00001
00002
00003
00004
00005
00006
00007
00008
00009
00012 #include "../stdafx.h"
00013 #include "../core/endian_func.hpp"
00014 #include "../string_func.h"
00015 #include "../table/control_codes.h"
00016
00017 #include "strgen.h"
00018
00019
00020 #include "../table/strgen_tables.h"
00021
00022
00023
00024 static bool _translated;
00025 static bool _translation;
00026 const char *_file = "(unknown file)";
00027 int _cur_line;
00028 int _errors, _warnings, _show_todo;
00029 LanguagePackHeader _lang;
00030
00031 static const ptrdiff_t MAX_COMMAND_PARAM_SIZE = 100;
00032 static const CmdStruct *ParseCommandString(const char **str, char *param, int *argno, int *casei);
00033
00040 Case::Case(int caseidx, const char *string, Case *next) :
00041 caseidx(caseidx), string(strdup(string)), next(next)
00042 {
00043 }
00044
00046 Case::~Case()
00047 {
00048 free(this->string);
00049 delete this->next;
00050 }
00051
00059 LangString::LangString(const char *name, const char *english, int index, int line) :
00060 name(strdup(name)), english(strdup(english)), translated(NULL),
00061 hash_next(0), index(index), line(line), translated_case(NULL)
00062 {
00063 }
00064
00066 LangString::~LangString()
00067 {
00068 free(this->name);
00069 free(this->english);
00070 free(this->translated);
00071 delete this->translated_case;
00072 }
00073
00075 void LangString::FreeTranslation()
00076 {
00077 free(this->translated);
00078 this->translated = NULL;
00079
00080 delete this->translated_case;
00081 this->translated_case = NULL;
00082 }
00083
00088 StringData::StringData(size_t tabs) : tabs(tabs), max_strings(tabs * TAB_SIZE)
00089 {
00090 this->strings = CallocT<LangString *>(max_strings);
00091 this->hash_heads = CallocT<uint16>(max_strings);
00092 this->next_string_id = 0;
00093 }
00094
00096 StringData::~StringData()
00097 {
00098 for (size_t i = 0; i < this->max_strings; i++) delete this->strings[i];
00099 free(this->strings);
00100 free(this->hash_heads);
00101 }
00102
00104 void StringData::FreeTranslation()
00105 {
00106 for (size_t i = 0; i < this->max_strings; i++) {
00107 LangString *ls = this->strings[i];
00108 if (ls != NULL) ls->FreeTranslation();
00109 }
00110 }
00111
00117 uint StringData::HashStr(const char *s) const
00118 {
00119 uint hash = 0;
00120 for (; *s != '\0'; s++) hash = ROL(hash, 3) ^ *s;
00121 return hash % this->max_strings;
00122 }
00123
00129 void StringData::Add(const char *s, LangString *ls)
00130 {
00131 uint hash = this->HashStr(s);
00132 ls->hash_next = this->hash_heads[hash];
00133
00134 this->hash_heads[hash] = ls->index + 1;
00135 this->strings[ls->index] = ls;
00136 }
00137
00143 LangString *StringData::Find(const char *s)
00144 {
00145 int idx = this->hash_heads[this->HashStr(s)];
00146
00147 while (--idx >= 0) {
00148 LangString *ls = this->strings[idx];
00149
00150 if (strcmp(ls->name, s) == 0) return ls;
00151 idx = ls->hash_next;
00152 }
00153 return NULL;
00154 }
00155
00162 uint StringData::VersionHashStr(uint hash, const char *s) const
00163 {
00164 for (; *s != '\0'; s++) {
00165 hash = ROL(hash, 3) ^ *s;
00166 hash = (hash & 1 ? hash >> 1 ^ 0xDEADBEEF : hash >> 1);
