| 1 | /**************************************************************************/ |
| 2 | /* translation_loader_po.cpp */ |
| 3 | /**************************************************************************/ |
| 4 | /* This file is part of: */ |
| 5 | /* GODOT ENGINE */ |
| 6 | /* https://godotengine.org */ |
| 7 | /**************************************************************************/ |
| 8 | /* Copyright (c) 2014-present Godot Engine contributors (see AUTHORS.md). */ |
| 9 | /* Copyright (c) 2007-2014 Juan Linietsky, Ariel Manzur. */ |
| 10 | /* */ |
| 11 | /* Permission is hereby granted, free of charge, to any person obtaining */ |
| 12 | /* a copy of this software and associated documentation files (the */ |
| 13 | /* "Software"), to deal in the Software without restriction, including */ |
| 14 | /* without limitation the rights to use, copy, modify, merge, publish, */ |
| 15 | /* distribute, sublicense, and/or sell copies of the Software, and to */ |
| 16 | /* permit persons to whom the Software is furnished to do so, subject to */ |
| 17 | /* the following conditions: */ |
| 18 | /* */ |
| 19 | /* The above copyright notice and this permission notice shall be */ |
| 20 | /* included in all copies or substantial portions of the Software. */ |
| 21 | /* */ |
| 22 | /* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, */ |
| 23 | /* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF */ |
| 24 | /* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. */ |
| 25 | /* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY */ |
| 26 | /* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, */ |
| 27 | /* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE */ |
| 28 | /* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */ |
| 29 | /**************************************************************************/ |
| 30 | |
| 31 | #include "translation_loader_po.h" |
| 32 | |
| 33 | #include "core/io/file_access.h" |
| 34 | #include "core/string/translation.h" |
| 35 | #include "core/string/translation_po.h" |
| 36 | |
| 37 | Ref<Resource> TranslationLoaderPO::load_translation(Ref<FileAccess> f, Error *r_error) { |
| 38 | if (r_error) { |
| 39 | *r_error = ERR_FILE_CORRUPT; |
| 40 | } |
| 41 | |
| 42 | const String path = f->get_path(); |
| 43 | Ref<TranslationPO> translation = Ref<TranslationPO>(memnew(TranslationPO)); |
| 44 | String config; |
| 45 | |
| 46 | uint32_t magic = f->get_32(); |
| 47 | if (magic == 0x950412de) { |
| 48 | // Load binary MO file. |
| 49 | |
| 50 | uint16_t version_maj = f->get_16(); |
| 51 | uint16_t version_min = f->get_16(); |
| 52 | ERR_FAIL_COND_V_MSG(version_maj > 1, Ref<Resource>(), vformat("Unsupported MO file %s, version %d.%d." , path, version_maj, version_min)); |
| 53 | |
| 54 | uint32_t num_strings = f->get_32(); |
| 55 | uint32_t id_table_offset = f->get_32(); |
| 56 | uint32_t trans_table_offset = f->get_32(); |
| 57 | |
| 58 | // Read string tables. |
| 59 | for (uint32_t i = 0; i < num_strings; i++) { |
| 60 | String msg_id; |
| 61 | String msg_id_plural; |
| 62 | String msg_context; |
| 63 | |
| 64 | // Read id strings and context. |
| 65 | { |
