#candidates + 1] = v |
| 152 |
end |
| 153 |
return candidates[math.random(1, #candidates)] |
| 154 |
end, |
| 155 |
__newindex = function(_, _, _) |
| 156 |
error("String tables are read-only", 2) |
| 157 |
end, |
| 158 |
__pairs = function(t) |
| 159 |
return pairs(shadows[t]) |
| 160 |
end, |
| 161 |
__ipairs = function(t) |
| 162 |
return ipairs(shadows[t]) |
| 163 |
end, |
| 164 |
} end |
| 165 |
|
| 166 |
-- no_restriction disables errors on access to non-existent strings (for debug purposes) |
| 167 |
-- no_inheritance disables inheritance except original_strings (for debug purposes) |
| 168 |
function Strings:load(language, no_restriction, no_inheritance) |
| 169 |
assert(language ~= "original_strings", "Original strings can not be loaded directly. Please select a proper language.") |
| 170 |
-- env is the table of globals to execute to the language file in, and hence |
| 171 |
-- it also stores the resulting strings. |
| 172 |
local env = {} |
| 173 |
shadows[env] = {} |
| 174 |
-- speech_file holds the result of any call to SetSpeechFile() |
| 175 |
local speech_file |
| 176 |
local default_encoding = id |
| 177 |
local encoding = default_encoding |
| 178 |
local language_called = false |
| 179 |
local functions; functions = { |
| 180 |
-- Convert UTF-8 to the file's default encoding |
| 181 |
utf8 = function(s) |
| 182 |
if encoding == cp437conv then |
| 183 |
return utf8conv(s) |
| 184 |
else |
| 185 |
return s |
| 186 |
end |
| 187 |
end, |
| 188 |
-- Convert CP437 to the file's default encoding |
| 189 |
cp437 = function(s) |
| 190 |
if encoding == cp437conv then |
| 191 |
return s |
| 192 |
else |
| 193 |
return cp437conv(s) |
| 194 |
end |
| 195 |
end, |
| 196 |
ipairs = ipairs, |
| 197 |
pairs = pairs, |
| 198 |
-- Calling the Language() function should have no effect any more |
| 199 |
Language = function() |
| 200 |
language_called = true |
| 201 |
end, |
| 202 |
Font = function() |
| 203 |
if language_called then |
| 204 |
error("Font declaration must occur before Language declaration") |
| 205 |
end |
| 206 |
end, |
| 207 |
-- Inherit() should evaluate the named language in the current environment |
| 208 |
-- NB: Inheritance of any but original_strings disabled when no_inheritance set |
| 209 |
Inherit = function(lang, ...) |
| 210 |
if no_inheritance and lang ~= "original_strings" then return end |
| 211 |
local old_encoding = encoding |
| 212 |
encoding = default_encoding |
| 213 |
local old_language_called = language_called |
| 214 |
language_called = false |
| 215 |
self:_loadPrivate(lang, env, ...) |
| 216 |
encoding = old_encoding |
| 217 |
language_called = old_language_called |
| 218 |
end, |
| 219 |
-- Encoding() should set the default encoding for the remainder of the file |
| 220 |
Encoding = function(new_encoding) |
| 221 |
if new_encoding == functions.utf8 then |
| 222 |
encoding = id |
| 223 |
elseif new_encoding == functions.cp437 then |
| 224 |
encoding = cp437conv |
| 225 |
else |
| 226 |
error("Invalid encoding; expected utf8 or cp437") |