00167 }
00168 return hash;
00169 }
00170
00175 uint StringData::Version() const
00176 {
00177 uint hash = 0;
00178
00179 for (size_t i = 0; i < this->max_strings; i++) {
00180 const LangString *ls = this->strings[i];
00181
00182 if (ls != NULL) {
00183 const CmdStruct *cs;
00184 const char *s;
00185 char buf[MAX_COMMAND_PARAM_SIZE];
00186 int argno;
00187 int casei;
00188
00189 s = ls->name;
00190 hash ^= i * 0x717239;
00191 hash = (hash & 1 ? hash >> 1 ^ 0xDEADBEEF : hash >> 1);
00192 hash = this->VersionHashStr(hash, s + 1);
00193
00194 s = ls->english;
00195 while ((cs = ParseCommandString(&s, buf, &argno, &casei)) != NULL) {
00196 if (cs->flags & C_DONTCOUNT) continue;
00197
00198 hash ^= (cs - _cmd_structs) * 0x1234567;
00199 hash = (hash & 1 ? hash >> 1 ^ 0xF00BAA4 : hash >> 1);
00200 }
00201 }
00202 }
00203
00204 return hash;
00205 }
00206
00211 uint StringData::CountInUse(uint tab) const
00212 {
00213 int i;
00214 for (i = TAB_SIZE; --i >= 0;) if (this->strings[(tab * TAB_SIZE) + i] != NULL) break;
00215 return i + 1;
00216 }
00217
00218 static const char *_cur_ident;
00219
00220 struct CmdPair {
00221 const CmdStruct *a;
00222 const char *v;
00223 };
00224
00225 struct ParsedCommandStruct {
00226 uint np;
00227 CmdPair pairs[32];
00228 const CmdStruct *cmd[32];
00229 };
00230
00231
00232 static ParsedCommandStruct _cur_pcs;
00233 static int _cur_argidx;
00234
00236 struct Buffer : SmallVector<byte, 256> {
00241 void AppendByte(byte value)
00242 {
00243 *this->Append() = value;
00244 }
00245
00250 void AppendUtf8(uint32 value)
00251 {
00252 if (value < 0x80) {
00253 *this->Append() = value;
00254 } else if (value < 0x800) {
00255 *this->Append() = 0xC0 + GB(value, 6, 5);
00256 *this->Append() = 0x80 + GB(value, 0, 6);
00257 } else if (value < 0x10000) {
00258 *this->Append() = 0xE0 + GB(value, 12, 4);
00259 *this->Append() = 0x80 + GB(value, 6, 6);
00260 *this->Append() = 0x80 + GB(value, 0, 6);
00261 } else if (value < 0x110000) {
00262 *this->Append() = 0xF0 + GB(value, 18, 3);
00263 *this->Append() = 0x80 + GB(value, 12, 6);
00264 *this->Append() = 0x80 + GB(value, 6, 6);
00265 *this->Append() = 0x80 + GB(value, 0, 6);
00266 } else {
00267 strgen_warning("Invalid unicode value U+0x%X", value);
00268 }
00269 }
00270 };
00271
00272 size_t Utf8Validate(const char *s)
00273 {
00274 uint32 c;
00275
00276 if (!HasBit(s[0], 7)) {
00277
00278 return 1;
00279 } else if (GB(s[0], 5, 3) == 6 && IsUtf8Part(s[1])) {
00280
00281 c = GB(s[0], 0, 5) << 6 | GB(s[1], 0, 6);
00282 if (c >= 0x80) return 2;
00283 } else if (GB(s[0], 4, 4) == 14 && IsUtf8Part(s[1]) && IsUtf8Part(s[2])) {
00284
00285 c = GB(s[0], 0, 4) << 12 | GB(s[1], 0, 6) << 6 | GB(s[2], 0, 6);
00286 if (c >= 0x800) return 3;
00287 } else if (GB(s[0], 3, 5) == 30 && IsUtf8Part(s[1]) && IsUtf8Part(s[2]) && IsUtf8Part(s[3])) {
00288
00289 c = GB(s[0], 0, 3) << 18 | GB(s[1], 0, 6) << 12 | GB(s[2], 0, 6) << 6 | GB(s[3], 0, 6);
00290 if (c >= 0x10000 && c <= 0x10FFFF) return 4;
00291 }
00292
00293 return 0;
00294 }
00295
00296
00297 void EmitSingleChar(Buffer *buffer, char *buf, int value)
00298 {
