| 1 | /* |
| 2 | * LegacyClonk |
| 3 | * |
| 4 | * Copyright (c) RedWolf Design |
| 5 | * Copyright (c) 2005, Sven2 |
| 6 | * Copyright (c) 2017-2022, The LegacyClonk Team and contributors |
| 7 | * |
| 8 | * Distributed under the terms of the ISC license; see accompanying file |
| 9 | * "COPYING" for details. |
| 10 | * |
| 11 | * "Clonk" is a registered trademark of Matthes Bender, used with permission. |
| 12 | * See accompanying file "TRADEMARK" for details. |
| 13 | * |
| 14 | * To redistribute this file separately, substitute the full license texts |
| 15 | * for the above references. |
| 16 | */ |
| 17 | |
| 18 | // RTF file parsing functionality |
| 19 | |
| 20 | #include "C4Include.h" |
| 21 | #include "C4RTF.h" |
| 22 | |
| 23 | #include <algorithm> |
| 24 | |
| 25 | extern C4RTFFile::KeywordTableEntry RTFKeywordTable[]; |
| 26 | |
| 27 | C4RTFFile::C4RTFFile() : pState(nullptr) {} |
| 28 | |
| 29 | C4RTFFile::~C4RTFFile() |
| 30 | { |
| 31 | ClearState(); |
| 32 | } |
| 33 | |
| 34 | void C4RTFFile::ClearState() |
| 35 | { |
| 36 | PropertyState *psNext = pState, *ps; |
| 37 | while (ps = psNext) |
| 38 | { |
| 39 | psNext = ps->pNext; |
| 40 | delete ps; |
| 41 | } |
| 42 | pState = nullptr; |
| 43 | fSkipDestIfUnknownKeyword = false; |
| 44 | } |
| 45 | |
| 46 | void C4RTFFile::AssertNoEOF(size_t iPos) |
| 47 | { |
| 48 | if (iPos >= sRTF.getSize()) throw ParserError("Unexpected end of file" ); |
| 49 | } |
| 50 | |
| 51 | void C4RTFFile::ChangeDest(StdStrBuf &sResult, int iDest) |
| 52 | { |
| 53 | // nothing to do if text is already skipped |
| 54 | if (pState->dest == dsSkip) return; |
| 55 | // otherwise, set new dest (always skip) |
| 56 | pState->dest = static_cast<DestState>(iDest); |
| 57 | } |
| 58 | |
| 59 | void C4RTFFile::SpecialKeyword(StdStrBuf &sResult, int iKeyw, int iParam) |
| 60 | { |
| 61 | switch (iKeyw) |
| 62 | { |
| 63 | case specBin: |
| 64 | if (iParam > 0) |
| 65 | { |
| 66 | pState->eState = psBinary; |
| 67 | pState->iHexBinCnt = iParam; |
| 68 | } |
| 69 | break; |
| 70 | case specHex: |
| 71 | pState->eState = psHex; |
| 72 | pState->iHexBinCnt = 2; |
| 73 | break; |
| 74 | case specSkipDest: |
| 75 | fSkipDestIfUnknownKeyword = true; |
| 76 | break; |
| 77 | }; |
| 78 | } |
| 79 | |
| 80 | void C4RTFFile::TranslateKeyword(StdStrBuf &sResult, const char *szKeyword, int iParam, bool fHasIntParam) |
| 81 | { |
| 82 | // get keyword from table |
| 83 | KeywordTableEntry *pKw = RTFKeywordTable; |
| 84 | while (pKw->szKeyword) if (!strcmp(s1: szKeyword, s2: pKw->szKeyword)) break; else ++pKw; |
| 85 | // no found? |
| 86 | if (!pKw->szKeyword) |
| 87 | { |
| 88 | // unknown destination: Skip |
| 89 | if (fSkipDestIfUnknownKeyword) |
| 90 | { |
| 91 | pState->dest = dsSkip; |
| 92 | fSkipDestIfUnknownKeyword = false; |
| 93 | } |
| 94 | return; |
| 95 | } |
| 96 | fSkipDestIfUnknownKeyword = false; |
| 97 | // keyword known - handle it |
| 98 | switch (pKw->eType) |
| 99 | { |
| 100 | case KeywordTableEntry::kwdProp: |
| 101 | // property: Use default param if none given or forced |
| 102 | ApplyPropChange(iProp: pKw->idx, iParam: (pKw->fForceDefaultParam || !fHasIntParam) ? pKw->iDefaultParam : iParam); |
| 103 | break; |
| 104 | case KeywordTableEntry::kwdChars: |
| 105 | // direct chars |
| 106 | ParseChars(sResult, szChars: pKw->szChars); |
| 107 | break; |
| 108 | case KeywordTableEntry::kwdDest: |
| 109 | ChangeDest(sResult, iDest: pKw->idx); |
| 110 | break; |
| 111 | case KeywordTableEntry::kwdSpec: |
| 112 | SpecialKeyword(sResult, iKeyw: pKw->idx, iParam); |
| 113 | break; |
| 114 | } |
| 115 | } |
| 116 | |
| 117 | void C4RTFFile::ParseKeyword(StdStrBuf &sResult, size_t &iPos) |
| 118 | { |
| 119 | bool fHasIntParam = false; |
| 120 | int iSign = +1; |
| 121 | int iParamInt = 0; // parameter as integer |
| 122 | char szKeyword[30 + 1]; *szKeyword = 0; |
| 123 | char szParameter[20 + 1]; *szParameter = 0; |
| 124 | |