| 66 | Vector<uint8_t> data; |
| 67 | f->seek(id_table_offset + i * 8); |
| 68 | uint32_t str_start = 0; |
| 69 | uint32_t str_len = f->get_32(); |
| 70 | uint32_t str_offset = f->get_32(); |
| 71 | |
| 72 | data.resize(str_len + 1); |
| 73 | f->seek(str_offset); |
| 74 | f->get_buffer(data.ptrw(), str_len); |
| 75 | data.write[str_len] = 0; |
| 76 | |
| 77 | bool is_plural = false; |
| 78 | for (uint32_t j = 0; j < str_len + 1; j++) { |
| 79 | if (data[j] == 0x04) { |
| 80 | msg_context.parse_utf8((const char *)data.ptr(), j); |
| 81 | str_start = j + 1; |
| 82 | } |
| 83 | if (data[j] == 0x00) { |
| 84 | if (is_plural) { |
| 85 | msg_id_plural.parse_utf8((const char *)(data.ptr() + str_start), j - str_start); |
| 86 | } else { |
| 87 | msg_id.parse_utf8((const char *)(data.ptr() + str_start), j - str_start); |
| 88 | is_plural = true; |
| 89 | } |
| 90 | str_start = j + 1; |
| 91 | } |
| 92 | } |
| 93 | } |
| 94 | |
| 95 | // Read translated strings. |
| 96 | { |
| 97 | Vector<uint8_t> data; |
| 98 | f->seek(trans_table_offset + i * 8); |
| 99 | uint32_t str_len = f->get_32(); |
| 100 | uint32_t str_offset = f->get_32(); |
| 101 | |
| 102 | data.resize(str_len + 1); |
| 103 | f->seek(str_offset); |
| 104 | f->get_buffer(data.ptrw(), str_len); |
| 105 | data.write[str_len] = 0; |
| 106 | |
| 107 | if (msg_id.is_empty()) { |
| 108 | config = String::utf8((const char *)data.ptr(), str_len); |
| 109 | // Record plural rule. |
| 110 | int p_start = config.find("Plural-Forms" ); |
| 111 | if (p_start != -1) { |
| 112 | int p_end = config.find("\n" , p_start); |
| 113 | translation->set_plural_rule(config.substr(p_start, p_end - p_start)); |
| 114 | } |
| 115 | } else { |
| 116 | uint32_t str_start = 0; |
| 117 | Vector<String> plural_msg; |
| 118 | for (uint32_t j = 0; j < str_len + 1; j++) { |
| 119 | if (data[j] == 0x00) { |
| 120 | if (msg_id_plural.is_empty()) { |
| 121 | translation->add_message(msg_id, String::utf8((const char *)(data.ptr() + str_start), j - str_start), msg_context); |
| 122 | } else { |
| 123 | plural_msg.push_back(String::utf8((const char *)(data.ptr() + str_start), j - str_start)); |
| 124 | } |
| 125 | str_start = j + 1; |
| 126 | } |
| 127 | } |
| 128 | if (!plural_msg.is_empty()) { |
| 129 | translation->add_plural_message(msg_id, plural_msg, msg_context); |
| 130 | } |
| 131 | } |
| 132 | } |
| 133 | } |
| 134 | |
| 135 | } else { |
| 136 | // Try to load as text PO file. |
| 137 | f->seek(0); |
| 138 | |
| 139 | enum Status { |
| 140 | STATUS_NONE, |
| 141 | STATUS_READING_ID, |
| 142 | STATUS_READING_STRING, |
| 143 | STATUS_READING_CONTEXT, |
| 144 | STATUS_READING_PLURAL, |
| 145 | }; |
| 146 | |
| 147 | Status status = STATUS_NONE; |
| 148 | |
| 149 | String msg_id; |
| 150 | String msg_str; |
| 151 | String msg_context; |
| 152 | Vector<String> msgs_plural; |
| 153 | |
| 154 | if (r_error) { |
| 155 | *r_error = ERR_FILE_CORRUPT; |
| 156 | } |
| 157 | |
| 158 | int line = 1; |
| 159 | int plural_forms = 0; |
| 160 | int plural_index = -1; |
| 161 | bool entered_context = false; |
| 162 | bool skip_this = false; |
| 163 | bool skip_next = false; |
| 164 | bool is_eof = false; |