00299 if (*buf != '\0') strgen_warning("Ignoring trailing letters in command");
00300 buffer->AppendUtf8(value);
00301 }
00302
00303
00304
00305
00306
00307
00308
00309
00310 bool ParseRelNum(char **buf, int *value, int *offset)
00311 {
00312 const char *s = *buf;
00313 char *end;
00314 bool rel = false;
00315
00316 while (*s == ' ' || *s == '\t') s++;
00317 if (*s == '+') {
00318 rel = true;
00319 s++;
00320 }
00321 int v = strtol(s, &end, 0);
00322 if (end == s) return false;
00323 if (rel || v < 0) {
00324 *value += v;
00325 } else {
00326 *value = v;
00327 }
00328 if (offset != NULL && *end == ':') {
00329
00330 s = end + 1;
00331 *offset = strtol(s, &end, 0);
00332 if (end == s) return false;
00333 }
00334 *buf = end;
00335 return true;
00336 }
00337
00338
00339 char *ParseWord(char **buf)
00340 {
00341 char *s = *buf, *r;
00342
00343 while (*s == ' ' || *s == '\t') s++;
00344 if (*s == '\0') return NULL;
00345
00346 if (*s == '"') {
00347 r = ++s;
00348
00349 for (;;) {
00350 if (*s == '\0') break;
00351 if (*s == '"') {
00352 *s++ = '\0';
00353 break;
00354 }
00355 s++;
00356 }
00357 } else {
00358
00359 r = s;
00360 for (;;) {
00361 if (*s == '\0') break;
00362 if (*s == ' ' || *s == '\t') {
00363 *s++ = '\0';
00364 break;
00365 }
00366 s++;
00367 }
00368 }
00369 *buf = s;
00370 return r;
00371 }
00372
00373
00374 static int TranslateArgumentIdx(int arg, int offset = 0);
00375
00376 static void EmitWordList(Buffer *buffer, const char * const *words, uint nw)
00377 {
00378 buffer->AppendByte(nw);
00379 for (uint i = 0; i < nw; i++) buffer->AppendByte((uint)strlen(words[i]) + 1);
00380 for (uint i = 0; i < nw; i++) {
00381 for (uint j = 0; words[i][j] != '\0'; j++) buffer->AppendByte(words[i][j]);
00382 buffer->AppendByte(0);
00383 }
00384 }
00385
00386 void EmitPlural(Buffer *buffer, char *buf, int value)
00387 {
00388 int argidx = _cur_argidx;
00389 int offset = 0;
00390 int expected = _plural_forms[_lang.plural_form].plural_count;
00391 const char **words = AllocaM(const char *, max(expected, MAX_PLURALS));
00392 int nw = 0;
00393
00394
00395 if (!ParseRelNum(&buf, &argidx, &offset)) argidx--;
00396
00397
00398 for (nw = 0; nw < MAX_PLURALS; nw++) {
00399 words[nw] = ParseWord(&buf);
00400 if (words[nw] == NULL) break;
00401 }
00402
00403 if (nw == 0) {
00404 strgen_fatal("%s: No plural words", _cur_ident);
00405 }
00406
00407 if (expected != nw) {
00408 if (_translated) {
00409 strgen_fatal("%s: Invalid number of plural forms. Expecting %d, found %d.", _cur_ident,
00410 expected, nw);
00411 } else {
00412 if ((_show_todo & 2) != 0) strgen_warning("'%s' is untranslated. Tweaking english string to allow compilation for plural forms", _cur_ident);
00413 if (nw > expected) {
00414 nw = expected;
00415 } else {
00416 for (; nw < expected; nw++) {
00417 words[nw] = words[nw - 1];
00418 }
00419 }
00420 }
00421 }
00422
00423 buffer->AppendUtf8(SCC_PLURAL_LIST);
00424 buffer->AppendByte(_lang.plural_form);
00425 buffer->AppendByte(TranslateArgumentIdx(argidx, offset));
00426 EmitWordList(buffer, words, nw);
00427 }
00428
00429
00430 void EmitGender(Buffer *buffer, char *buf, int value)
00431 {