| 227 |
end |
| 228 |
end, |
| 229 |
-- LoadStrings() should return the original game string table |
| 230 |
LoadStrings = function(filename) |
| 231 |
return assert(TH.LoadStrings(self.app:readDataFile(filename)), |
| 232 |
"Cannot load original string file '" .. filename .. "'") |
| 233 |
end, |
| 234 |
-- SetSpeechFile() should remember the named file to return to our caller |
| 235 |
SetSpeechFile = function(...) |
| 236 |
speech_file = ... |
| 237 |
end, |
| 238 |
IsArabicNumerals = function(is_enabled) |
| 239 |
self.languages_with_arabic_numerals[language] = is_enabled |
| 240 |
end, |
| 241 |
_G = env, |
| 242 |
type = type, |
| 243 |
} |
| 244 |
-- The metatable on the environment is set so that the above functions look |
| 245 |
-- like top-level level globals, so that the environment behaves like an |
| 246 |
-- infinite table, and so that assignments are merged into the string table. |
| 247 |
local metatable |
| 248 |
metatable = { |
| 249 |
__index = function(t, k) |
| 250 |
-- Make the functions look like top-level globals |
| 251 |
local shadow = shadows[t] |
| 252 |
local value = ((t == env) and functions[k]) or shadow[k] |
| 253 |
if value == nil then |
| 254 |
value = setmetatable({}, metatable) |
| 255 |
shadow[k] = value |
| 256 |
shadows[value] = {} |
| 257 |
end |
| 258 |
return value |
| 259 |
end, |
| 260 |
__newindex = function(t, k, v) |
| 261 |
local ty = type(v) |
| 262 |
if ty ~= "table" then |
| 263 |
-- non-table values cannot be merged |
| 264 |
if ty == "string" then |
| 265 |
-- convert from file's default encoding to UTF-8 |
| 266 |
v = encoding(v) |
| 267 |
end |
| 268 |
shadows[t][k] = v |
| 269 |
else |
| 270 |
-- v should be merged into t[k] |
| 271 |
-- Perform t[k][k2] = v2 for each (k2, v2) in v to recursively merge |
| 272 |
t = t[k] |
| 273 |
for k2, v2 in pairs(v) do |
| 274 |
t[k2] = v2 |
| 275 |
end |
| 276 |
end |
| 277 |
end, |
| 278 |
} |
| 279 |
metatable.__pairs = strings_metatable(no_restriction).__pairs |
| 280 |
-- Evaluate the language file |
| 281 |
setmetatable(env, metatable) |
| 282 |
self:_loadPrivate(language, env) |
| 283 |
-- Change the metamethods on every string table to match strings_metatable |
| 284 |
for k, v in pairs(strings_metatable(no_restriction)) do |
| 285 |
metatable[k] = v |
| 286 |
end |
| 287 |
return env, speech_file |
| 288 |
end |
| 289 |
|
| 290 |
--! Get the Font() declaration of a language, if there was one. |
| 291 |
function Strings:getFont(language) |
| 292 |
local chunk = self.language_to_chunk[language:lower()] |
| 293 |
return chunk and self.chunk_to_font[chunk] |
| 294 |
end |
| 295 |
|
| 296 |
function Strings:getLanguageNames(language) |
| 297 |
local chunk = self.language_to_chunk[language:lower()] |
| 298 |
return chunk and self.chunk_to_names[chunk] |
| 299 |
end |
| 300 |
|
| 301 |
function Strings:getLangCode(language) |
| 302 |
local lang = language or self.app.config.language |
| 303 |
return self.language_to_lang_code[lang:lower()] |
| 304 |
end |
| 305 |
|
| 306 |