| 125 | AssertNoEOF(iPos); |
| 126 | const auto rtfString = static_cast<const char *>(sRTF.getData()); |
| 127 | char c = rtfString[iPos++]; |
| 128 | if (!isalpha(static_cast<unsigned char>(c))) |
| 129 | { |
| 130 | // parse direct control symbol |
| 131 | szKeyword[0] = c; |
| 132 | szKeyword[1] = 0; |
| 133 | } |
| 134 | else |
| 135 | { |
| 136 | // get keyword string |
| 137 | char *szWrite = szKeyword; |
| 138 | do |
| 139 | { |
| 140 | *szWrite = c; |
| 141 | // do not overflow buffer - longer keywords will be read, not recognized and silently discarded |
| 142 | if (szWrite - szKeyword < 30) ++szWrite; |
| 143 | // do not go past rtf file |
| 144 | if (iPos >= sRTF.getSize()) break; |
| 145 | // next char |
| 146 | c = rtfString[iPos++]; |
| 147 | } while (isalpha(static_cast<unsigned char>(c))); |
| 148 | *szWrite = 0; |
| 149 | // parameter is a negative number? |
| 150 | if (c == '-') |
| 151 | { |
| 152 | iSign = -1; |
| 153 | if (iPos < sRTF.getSize()) c = rtfString[iPos++]; |
| 154 | } |
| 155 | if (isdigit(static_cast<unsigned char>(c))) |
| 156 | { |
| 157 | // get parameter as number |
| 158 | char *szWrite = szParameter; |
| 159 | do |
| 160 | { |
| 161 | *szWrite = c; |
| 162 | // do not overflow buffer - longer parameters will be read, not recognized and silently discarded |
| 163 | if (szWrite - szParameter < 20) ++szWrite; |
| 164 | // do not go past rtf file |
| 165 | if (iPos >= sRTF.getSize()) break; |
| 166 | // next char |
| 167 | c = rtfString[iPos++]; |
| 168 | } while (isdigit(static_cast<unsigned char>(c))); |
| 169 | *szWrite = 0; |
| 170 | iParamInt = atoi(nptr: szParameter) * iSign; |
| 171 | fHasIntParam = true; |
| 172 | } |
| 173 | // if next char is not a spacing for the command, it does not belong to the keyword and must be re-parsed |
| 174 | if (c != ' ') --iPos; |
| 175 | } |
| 176 | // execute keyword action |
| 177 | TranslateKeyword(sResult, szKeyword, iParam: iParamInt, fHasIntParam); |
| 178 | } |
| 179 | |
| 180 | void C4RTFFile::ParseChar(StdStrBuf &sResult, char c) |
| 181 | { |
| 182 | // parse as 1-char-string |
| 183 | char buf[2]; |
| 184 | buf[0] = c; buf[1] = '\0'; |
| 185 | ParseChars(sResult, szChars: buf); |
| 186 | } |
| 187 | |
| 188 | void C4RTFFile::ParseChars(StdStrBuf &sResult, const char *szChars) |
| 189 | { |
| 190 | // route the characters to the appropriate destination stream. |
| 191 | switch (pState->dest) |
| 192 | { |
| 193 | case dsNormal: |
| 194 | // process characters: Append to result buffer |
| 195 | sResult.Append(pnData: szChars); |
| 196 | break; |
| 197 | |
| 198 | case dsSkip: |
| 199 | // skip character |
| 200 | break; |
| 201 | } |
| 202 | } |
| 203 | |
| 204 | void C4RTFFile::ParseHexChar(StdStrBuf &sResult, char c) |
| 205 | { |
| 206 | pState->bHex = pState->bHex << 4; |
| 207 | if (isdigit(static_cast<unsigned char>(c))) |
| 208 | pState->bHex += c - '0'; |
| 209 | else if (Inside<char>(ival: c, lbound: 'a', rbound: 'f')) |
| 210 | pState->bHex += c - 'a' + 10; |
| 211 | else if (Inside<char>(ival: c, lbound: 'A', rbound: 'F')) |
| 212 | pState->bHex += c - 'A' + 10; |
| 213 | else |
| 214 | throw ParserError("Invalid hex character" ); |
| 215 | if (!--pState->iHexBinCnt) |
| 216 | { |
| 217 | pState->eState = psNormal; |
| 218 | ParseChar(sResult, c: pState->bHex); |
| 219 | } |
| 220 | } |
| 221 | |
| 222 | void C4RTFFile::PushState() |
| 223 | { |
| 224 | // store current state to new |
| 225 | PropertyState *pNew = new PropertyState(*pState); |
| 226 | pNew->pNext = pState; |
| 227 | // update current state to new; beginning in default parser mode |
| 228 | pState = pNew; |
| 229 | pState->eState = psNormal; |
| 230 | } |
| 231 | |
| 232 | void C4RTFFile::PopState() |
| 233 | { |
| 234 | if (!pState->pNext) throw ParserError("Too many brackets closed" ); |
| 235 | // if the destination ends, finish it |
| 236 | if (pState->dest != pState->pNext->dest) EndGroupAction(); |