00432 int argidx = _cur_argidx;
00433 int offset = 0;
00434 uint nw;
00435
00436 if (buf[0] == '=') {
00437 buf++;
00438
00439
00440 nw = _lang.GetGenderIndex(buf);
00441 if (nw >= MAX_NUM_GENDERS) strgen_fatal("G argument '%s' invalid", buf);
00442
00443
00444 buffer->AppendUtf8(SCC_GENDER_INDEX);
00445 buffer->AppendByte(nw);
00446 } else {
00447 const char *words[MAX_NUM_GENDERS];
00448
00449
00450
00451 if (!ParseRelNum(&buf, &argidx, &offset)) {}
00452
00453 const CmdStruct *cmd = _cur_pcs.cmd[argidx];
00454 if (cmd == NULL || (cmd->flags & C_GENDER) == 0) {
00455 strgen_fatal("Command '%s' can't have a gender", cmd == NULL ? "<empty>" : cmd->cmd);
00456 }
00457
00458 for (nw = 0; nw < MAX_NUM_GENDERS; nw++) {
00459 words[nw] = ParseWord(&buf);
00460 if (words[nw] == NULL) break;
00461 }
00462 if (nw != _lang.num_genders) strgen_fatal("Bad # of arguments for gender command");
00463
00464 assert(IsInsideBS(cmd->value, SCC_CONTROL_START, UINT8_MAX));
00465 buffer->AppendUtf8(SCC_GENDER_LIST);
00466 buffer->AppendByte(TranslateArgumentIdx(argidx, offset));
00467 EmitWordList(buffer, words, nw);
00468 }
00469 }
00470
00471 static const CmdStruct *FindCmd(const char *s, int len)
00472 {
00473 for (const CmdStruct *cs = _cmd_structs; cs != endof(_cmd_structs); cs++) {
00474 if (strncmp(cs->cmd, s, len) == 0 && cs->cmd[len] == '\0') return cs;
00475 }
00476 return NULL;
00477 }
00478
00479 static uint ResolveCaseName(const char *str, size_t len)
00480 {
00481
00482 char case_str[CASE_GENDER_LEN];
00483 len = min(lengthof(case_str) - 1, len);
00484 memcpy(case_str, str, len);
00485 case_str[len] = '\0';
00486
00487 uint8 case_idx = _lang.GetCaseIndex(case_str);
00488 if (case_idx >= MAX_NUM_CASES) strgen_fatal("Invalid case-name '%s'", case_str);
00489 return case_idx + 1;
00490 }
00491
00492
00493
00494
00495 static const CmdStruct *ParseCommandString(const char **str, char *param, int *argno, int *casei)
00496 {
00497 const char *s = *str, *start;
00498 char c;
00499
00500 *argno = -1;
00501 *casei = -1;
00502
00503
00504 for (; *s != '{'; s++) {
00505 if (*s == '\0') return NULL;
00506 }
00507 s++;
00508
00509 if (*s >= '0' && *s <= '9') {
00510 char *end;
00511
00512 *argno = strtoul(s, &end, 0);
00513 if (*end != ':') strgen_fatal("missing arg #");
00514 s = end + 1;
00515 }
00516
00517
00518 start = s;
00519 do {
00520 c = *s++;
00521 } while (c != '}' && c != ' ' && c != '=' && c != '.' && c != 0);
00522
00523 const CmdStruct *cmd = FindCmd(start, s - start - 1);
00524 if (cmd == NULL) {
00525 strgen_error("Undefined command '%.*s'", (int)(s - start - 1), start);
00526 return NULL;
00527 }
00528
00529 if (c == '.') {
00530 const char *casep = s;
00531
00532 if (!(cmd->flags & C_CASE)) {
00533 strgen_fatal("Command '%s' can't have a case", cmd->cmd);
00534 }
00535
00536 do {
00537 c = *s++;
00538 } while (c != '}' && c != ' ' && c != '\0');
00539 *casei = ResolveCaseName(casep, s - casep - 1);
00540 }
00541
00542 if (c == '\0') {
00543 strgen_error("Missing } from command '%s'", start);
00544 return NULL;
00545 }
00546
00547
00548 if (c != '}') {
00549 if (c == '=') s--;