function Strings:isArabicNumerals(language) |
| 307 |
return self.languages_with_arabic_numerals[language:lower()] |
| 308 |
end |
| 309 |
|
| 310 |
--! Use local language text where possible. |
| 311 |
--!param string (string) The default, likely English, text |
| 312 |
--!param table (table) A table of translated text in language code fields |
| 313 |
--!return (string) The text in the current language if available, or in English, or the default string. |
| 314 |
function Strings:getLocalisedText(string, table) |
| 315 |
if string and not table then return string |
| 316 |
elseif table[self:getLangCode()] then return table[self:getLangCode()] |
| 317 |
elseif table.en then return table.en |
| 318 |
else return string |
| 319 |
end |
| 320 |
end |
| 321 |
|
| 322 |
--! Checks the language we specify is known to the application. |
| 323 |
--!param language (string) Language to check |
| 324 |
--!return true if known |
| 325 |
function Strings:checkLanguageExists(language) |
| 326 |
return not not self.language_to_chunk[language:lower()] |
| 327 |
end |
| 328 |
|
| 329 |
function Strings:_loadPrivate(language, env, ...) |
| 330 |
local chunk = self.language_to_chunk[language:lower()] |
| 331 |
if not chunk then -- If selected language could not be found, try to revert to English |
| 332 |
print(serialize(self.language_to_chunk, {detect_cycles=true, pretty=true})) |
| 333 |
print("Language '" .. language .. "' could not be found. Reverting to English.") |
| 334 |
chunk = self.language_to_chunk["english"] |
| 335 |
if not chunk then -- If english could not be found, raise an error |
| 336 |
error("Language 'English' could not be found. Please verify your installation.") |
| 337 |
end |
| 338 |
end |
| 339 |
local status, err = pcall(chunk, env, ...) |
| 340 |
if not status and self.app.good_install_folder then |
| 341 |
print("Error evaluating " .. self.language_chunks[chunk] .. ":\n" .. tostring(err)) |
| 342 |
end |
| 343 |
end |
| 344 |
|
| 345 |
function Strings:setupAdviserMessage(messages) |
| 346 |
local prioTable |
| 347 |
prioTable = { |
| 348 |
_priority = 5, -- Default priority |
| 349 |
tutorial = { _priority = 11}, |
| 350 |
epidemic = { _priority = 6 }, |
| 351 |
--staff_advice |
| 352 |
--earthquake |
| 353 |
--multiplayer |
| 354 |
--surgery_requirements |
| 355 |
--vomit_wave |
| 356 |
--level_progress |
| 357 |
--staff_place_advice |
| 358 |
--room_forbidden_non_reachable_parts |
| 359 |
--research |
| 360 |
--boiler_issue |
| 361 |
--room_requirements |
| 362 |
--goals |
| 363 |
warnings = { _priority = 10 }, |
| 364 |
--placement_info |
| 365 |
praise = { _priority = 1 }, |
| 366 |
--information |
| 367 |
--build_advice |
| 368 |
--cheats |
| 369 |
} |
| 370 |
-- Passes adviser message through string.format process |
| 371 |
local formatFunc |
| 372 |
formatFunc = function(format_self, ...) |
| 373 |
-- After 'format', it is not useful to have indexing magic anymore. |
| 374 |
return { text = format_self.text:format(...), priority = format_self.priority } |
| 375 |