| 237 | // return to last state |
| 238 | PropertyState *pKill = pState; |
| 239 | pState = pState->pNext; |
| 240 | delete pKill; |
| 241 | pState->eState = psNormal; |
| 242 | } |
| 243 | |
| 244 | StdStrBuf C4RTFFile::GetPlainText() |
| 245 | { |
| 246 | // clear any previous crap |
| 247 | ClearState(); |
| 248 | // start with a fresh state |
| 249 | pState = new PropertyState(); |
| 250 | pState->eState = psNormal; |
| 251 | StdStrBuf sResult; |
| 252 | // nothing to do for empty RTFs |
| 253 | if (sRTF.getSize() <= 0) return sResult; |
| 254 | // parse through all chars |
| 255 | try |
| 256 | { |
| 257 | char c; size_t iPos = 0; |
| 258 | while (iPos < sRTF.getSize()) |
| 259 | { |
| 260 | c = reinterpret_cast<const char *>(sRTF.getData())[iPos++]; |
| 261 | // binary parsing? |
| 262 | if (pState->eState == psBinary) |
| 263 | { |
| 264 | if (!--pState->iHexBinCnt) pState->eState = psNormal; |
| 265 | ParseChar(sResult, c); |
| 266 | continue; |
| 267 | } |
| 268 | // normal parsing: Handle state blocks |
| 269 | switch (c) |
| 270 | { |
| 271 | case '{': PushState(); break; |
| 272 | case '}': PopState(); break; |
| 273 | case '\\': |
| 274 | ParseKeyword(sResult, iPos); |
| 275 | break; |
| 276 | case 0x0d: case 0x0a: // ignored chars |
| 277 | break; |
| 278 | default: |
| 279 | // regular char parsing |
| 280 | if (pState->eState == psNormal) |
| 281 | // normal mode |
| 282 | ParseChar(sResult, c); |
| 283 | else if (pState->eState == psHex) |
| 284 | ParseHexChar(sResult, c); |
| 285 | else |
| 286 | throw ParserError("Invalid State" ); |
| 287 | break; |
| 288 | } |
| 289 | // next char |
| 290 | } |
| 291 | // all states must be closed in the end |
| 292 | if (pState->pNext) throw ParserError("Block not closed" ); |
| 293 | } |
| 294 | catch (const ParserError &pe) |
| 295 | { |
| 296 | // invalid RTF file: Display error message instead |
| 297 | sResult = "Invalid RTF file: " ; |
| 298 | sResult.Append(Buf2: pe.ErrorText); |
| 299 | } |
| 300 | // cleanup |
| 301 | ClearState(); |
| 302 | // return result |
| 303 | return sResult; |
| 304 | } |
| 305 | |
| 306 | #define kwdChars C4RTFFile::KeywordTableEntry::kwdChars |
| 307 | #define kwdSpec C4RTFFile::KeywordTableEntry::kwdSpec |
| 308 | #define kwdDest C4RTFFile::KeywordTableEntry::kwdDest |
| 309 | |
| 310 | // Keyword descriptions |
| 311 | C4RTFFile::KeywordTableEntry RTFKeywordTable[] = |
| 312 | { |
| 313 | // keyword iDefaultPar fForceDef eType idx |
| 314 | { .szKeyword: "par" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\n" , .idx: 0 }, |
| 315 | { .szKeyword: "\0x0a" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\n" , .idx: 0 }, |
| 316 | { .szKeyword: "\0x0d" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\n" , .idx: 0 }, |
| 317 | { .szKeyword: "tab" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\n" , .idx: 0 }, |
| 318 | { .szKeyword: "ldblquote" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\"" , .idx: 0 }, |
| 319 | { .szKeyword: "rdblquote" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\"" , .idx: 0 }, |
| 320 | { .szKeyword: "lquote" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "'" , .idx: 0 }, |
| 321 | { .szKeyword: "rquote" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "'" , .idx: 0 }, |
| 322 | { .szKeyword: "bin" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdSpec, .szChars: nullptr, .idx: C4RTFFile::specBin }, |
| 323 | { .szKeyword: "*" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdSpec, .szChars: nullptr, .idx: C4RTFFile::specSkipDest }, |
| 324 | { .szKeyword: "'" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdSpec, .szChars: nullptr, .idx: C4RTFFile::specHex }, |