00550
00551 start = s;
00552 for (;;) {
00553 c = *s++;
00554 if (c == '}') break;
00555 if (c == '\0') {
00556 strgen_error("Missing } from command '%s'", start);
00557 return NULL;
00558 }
00559 if (s - start == MAX_COMMAND_PARAM_SIZE) error("param command too long");
00560 *param++ = c;
00561 }
00562 }
00563 *param = '\0';
00564
00565 *str = s;
00566
00567 return cmd;
00568 }
00569
00577 StringReader::StringReader(StringData &data, const char *file, bool master, bool translation) :
00578 data(data), file(strdup(file)), master(master), translation(translation)
00579 {
00580 }
00581
00583 StringReader::~StringReader()
00584 {
00585 free(file);
00586 }
00587
00588 static void ExtractCommandString(ParsedCommandStruct *p, const char *s, bool warnings)
00589 {
00590 char param[MAX_COMMAND_PARAM_SIZE];
00591 int argno;
00592 int argidx = 0;
00593 int casei;
00594
00595 memset(p, 0, sizeof(*p));
00596
00597 for (;;) {
00598
00599 const CmdStruct *ar = ParseCommandString(&s, param, &argno, &casei);
00600
00601 if (ar == NULL) break;
00602
00603
00604 if (argno != -1 && ar->consumes == 0) strgen_fatal("Non consumer param can't have a paramindex");
00605
00606 if (ar->consumes) {
00607 if (argno != -1) argidx = argno;
00608 if (argidx < 0 || (uint)argidx >= lengthof(p->cmd)) strgen_fatal("invalid param idx %d", argidx);
00609 if (p->cmd[argidx] != NULL && p->cmd[argidx] != ar) strgen_fatal("duplicate param idx %d", argidx);
00610
00611 p->cmd[argidx++] = ar;
00612 } else if (!(ar->flags & C_DONTCOUNT)) {
00613 if (p->np >= lengthof(p->pairs)) strgen_fatal("too many commands in string, max " PRINTF_SIZE, lengthof(p->pairs));
00614 p->pairs[p->np].a = ar;
00615 p->pairs[p->np].v = param[0] != '\0' ? strdup(param) : "";
00616 p->np++;
00617 }
00618 }
00619 }
00620
00621
00622 static const CmdStruct *TranslateCmdForCompare(const CmdStruct *a)
00623 {
00624 if (a == NULL) return NULL;
00625
00626 if (strcmp(a->cmd, "STRING1") == 0 ||
00627 strcmp(a->cmd, "STRING2") == 0 ||
00628 strcmp(a->cmd, "STRING3") == 0 ||
00629 strcmp(a->cmd, "STRING4") == 0 ||
00630 strcmp(a->cmd, "STRING5") == 0 ||
00631 strcmp(a->cmd, "STRING6") == 0 ||
00632 strcmp(a->cmd, "STRING7") == 0 ||
00633 strcmp(a->cmd, "RAW_STRING") == 0) {
00634 return FindCmd("STRING", 6);
00635 }
00636
00637 return a;
00638 }
00639
00640
00641 static bool CheckCommandsMatch(char *a, char *b, const char *name)
00642 {
00643
00644
00645
00646
00647 if (!_translation) return true;
00648
00649 ParsedCommandStruct templ;
00650 ParsedCommandStruct lang;
00651 bool result = true;
00652
00653 ExtractCommandString(&templ, b, true);
00654 ExtractCommandString(&lang, a, true);
00655
00656
00657 if (templ.np != lang.np) {
00658 strgen_warning("%s: template string and language string have a different # of commands", name);
00659 result = false;
00660 }
00661
00662 for (uint i = 0; i < templ.np; i++) {
00663
00664 bool found = false;
00665 for (uint j = 0; j < lang.np; j++) {
00666 if (templ.pairs[i].a == lang.pairs[j].a &&
00667 strcmp(templ.pairs[i].v, lang.pairs[j].v) == 0) {
00668
00669 lang.pairs[j].a = NULL;
00670 found = true;
00671 break;