end |
| 376 |
local indexFunc |
| 377 |
indexFunc = function(index_self, field) |
| 378 |
-- Since prioTable is a partial table, index_self.table may disappear, prevent infinite recursion. |
| 379 |
if field == "table" then |
| 380 |
return nil |
| 381 |
end |
| 382 |
local val = {} |
| 383 |
val.text = index_self.text[field] |
| 384 |
val.format = formatFunc |
| 385 |
val.table = index_self.table |
| 386 |
val.priority = index_self.priority |
| 387 |
if val.table ~= nil then |
| 388 |
val.table = val.table[field] |
| 389 |
if val.table ~= nil and val.table._priority ~= nil then |
| 390 |
val.priority = val.table._priority |
| 391 |
end |
| 392 |
end |
| 393 |
setmetatable(val, {__index = indexFunc}) |
| 394 |
return val |
| 395 |
end |
| 396 |
-- Initial value |
| 397 |
local adviserMessage = { |
| 398 |
text = messages, |
| 399 |
table = prioTable, |
| 400 |
priority = prioTable._priority, |
| 401 |
format = formatFunc |
| 402 |
} |
| 403 |
setmetatable(adviserMessage, {__index = indexFunc}) |
| 404 |
return adviserMessage |
| 405 |
end |
| 406 |
|
| 407 |
|
| 408 |
-- Primitive system to map UTF-8 characters onto Code Page 437. |
| 409 |
-- Provided so that language scripts can encode text in a modern and well |
| 410 |
-- supported manner, and have the text automatically transcoded into the |
| 411 |
-- encoding which the Theme Hospital fonts use. |
| 412 |
|
| 413 |
local codepoints_to_cp437 = { |
| 414 |
-- Below 0x80 need no translation |
| 415 |
[0xC7] = 0x80, -- majuscule c-cedilla |
| 416 |
[0xFC] = 0x81, -- minuscule u-umlaut |
| 417 |
[0xE9] = 0x82, -- minuscule e-acute |
| 418 |
[0xE2] = 0x83, -- minuscule a-circumflex |
| 419 |
[0xE4] = 0x84, -- minuscule a-umlaut |
| 420 |
[0xE0] = 0x85, -- minuscule a-grave |
| 421 |
[0xE5] = 0x86, -- minuscule a-ring |
| 422 |
[0xE7] = 0x87, -- minuscule c-cedilla |
| 423 |
[0xEA] = 0x88, -- minuscule e-circumflex |
| 424 |
[0xEB] = 0x89, -- minuscule e-umlaut |
| 425 |
[0xE8] = 0x8A, -- minuscule e-grave |
| 426 |
[0xEF] = 0x8B, -- minuscule i-umlaut |
| 427 |
[0xEE] = 0x8C, -- minuscule i-circumflex |
| 428 |
[0xEC] = 0x8D, -- minuscule i-grave |
| 429 |
[0xC4] = 0x8E, -- majuscule a-umlaut |
| 430 |
[0xC5] = 0x8F, -- majuscule a-ring |
| 431 |
[0xC9] = 0x90, -- majuscule e-acute |
| 432 |
[0xE6] = 0x91, -- minuscule ae |
| 433 |
[0xC6] = 0x91, -- majuscule ae (not in TH fonts - mapped to minuscule ae) |
| 434 |
[0xF4] = 0x93, -- minuscule o-circumflex |
| 435 |
[0xF6] = 0x94, -- minuscule o-umlaut |
| 436 |
[0xF2] = 0x95, -- minuscule o-grave |
| 437 |
[0xFB] = 0x96, -- minuscule u-circumflex |
| 438 |
[0xF9] = 0x97, -- minuscule u-grave |
| 439 |
[0xFF] = 0x98, -- minuscule y-umlaut |
| 440 |
[0xD6] = 0x99, -- majuscule o-umlaut |
| 441 |
[0xDC] = 0x9A, -- majuscule u-umlaut |
| 442 |
-- 0x9B through 0x9F are currency symbols and not present in TH fonts |
| 443 |
[0xE1] = 0xA0, -- minuscule a-acute |
| 444 |
[0xED] = 0xA1, -- minuscule i-acute |
| 445 |
[0xF3] = 0xA2, -- minuscule o-acute |
| 446 |
[0xFA] = 0xA3, -- minuscule u-acute |