| 165 | |
| 166 | while (!is_eof) { |
| 167 | String l = f->get_line().strip_edges(); |
| 168 | is_eof = f->eof_reached(); |
| 169 | |
| 170 | // If we reached last line and it's not a content line, break, otherwise let processing that last loop |
| 171 | if (is_eof && l.is_empty()) { |
| 172 | if (status == STATUS_READING_ID || status == STATUS_READING_CONTEXT || (status == STATUS_READING_PLURAL && plural_index != plural_forms - 1)) { |
| 173 | ERR_FAIL_V_MSG(Ref<Resource>(), "Unexpected EOF while reading PO file at: " + path + ":" + itos(line)); |
| 174 | } else { |
| 175 | break; |
| 176 | } |
| 177 | } |
| 178 | |
| 179 | if (l.begins_with("msgctxt" )) { |
| 180 | ERR_FAIL_COND_V_MSG(status != STATUS_READING_STRING && status != STATUS_READING_PLURAL, Ref<Resource>(), "Unexpected 'msgctxt', was expecting 'msgid_plural' or 'msgstr' before 'msgctxt' while parsing: " + path + ":" + itos(line)); |
| 181 | |
| 182 | // In PO file, "msgctxt" appears before "msgid". If we encounter a "msgctxt", we add what we have read |
| 183 | // and set "entered_context" to true to prevent adding twice. |
| 184 | if (!skip_this && !msg_id.is_empty()) { |
| 185 | if (status == STATUS_READING_STRING) { |
| 186 | translation->add_message(msg_id, msg_str, msg_context); |
| 187 | } else if (status == STATUS_READING_PLURAL) { |
| 188 | ERR_FAIL_COND_V_MSG(plural_index != plural_forms - 1, Ref<Resource>(), "Number of 'msgstr[]' doesn't match with number of plural forms: " + path + ":" + itos(line)); |
| 189 | translation->add_plural_message(msg_id, msgs_plural, msg_context); |
| 190 | } |
| 191 | } |
| 192 | msg_context = "" ; |
| 193 | l = l.substr(7, l.length()).strip_edges(); |
| 194 | status = STATUS_READING_CONTEXT; |
| 195 | entered_context = true; |
| 196 | } |
| 197 | |
| 198 | if (l.begins_with("msgid_plural" )) { |
| 199 | if (plural_forms == 0) { |
| 200 | ERR_FAIL_V_MSG(Ref<Resource>(), "PO file uses 'msgid_plural' but 'Plural-Forms' is invalid or missing in header: " + path + ":" + itos(line)); |
| 201 | } else if (status != STATUS_READING_ID) { |
| 202 | ERR_FAIL_V_MSG(Ref<Resource>(), "Unexpected 'msgid_plural', was expecting 'msgid' before 'msgid_plural' while parsing: " + path + ":" + itos(line)); |
| 203 | } |
| 204 | // We don't record the message in "msgid_plural" itself as tr_n(), TTRN(), RTRN() interfaces provide the plural string already. |
| 205 | // We just have to reset variables related to plurals for "msgstr[]" later on. |
| 206 | l = l.substr(12, l.length()).strip_edges(); |
| 207 | plural_index = -1; |
| 208 | msgs_plural.clear(); |
| 209 | msgs_plural.resize(plural_forms); |
| 210 | status = STATUS_READING_PLURAL; |
| 211 | } else if (l.begins_with("msgid" )) { |
| 212 | ERR_FAIL_COND_V_MSG(status == STATUS_READING_ID, Ref<Resource>(), "Unexpected 'msgid', was expecting 'msgstr' while parsing: " + path + ":" + itos(line)); |
| 213 | |
| 214 | if (!msg_id.is_empty()) { |
| 215 | if (!skip_this && !entered_context) { |
| 216 | if (status == STATUS_READING_STRING) { |
| 217 | translation->add_message(msg_id, msg_str, msg_context); |
| 218 | } else if (status == STATUS_READING_PLURAL) { |