00672 }
00673 }
00674
00675 if (!found) {
00676 strgen_warning("%s: command '%s' exists in template file but not in language file", name, templ.pairs[i].a->cmd);
00677 result = false;
00678 }
00679 }
00680
00681
00682
00683 for (uint i = 0; i < lengthof(templ.cmd); i++) {
00684 if (TranslateCmdForCompare(templ.cmd[i]) != lang.cmd[i]) {
00685 strgen_warning("%s: Param idx #%d '%s' doesn't match with template command '%s'", name, i,
00686 lang.cmd[i] == NULL ? "<empty>" : TranslateCmdForCompare(lang.cmd[i])->cmd,
00687 templ.cmd[i] == NULL ? "<empty>" : templ.cmd[i]->cmd);
00688 result = false;
00689 }
00690 }
00691
00692 return result;
00693 }
00694
00695 void StringReader::HandleString(char *str)
00696 {
00697 if (*str == '#') {
00698 if (str[1] == '#' && str[2] != '#') this->HandlePragma(str + 2);
00699 return;
00700 }
00701
00702
00703 if (*str == ';' || *str == ' ' || *str == '\0') return;
00704
00705 char *s = strchr(str, ':');
00706 if (s == NULL) {
00707 strgen_error("Line has no ':' delimiter");
00708 return;
00709 }
00710
00711 char *t;
00712
00713
00714 for (t = s; t > str && (t[-1] == ' ' || t[-1] == '\t'); t--) {}
00715 *t = 0;
00716 s++;
00717
00718
00719 const char *tmp;
00720 for (tmp = s; *tmp != '\0';) {
00721 size_t len = Utf8Validate(tmp);
00722 if (len == 0) strgen_fatal("Invalid UTF-8 sequence in '%s'", s);
00723
00724 WChar c;
00725 Utf8Decode(&c, tmp);
00726 if (c <= 0x001F ||
00727 c == 0x200B ||
00728 (c >= 0xE000 && c <= 0xF8FF) ||
00729 (c >= 0xFFF0 && c <= 0xFFFF)) {
00730 strgen_fatal("Unwanted UTF-8 character U+%04X in sequence '%s'", c, s);
00731 }
00732
00733 tmp += len;
00734 }
00735
00736
00737
00738 char *casep = strchr(str, '.');
00739 if (casep != NULL) *casep++ = '\0';
00740
00741
00742 LangString *ent = this->data.Find(str);
00743
00744 if (this->master) {
00745 if (casep != NULL) {
00746 strgen_error("Cases in the base translation are not supported.");
00747 return;
00748 }
00749
00750 if (ent != NULL) {
00751 strgen_error("String name '%s' is used multiple times", str);
00752 return;
00753 }
00754
00755 if (this->data.strings[this->data.next_string_id] != NULL) {
00756 strgen_error("String ID 0x%X for '%s' already in use by '%s'", this->data.next_string_id, str, this->data.strings[this->data.next_string_id]->name);
00757 return;
00758 }
00759
00760
00761 this->data.Add(str, new LangString(str, s, this->data.next_string_id++, _cur_line));
00762 } else {
00763 if (ent == NULL) {
00764 strgen_warning("String name '%s' does not exist in master file", str);
00765 return;
00766 }
00767
00768 if (ent->translated && casep == NULL) {
00769 strgen_error("String name '%s' is used multiple times", str);
00770 return;
00771 }
00772
00773
00774 if (!CheckCommandsMatch(s, ent->english, str)) return;
00775
00776 if (casep != NULL) {
00777 ent->translated_case = new Case(ResolveCaseName(casep, strlen(casep)), s, ent->translated_case);
00778 } else {
00779 ent->translated = strdup(s);
00780
00781
00782
00783 ent->line = _cur_line;
00784 }
00785 }
00786 }
00787
00788 void StringReader::HandlePragma(char *str)
00789 {
00790 if (!memcmp(str, "plural ", 7)) {