| 447 |
[0xF1] = 0xA4, -- minuscule n-tilde |
| 448 |
[0xD1] = 0xA5, -- majuscule n-tilde |
| 449 |
-- 0xA6 and 0xA7 are ordinal indicators and not present in TH fonts |
| 450 |
[0xBF] = 0xA8, -- inverted question mark |
| 451 |
-- 0xA9 through 0xAC are not present in TH fonts |
| 452 |
[0xA1] = 0xAD, -- inverted exclaimation mark |
| 453 |
-- 0xAE through 0xE0 are not present in TH fonts |
| 454 |
[0xDF] = 0xE1, -- eszett / sharp-S / lowercase-beta |
| 455 |
-- 0xE2 through 0xFF are not present in TH fonts |
| 456 |
} |
| 457 |
|
| 458 |
local cp437_to_codepoints = { |
| 459 |
[0x91] = 0xE6, -- minuscule ae |
| 460 |
[0x9B] = 0xA2, -- cent sign |
| 461 |
[0x9C] = 0xA3, -- pound sign |
| 462 |
[0x9D] = 0xA5, -- yen sign |
| 463 |
[0x9E] = 0x20A7, -- peseta sign |
| 464 |
[0x9F] = 0x192, -- florin sign |
| 465 |
[0xA6] = 0xAA, -- a ordinal indicator |
| 466 |
[0xA7] = 0xBA, -- o ordinal indicator |
| 467 |
[0xA9] = 0x2310, -- negation |
| 468 |
[0xAA] = 0xAC, -- negation |
| 469 |
[0xAB] = 0xBD, -- 1/2 |
| 470 |
[0xAC] = 0xBC, -- 1/4 |
| 471 |
[0xAE] = 0xAB, -- << guillemets |
| 472 |
[0xAF] = 0xBB, -- >> guillemets |
| 473 |
-- 0xB0 through 0xDF omitted |
| 474 |
[0xE0] = 0x3B1, -- alpha |
| 475 |
[0xE2] = 0x393, -- gamma |
| 476 |
[0xE3] = 0x3C0, -- pi |
| 477 |
[0xE4] = 0x3A3, -- majuscule sigma |
| 478 |
[0xE5] = 0x3C3, -- minuscule sigma |
| 479 |
[0xE6] = 0x3BC, -- minuscule mu |
| 480 |
[0xE7] = 0x3C4, -- tau |
| 481 |
[0xE8] = 0x3A6, -- majuscule phi |
| 482 |
[0xE9] = 0x398, -- theta |
| 483 |
[0xEA] = 0x3A9, -- omega |
| 484 |
[0xEB] = 0x3B4, -- delta |
| 485 |
[0xEC] = 0x221E, -- infinity |
| 486 |
[0xED] = 0x3C6, -- minuscule phi |
| 487 |
[0xEE] = 0x3B5, -- epsilon |
| 488 |
[0xEF] = 0x2229, -- set intersection |
| 489 |
[0xF0] = 0x2261, -- modular congruence / triple equals |
| 490 |
[0xF1] = 0xB1, -- plus / minus |
| 491 |
[0xF2] = 0x2265, -- >= |
| 492 |
[0xF3] = 0x2264, -- <= |
| 493 |
[0xF4] = 0x2320, -- upper integral sign |
| 494 |
[0xF5] = 0x2321, -- lower integral sign |
| 495 |
[0xF6] = 0xF7, -- obelus |
| 496 |
[0xF7] = 0x2248, -- approximate equality |
| 497 |
[0xF8] = 0xB0, -- degree symbol |
| 498 |
[0xF9] = 0x2219, -- bullet |
| 499 |
[0xFA] = 0xB7, -- interpunct |
| 500 |
[0xFB] = 0x221A, -- square root |
| 501 |
[0xFC] = 0x207F, -- exponentiation |
| 502 |
[0xFD] = 0xB2, -- superscript 2 |
| 503 |
[0xFE] = 0x25A0, -- filled square |
| 504 |
[0xFF] = 0xA0, -- non-breaking space |
| 505 |
} |
| 506 |
|
| 507 |
local function utf8encode(codepoint) |
| 508 |
if codepoint <= 0x7F then |
| 509 |
return string.char(codepoint) |
| 510 |
elseif codepoint <= 0x7FF then |
| 511 |
local sextet = codepoint % 64 |
| 512 |
codepoint = math.floor((codepoint - sextet) / 64) |
| 513 |
return string.char(0xC0 + codepoint, 0x80 + sextet) |
| 514 |
elseif codepoint <= 0xFFFF then |
| 515 |
local sextet2 = codepoint % 64 |
| 516 |
codepoint = math.floor((codepoint - sextet2) / 64) |
| 517 |
local sextet1 = codepoint % 64 |
| 518 |
codepoint = math.floor((codepoint - sextet2) / 64) |