| 219 | ERR_FAIL_COND_V_MSG(plural_index != plural_forms - 1, Ref<Resource>(), "Number of 'msgstr[]' doesn't match with number of plural forms: " + path + ":" + itos(line)); |
| 220 | translation->add_plural_message(msg_id, msgs_plural, msg_context); |
| 221 | } |
| 222 | } |
| 223 | } else if (config.is_empty()) { |
| 224 | config = msg_str; |
| 225 | // Record plural rule. |
| 226 | int p_start = config.find("Plural-Forms" ); |
| 227 | if (p_start != -1) { |
| 228 | int p_end = config.find("\n" , p_start); |
| 229 | translation->set_plural_rule(config.substr(p_start, p_end - p_start)); |
| 230 | plural_forms = translation->get_plural_forms(); |
| 231 | } |
| 232 | } |
| 233 | |
| 234 | l = l.substr(5, l.length()).strip_edges(); |
| 235 | status = STATUS_READING_ID; |
| 236 | // If we did not encounter msgctxt, we reset context to empty to reset it. |
| 237 | if (!entered_context) { |
| 238 | msg_context = "" ; |
| 239 | } |
| 240 | msg_id = "" ; |
| 241 | msg_str = "" ; |
| 242 | skip_this = skip_next; |
| 243 | skip_next = false; |
| 244 | entered_context = false; |
| 245 | } |
| 246 | |
| 247 | if (l.begins_with("msgstr[" )) { |
| 248 | ERR_FAIL_COND_V_MSG(status != STATUS_READING_PLURAL, Ref<Resource>(), "Unexpected 'msgstr[]', was expecting 'msgid_plural' before 'msgstr[]' while parsing: " + path + ":" + itos(line)); |
| 249 | plural_index++; // Increment to add to the next slot in vector msgs_plural. |
| 250 | l = l.substr(9, l.length()).strip_edges(); |
| 251 | } else if (l.begins_with("msgstr" )) { |
| 252 | ERR_FAIL_COND_V_MSG(status != STATUS_READING_ID, Ref<Resource>(), "Unexpected 'msgstr', was expecting 'msgid' before 'msgstr' while parsing: " + path + ":" + itos(line)); |
| 253 | l = l.substr(6, l.length()).strip_edges(); |
| 254 | status = STATUS_READING_STRING; |
| 255 | } |
| 256 | |
| 257 | if (l.is_empty() || l.begins_with("#" )) { |
| 258 | if (l.contains("fuzzy" )) { |
| 259 | skip_next = true; |
| 260 | } |
| 261 | line++; |
| 262 | continue; // Nothing to read or comment. |
| 263 | } |
| 264 | |
| 265 | ERR_FAIL_COND_V_MSG(!l.begins_with("\"" ) || status == STATUS_NONE, Ref<Resource>(), "Invalid line '" + l + "' while parsing: " + path + ":" + itos(line)); |
| 266 | |
| 267 | l = l.substr(1, l.length()); |
| 268 | // Find final quote, ignoring escaped ones (\"). |
| 269 | // The escape_next logic is necessary to properly parse things like \\" |
| 270 | // where the backslash is the one being escaped, not the quote. |
| 271 | int end_pos = -1; |
| 272 | bool escape_next = false; |
| 273 | for (int i = 0; i < l.length(); i++) { |
| 274 | if (l[i] == '\\' && !escape_next) { |
| 275 | escape_next = true; |
| 276 | continue; |
| 277 | } |
| 278 | |
| 279 | if (l[i] == '"' && !escape_next) { |
| 280 | end_pos = i; |
| 281 | break; |
| 282 | } |
| 283 | |
| 284 | escape_next = false; |
| 285 | } |
| 286 | |
| 287 | ERR_FAIL_COND_V_MSG(end_pos == -1, Ref<Resource>(), "Expected '\"' at end of message while parsing: " + path + ":" + itos(line)); |
| 288 | |
| 289 | l = l.substr(0, end_pos); |
| 290 | l = l.c_unescape(); |
| 291 | |
| 292 | if (status == STATUS_READING_ID) { |
| 293 | msg_id += l; |