00791 _lang.plural_form = atoi(str + 7);
00792 if (_lang.plural_form >= lengthof(_plural_forms)) {
00793 strgen_fatal("Invalid pluralform %d", _lang.plural_form);
00794 }
00795 } else {
00796 strgen_fatal("unknown pragma '%s'", str);
00797 }
00798 }
00799
00800 static void rstrip(char *buf)
00801 {
00802 size_t i = strlen(buf);
00803 while (i > 0 && (buf[i - 1] == '\r' || buf[i - 1] == '\n' || buf[i - 1] == ' ')) i--;
00804 buf[i] = '\0';
00805 }
00806
00807 void StringReader::ParseFile()
00808 {
00809 char buf[2048];
00810 _warnings = _errors = 0;
00811
00812 _translation = this->master || this->translation;
00813 _file = this->file;
00814
00815
00816 MemSetT(&_lang, 0);
00817 strecpy(_lang.digit_group_separator, ",", lastof(_lang.digit_group_separator));
00818 strecpy(_lang.digit_group_separator_currency, ",", lastof(_lang.digit_group_separator_currency));
00819 strecpy(_lang.digit_decimal_separator, ".", lastof(_lang.digit_decimal_separator));
00820
00821 _cur_line = 1;
00822 while (this->ReadLine(buf, sizeof(buf)) != NULL) {
00823 rstrip(buf);
00824 this->HandleString(buf);
00825 _cur_line++;
00826 }
00827 }
00828
00833 void HeaderWriter::WriteHeader(const StringData &data)
00834 {
00835 int last = 0;
00836 for (size_t i = 0; i < data.max_strings; i++) {
00837 if (data.strings[i] != NULL) {
00838 this->WriteStringID(data.strings[i]->name, (int)i);
00839 last = (int)i;
00840 }
00841 }
00842
00843 this->WriteStringID("STR_LAST_STRINGID", last);
00844 }
00845
00846 static int TranslateArgumentIdx(int argidx, int offset)
00847 {
00848 int sum;
00849
00850 if (argidx < 0 || (uint)argidx >= lengthof(_cur_pcs.cmd)) {
00851 strgen_fatal("invalid argidx %d", argidx);
00852 }
00853 const CmdStruct *cs = _cur_pcs.cmd[argidx];
00854 if (cs != NULL && cs->consumes <= offset) {
00855 strgen_fatal("invalid argidx offset %d:%d", argidx, offset);
00856 }
00857
00858 if (_cur_pcs.cmd[argidx] == NULL) {
00859 strgen_fatal("no command for this argidx %d", argidx);
00860 }
00861
00862 for (int i = sum = 0; i < argidx; i++) {
00863 const CmdStruct *cs = _cur_pcs.cmd[i];
00864
00865 sum += (cs != NULL) ? cs->consumes : 1;
00866 }
00867
00868 return sum + offset;
00869 }
00870
00871 static void PutArgidxCommand(Buffer *buffer)
00872 {
00873 buffer->AppendUtf8(SCC_ARG_INDEX);
00874 buffer->AppendByte(TranslateArgumentIdx(_cur_argidx));
00875 }
00876
00877
00878 static void PutCommandString(Buffer *buffer, const char *str)
00879 {
00880 _cur_argidx = 0;
00881
00882 while (*str != '\0') {
00883
00884 if (*str != '{') {
00885 buffer->AppendByte(*str++);
00886 continue;
00887 }
00888
00889 char param[MAX_COMMAND_PARAM_SIZE];
00890 int argno;
00891 int casei;
00892 const CmdStruct *cs = ParseCommandString(&str, param, &argno, &casei);
00893 if (cs == NULL) break;
00894
00895 if (casei != -1) {
00896 buffer->AppendUtf8(SCC_SET_CASE);
00897 buffer->AppendByte(casei);
00898 }
00899
00900
00901 if (cs->consumes > 0) {
00902
00903 if (argno != -1 && argno != _cur_argidx) {
00904 _cur_argidx = argno;
00905 PutArgidxCommand(buffer);
00906 }
00907
00908
00909 cs = _cur_pcs.cmd[_cur_argidx++];
00910 if (cs == NULL) {