| 519 |
return string.char(0xE0 + codepoint, 0x80 + sextet1, 0x80 + sextet2) |
| 520 |
else |
| 521 |
local sextet3 = codepoint % 64 |
| 522 |
codepoint = math.floor((codepoint - sextet3) / 64) |
| 523 |
local sextet2 = codepoint % 64 |
| 524 |
codepoint = math.floor((codepoint - sextet2) / 64) |
| 525 |
local sextet1 = codepoint % 64 |
| 526 |
codepoint = math.floor((codepoint - sextet2) / 64) |
| 527 |
return string.char(0xF0 + codepoint, 0x80 + sextet1, 0x80 + sextet2, |
| 528 |
0x80 + sextet3) |
| 529 |
end |
| 530 |
end |
| 531 |
|
| 532 |
local cp437_to_utf8_pattern = {"["} |
| 533 |
local cp437_to_utf8_replacement = {} |
| 534 |
for codepoint, cp437 in pairs(codepoints_to_cp437) do |
| 535 |
if not cp437_to_codepoints[cp437] then |
| 536 |
cp437_to_utf8_pattern[#cp437_to_utf8_pattern + 1] = string.char(cp437) |
| 537 |
cp437_to_utf8_replacement[string.char(cp437)] = utf8encode(codepoint) |
| 538 |
end |
| 539 |
end |
| 540 |
for cp437, codepoint in pairs(cp437_to_codepoints) do |
| 541 |
cp437_to_utf8_pattern[#cp437_to_utf8_pattern + 1] = string.char(cp437) |
| 542 |
cp437_to_utf8_replacement[string.char(cp437)] = utf8encode(codepoint) |
| 543 |
end |
| 544 |
cp437_to_utf8_pattern[#cp437_to_utf8_pattern + 1] = "]" |
| 545 |
cp437_to_utf8_pattern = table.concat(cp437_to_utf8_pattern) |
| 546 |
cp437conv = function(s) |
| 547 |
return (s:gsub(cp437_to_utf8_pattern, cp437_to_utf8_replacement)) |
| 548 |
end |
| 549 |
|
| 550 |
-- Table which maps a single character and a single unicode combining |
| 551 |
-- diacritical mark to a single unicode codepoint |
| 552 |
local circumflex = 0x302 |
| 553 |
local cedilla = 0x327 |
| 554 |
local umlaut = 0x308 |
| 555 |
local acute = 0x301 |
| 556 |
local grave = 0x300 |
| 557 |
local ring = 0x30A |
| 558 |
local tilde = 0x303 |
| 559 |
local combine_diacritical_marks = { |
| 560 |
a = { |
| 561 |
[grave] = 0xE0, |
| 562 |
[acute] = 0xE1, |
| 563 |
[circumflex] = 0xE2, |
| 564 |
[umlaut] = 0xE4, |
| 565 |
[ring] = 0xE5, |
| 566 |
}, |
| 567 |
e = { |
| 568 |
[grave] = 0xE8, |
| 569 |
[acute] = 0xE9, |
| 570 |
[circumflex] = 0xEA, |
| 571 |
[umlaut] = 0xEB, |
| 572 |
}, |
| 573 |
i = { |
| 574 |
[grave] = 0xEC, |
| 575 |
[acute] = 0xED, |
| 576 |
[circumflex] = 0xEE, |
| 577 |
[umlaut] = 0xEF, |
| 578 |
}, |
| 579 |
o = { |
| 580 |
[grave] = 0xF2, |
| 581 |
[acute] = 0xF3, |
| 582 |
}, |
| 583 |
u = { |
| 584 |
[grave] = 0xF9, |
| 585 |
[acute] = 0xFA, |
| 586 |
[circumflex] = 0xFB, |
| 587 |
[umlaut] = 0xFC, |
| 588 |
}, |
| 589 |
c = { |
| 590 |
[cedilla] = 0xE7, |
| 591 |
}, |
| 592 |
n = { |
| 593 |
[tilde] = 0xF1, |
| 594 |
}, |
| 595 |
y = { |
| 596 |
[umlaut] = 0xFF, |
| 597 |
}, |
| 598 |
A = { |
| 599 |
[umlaut] = 0xC4, |
| 600 |
[ring] = 0xC5, |
| 601 |
}, |
| 602 |
E = { |
| 603 |
[acute] = 0xC9, |
| 604 |
}, |
| 605 |
O = { |
| 606 |
[umlaut] = 0xD6, |
| 607 |
}, |
| 608 |
U = { |
| 609 |
[umlaut] = 0xDC, |
| 610 |
}, |
| 611 |
C = { |
| 612 |
[cedilla] = 0xC7, |
| 613 |
}, |
| 614 |
N = { |
| 615 |