| 325 | { .szKeyword: "author" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 326 | { .szKeyword: "buptim" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 327 | { .szKeyword: "colortbl" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 328 | { .szKeyword: "comment" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 329 | { .szKeyword: "creatim" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 330 | { .szKeyword: "doccomm" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 331 | { .szKeyword: "fonttbl" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 332 | { .szKeyword: "footer" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 333 | { .szKeyword: "footerf" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 334 | { .szKeyword: "footerl" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 335 | { .szKeyword: "footerr" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 336 | { .szKeyword: "footnote" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 337 | { .szKeyword: "ftncn" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 338 | { .szKeyword: "ftnsep" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 339 | { .szKeyword: "ftnsepc" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 340 | { .szKeyword: "header" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 341 | { .szKeyword: "headerf" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 342 | { .szKeyword: "headerl" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 343 | { .szKeyword: "headerr" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 344 | { .szKeyword: "info" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 345 | { .szKeyword: "keywords" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 346 | { .szKeyword: "operator" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 347 | { .szKeyword: "pict" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 348 | { .szKeyword: "printim" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 349 | { .szKeyword: "private1" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 350 | { .szKeyword: "revtim" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 351 | { .szKeyword: "rxe" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 352 | { .szKeyword: "stylesheet" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 353 | { .szKeyword: "subject" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 354 | { .szKeyword: "tc" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 355 | { .szKeyword: "title" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 356 | { .szKeyword: "txe" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 357 | { .szKeyword: "xe" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdDest, .szChars: nullptr, .idx: C4RTFFile::dsSkip }, |
| 358 | { .szKeyword: "{" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "{" , .idx: 0 }, |
| 359 | { .szKeyword: "}" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "}" , .idx: 0 }, |
| 360 | { .szKeyword: "\\" , .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: "\\" , .idx: 0 }, |
| 361 | { .szKeyword: nullptr, .iDefaultParam: 0, .fForceDefaultParam: false, kwdChars, .szChars: nullptr, .idx: 0 } |
| 362 | }; |
| 363 | |
| 364 | std::string RtfEscape(std::string_view plainText) |
| 365 | { |
| 366 | static constexpr std::string_view needEscapeChars{"\\{}" }; |
| 367 | constexpr auto needsEscape = [](char c) constexpr noexcept |
| 368 | { |
| 369 | return needEscapeChars.find(c: c) != std::string_view::npos; |
| 370 | }; |
| 371 | std::string result; |
| 372 | result.reserve(res: plainText.size() + std::count_if(first: plainText.begin(), last: plainText.end(), pred: needsEscape)); |
| 373 | |
| 374 | for (std::size_t last = 0; last < plainText.size(); ) |
| 375 | { |
| 376 | const auto next = plainText.find_first_of(str: needEscapeChars, pos: last); |
| 377 | result.append(svt: plainText.substr(pos: last, n: next - last)); |
| 378 | last = next; |
| 379 | if (next != std::string_view::npos) |
| 380 | { |
| 381 | result.push_back(c: '\\'); |
| 382 | result.push_back(c: plainText[next]); |
| 383 | ++last; |
| 384 | } |
| 385 | } |
| 386 | |
| 387 | return result; |
| 388 | } |
| 389 | |