00911 strgen_fatal("%s: No argument exists at position %d", _cur_ident, _cur_argidx - 1);
00912 }
00913 }
00914
00915 cs->proc(buffer, param, cs->value);
00916 }
00917 }
00918
00923 void LanguageWriter::WriteLength(uint length)
00924 {
00925 char buffer[2];
00926 int offs = 0;
00927 if (length >= 0x4000) {
00928 strgen_fatal("string too long");
00929 }
00930
00931 if (length >= 0xC0) {
00932 buffer[offs++] = (length >> 8) | 0xC0;
00933 }
00934 buffer[offs++] = length & 0xFF;
00935 this->Write((byte*)buffer, offs);
00936 }
00937
00942 void LanguageWriter::WriteLang(const StringData &data)
00943 {
00944 uint *in_use = AllocaM(uint, data.tabs);
00945 for (size_t tab = 0; tab < data.tabs; tab++) {
00946 uint n = data.CountInUse((uint)tab);
00947
00948 in_use[tab] = n;
00949 _lang.offsets[tab] = TO_LE16(n);
00950
00951 for (uint j = 0; j != in_use[tab]; j++) {
00952 const LangString *ls = data.strings[(tab * TAB_SIZE) + j];
00953 if (ls != NULL && ls->translated == NULL) _lang.missing++;
00954 }
00955 }
00956
00957 _lang.ident = TO_LE32(LanguagePackHeader::IDENT);
00958 _lang.version = TO_LE32(data.Version());
00959 _lang.missing = TO_LE16(_lang.missing);
00960 _lang.winlangid = TO_LE16(_lang.winlangid);
00961
00962 this->WriteHeader(&_lang);
00963 Buffer buffer;
00964
00965 for (size_t tab = 0; tab < data.tabs; tab++) {
00966 for (uint j = 0; j != in_use[tab]; j++) {
00967 const LangString *ls = data.strings[(tab * TAB_SIZE) + j];
00968 const Case *casep;
00969 const char *cmdp;
00970
00971
00972 if (ls == NULL) {
00973 this->WriteLength(0);
00974 continue;
00975 }
00976
00977 _cur_ident = ls->name;
00978 _cur_line = ls->line;
00979
00980
00981 if (_show_todo > 0 && ls->translated == NULL) {
00982 if ((_show_todo & 2) != 0) {
00983 strgen_warning("'%s' is untranslated", ls->name);
00984 }
00985 if ((_show_todo & 1) != 0) {
00986 const char *s = "<TODO> ";
00987 while (*s != '\0') buffer.AppendByte(*s++);
00988 }
00989 }
00990
00991
00992 ExtractCommandString(&_cur_pcs, ls->english, false);
00993
00994 if (ls->translated_case != NULL || ls->translated != NULL) {
00995 casep = ls->translated_case;
00996 cmdp = ls->translated;
00997 } else {
00998 casep = NULL;
00999 cmdp = ls->english;
01000 }
01001
01002 _translated = cmdp != ls->english;
01003
01004 if (casep != NULL) {
01005 const Case *c;
01006 uint num;
01007
01008
01009
01010
01011
01012 buffer.AppendUtf8(SCC_SWITCH_CASE);
01013
01014 for (num = 0, c = casep; c; c = c->next) num++;
01015 buffer.AppendByte(num);
01016
01017
01018 for (c = casep; c != NULL; c = c->next) {
01019 buffer.AppendByte(c->caseidx);
01020
01021 uint pos = buffer.Length();
01022 buffer.AppendByte(0);
01023 buffer.AppendByte(0);
01024
01025 PutCommandString(&buffer, c->string);
01026 buffer.AppendByte(0);
01027
01028 uint size = buffer.Length() - (pos + 2);
01029 buffer[pos + 0] = GB(size, 8, 8);
01030 buffer[pos + 1] = GB(size, 0, 8);
01031 }
01032 }
01033
01034 if (cmdp != NULL) PutCommandString(&buffer, cmdp);
01035
01036 this->WriteLength(buffer.Length());
01037 this->Write(buffer.Begin(), buffer.Length());
01038 buffer.Clear();
01039 }
01040 }
01041 }