| 294 | } else if (status == STATUS_READING_STRING) { |
| 295 | msg_str += l; |
| 296 | } else if (status == STATUS_READING_CONTEXT) { |
| 297 | msg_context += l; |
| 298 | } else if (status == STATUS_READING_PLURAL && plural_index >= 0) { |
| 299 | ERR_FAIL_COND_V_MSG(plural_index >= plural_forms, Ref<Resource>(), "Unexpected plural form while parsing: " + path + ":" + itos(line)); |
| 300 | msgs_plural.write[plural_index] = msgs_plural[plural_index] + l; |
| 301 | } |
| 302 | |
| 303 | line++; |
| 304 | } |
| 305 | |
| 306 | // Add the last set of data from last iteration. |
| 307 | if (status == STATUS_READING_STRING) { |
| 308 | if (!msg_id.is_empty()) { |
| 309 | if (!skip_this) { |
| 310 | translation->add_message(msg_id, msg_str, msg_context); |
| 311 | } |
| 312 | } else if (config.is_empty()) { |
| 313 | config = msg_str; |
| 314 | } |
| 315 | } else if (status == STATUS_READING_PLURAL) { |
| 316 | if (!skip_this && !msg_id.is_empty()) { |
| 317 | ERR_FAIL_COND_V_MSG(plural_index != plural_forms - 1, Ref<Resource>(), "Number of 'msgstr[]' doesn't match with number of plural forms: " + path + ":" + itos(line)); |
| 318 | translation->add_plural_message(msg_id, msgs_plural, msg_context); |
| 319 | } |
| 320 | } |
| 321 | } |
| 322 | |
| 323 | ERR_FAIL_COND_V_MSG(config.is_empty(), Ref<Resource>(), "No config found in file: " + path + "." ); |
| 324 | |
| 325 | Vector<String> configs = config.split("\n" ); |
| 326 | for (int i = 0; i < configs.size(); i++) { |
| 327 | String c = configs[i].strip_edges(); |
| 328 | int p = c.find(":" ); |
| 329 | if (p == -1) { |
| 330 | continue; |
| 331 | } |
| 332 | String prop = c.substr(0, p).strip_edges(); |
| 333 | String value = c.substr(p + 1, c.length()).strip_edges(); |
| 334 | |
| 335 | if (prop == "X-Language" || prop == "Language" ) { |
| 336 | translation->set_locale(value); |
| 337 | } |
| 338 | } |
| 339 | |
| 340 | if (r_error) { |
| 341 | *r_error = OK; |
| 342 | } |
| 343 | |
| 344 | return translation; |
| 345 | } |
| 346 | |
| 347 | Ref<Resource> TranslationLoaderPO::load(const String &p_path, const String &p_original_path, Error *r_error, bool p_use_sub_threads, float *r_progress, CacheMode p_cache_mode) { |
| 348 | if (r_error) { |
| 349 | *r_error = ERR_CANT_OPEN; |
| 350 | } |
| 351 | |
| 352 | Ref<FileAccess> f = FileAccess::open(p_path, FileAccess::READ); |
| 353 | ERR_FAIL_COND_V_MSG(f.is_null(), Ref<Resource>(), "Cannot open file '" + p_path + "'." ); |
| 354 | |
| 355 | return load_translation(f, r_error); |
| 356 | } |
| 357 | |
| 358 | void TranslationLoaderPO::get_recognized_extensions(List<String> *p_extensions) const { |
| 359 | p_extensions->push_back("po" ); |
| 360 | p_extensions->push_back("mo" ); |
| 361 | } |
| 362 | |
| 363 | bool TranslationLoaderPO::handles_type(const String &p_type) const { |
| 364 | return (p_type == "Translation" ); |
| 365 | } |
| 366 | |
| 367 | String TranslationLoaderPO::get_resource_type(const String &p_path) const { |
| 368 | if (p_path.get_extension().to_lower() == "po" || p_path.get_extension().to_lower() == "mo" ) { |
| 369 | return "Translation" ; |
| 370 | } |
| 371 | return "" ; |
| 372 | } |
| 373 | |