[tilde] = 0xD1, |
| 616 |
}, |
| 617 |
} |
| 618 |
|
| 619 |
local function utf8char(c) |
| 620 |
-- Test for presence of a normal character prior to the utf-8 character |
| 621 |
local prechar |
| 622 |
if c:byte() < 128 then |
| 623 |
prechar = c:sub(1, 1) |
| 624 |
c = c:sub(2, -1) |
| 625 |
end |
| 626 |
-- Extract the codepoint of the utf-8 character |
| 627 |
local codepoint = 0 |
| 628 |
local multiplier = 1 |
| 629 |
for i = #c, 2, -1 do |
| 630 |
codepoint = codepoint + (c:byte(i) - 128) * multiplier |
| 631 |
multiplier = multiplier * 2^6 |
| 632 |
end |
| 633 |
codepoint = codepoint + (c:byte(1) % 2^(7 - #c)) * multiplier |
| 634 |
-- If the utf-8 character is a combining diacritical mark, merge it with the |
| 635 |
-- preceding normal character |
| 636 |
if prechar and (0x300 <= codepoint and codepoint < 0x370) then |
| 637 |
if combine_diacritical_marks[prechar] then |
| 638 |
if combine_diacritical_marks[prechar][codepoint] then |
| 639 |
codepoint = combine_diacritical_marks[prechar][codepoint] |
| 640 |
prechar = nil |
| 641 |
else |
| 642 |
return prechar |
| 643 |
end |
| 644 |
else |
| 645 |
return prechar |
| 646 |
end |
| 647 |
end |
| 648 |
-- Convert to Code Page 437 |
| 649 |
return (prechar or "") .. string.char(codepoints_to_cp437[codepoint] or 0x3F) |
| 650 |
end |
| 651 |
|
| 652 |
utf8conv = function(s) |
| 653 |
-- Pull out each individual utf-8 character and pass it through utf8char |
| 654 |
-- [\1-\127] picks up a preceding ASCII character to combine diacritics |
| 655 |
-- [\192-\253] picks up the first byte of a utf-8 character (technically |
| 656 |
-- only 194 through 244 should be used) |
| 657 |
-- [\128-\191] picks up the remaining bytes of a utf-8 character |
| 658 |
return (s:gsub("[\1-\127]?[\192-\253][\128-\191]*", utf8char)) |
| 659 |
end |
| 660 |
|
| 661 |
-- Fix string.upper and string.lower to work with CP437 rather than current locale |
| 662 |
local lower_to_upper, upper_to_lower = {}, {} |
| 663 |
local function case(lower, upper) |
| 664 |
lower = cp437conv(string.char(lower)) |
| 665 |
upper = cp437conv(string.char(upper)) |
| 666 |
lower_to_upper[lower] = upper |
| 667 |
lower_to_upper[upper] = upper |
| 668 |
upper_to_lower[lower] = lower |
| 669 |
upper_to_lower[upper] = lower |
| 670 |
end |
| 671 |
case(0x87, 0x80) -- c-cedilla |
| 672 |
case(0x81, 0x9A) -- u-umlaut |
| 673 |
case(0x82, 0x90) -- e-acute |
| 674 |
case(0x84, 0x8E) -- a-umlaut |
| 675 |
case(0x86, 0x8F) -- a-ring |
| 676 |
case(0x94, 0x99) -- o-umlaut |
| 677 |
case(0xA4, 0xA5) -- n-tilde |
| 678 |
local case_pattern = "\195[\128-\191]" -- Unicode range [0xC0, 0xFF] as UTF-8 |
| 679 |
|
| 680 |
local orig_upper = string.upper |
| 681 |
function string.upper(s) -- luacheck: ignore 122 |
| 682 |
return orig_upper(s:gsub(case_pattern, lower_to_upper)) |
| 683 |
end |
| 684 |
|
| 685 |
local orig_lower = string.lower |
| 686 |
function string.lower(s) -- luacheck: ignore 122 |
| 687 |
return orig_lower(s:gsub(case_pattern, upper_to_lower)) |
| 688 |
end |