1 //===- llvm/ADT/StringExtras.h - Useful string functions --------*- C++ -*-===// 2 // 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 4 // See https://llvm.org/LICENSE.txt for license information. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 6 // 7 //===----------------------------------------------------------------------===// 8 /// 9 /// \file 10 /// This file contains some functions that are useful when dealing with strings. 11 /// 12 //===----------------------------------------------------------------------===// 13 14 #ifndef LLVM_ADT_STRINGEXTRAS_H 15 #define LLVM_ADT_STRINGEXTRAS_H 16 17 #include "llvm/ADT/APSInt.h" 18 #include "llvm/ADT/ArrayRef.h" 19 #include "llvm/ADT/SmallString.h" 20 #include "llvm/ADT/StringRef.h" 21 #include "llvm/ADT/Twine.h" 22 #include <cassert> 23 #include <cstddef> 24 #include <cstdint> 25 #include <cstdlib> 26 #include <cstring> 27 #include <iterator> 28 #include <string> 29 #include <utility> 30 31 namespace llvm { 32 33 class raw_ostream; 34 35 /// hexdigit - Return the hexadecimal character for the 36 /// given number \p X (which should be less than 16). 37 inline char hexdigit(unsigned X, bool LowerCase = false) { 38 assert(X < 16); 39 static const char LUT[] = "0123456789ABCDEF"; 40 const uint8_t Offset = LowerCase ? 32 : 0; 41 return LUT[X] | Offset; 42 } 43 44 /// Given an array of c-style strings terminated by a null pointer, construct 45 /// a vector of StringRefs representing the same strings without the terminating 46 /// null string. 47 inline std::vector<StringRef> toStringRefArray(const char *const *Strings) { 48 std::vector<StringRef> Result; 49 while (*Strings) 50 Result.push_back(*Strings++); 51 return Result; 52 } 53 54 /// Construct a string ref from a boolean. 55 inline StringRef toStringRef(bool B) { return StringRef(B ? "true" : "false"); } 56 57 /// Construct a string ref from an array ref of unsigned chars. 58 inline StringRef toStringRef(ArrayRef<uint8_t> Input) { 59 return StringRef(reinterpret_cast<const char *>(Input.begin()), Input.size()); 60 } 61 62 /// Construct a string ref from an array ref of unsigned chars. 63 inline ArrayRef<uint8_t> arrayRefFromStringRef(StringRef Input) { 64 return {Input.bytes_begin(), Input.bytes_end()}; 65 } 66 67 /// Interpret the given character \p C as a hexadecimal digit and return its 68 /// value. 69 /// 70 /// If \p C is not a valid hex digit, -1U is returned. 71 inline unsigned hexDigitValue(char C) { 72 /* clang-format off */ 73 static const int16_t LUT[256] = { 74 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 75 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 76 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 77 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, -1, -1, -1, -1, -1, // '0'..'9' 78 -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 'A'..'F' 79 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 80 -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 'a'..'f' 81 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 82 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 83 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 84 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 85 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 86 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 87 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 88 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 89 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 90 }; 91 /* clang-format on */ 92 return LUT[static_cast<unsigned char>(C)]; 93 } 94 95 /// Checks if character \p C is one of the 10 decimal digits. 96 inline bool isDigit(char C) { return C >= '0' && C <= '9'; } 97 98 /// Checks if character \p C is a hexadecimal numeric character. 99 inline bool isHexDigit(char C) { return hexDigitValue(C) != ~0U; } 100 101 /// Checks if character \p C is a valid letter as classified by "C" locale. 102 inline bool isAlpha(char C) { 103 return ('a' <= C && C <= 'z') || ('A' <= C && C <= 'Z'); 104 } 105 106 /// Checks whether character \p C is either a decimal digit or an uppercase or 107 /// lowercase letter as classified by "C" locale. 108 inline bool isAlnum(char C) { return isAlpha(C) || isDigit(C); } 109 110 /// Checks whether character \p C is valid ASCII (high bit is zero). 111 inline bool isASCII(char C) { return static_cast<unsigned char>(C) <= 127; } 112 113 /// Checks whether all characters in S are ASCII. 114 inline bool isASCII(llvm::StringRef S) { 115 for (char C : S) 116 if (LLVM_UNLIKELY(!isASCII(C))) 117 return false; 118 return true; 119 } 120 121 /// Checks whether character \p C is printable. 122 /// 123 /// Locale-independent version of the C standard library isprint whose results 124 /// may differ on different platforms. 125 inline bool isPrint(char C) { 126 unsigned char UC = static_cast<unsigned char>(C); 127 return (0x20 <= UC) && (UC <= 0x7E); 128 } 129 130 /// Checks whether character \p C is whitespace in the "C" locale. 131 /// 132 /// Locale-independent version of the C standard library isspace. 133 inline bool isSpace(char C) { 134 return C == ' ' || C == '\f' || C == '\n' || C == '\r' || C == '\t' || 135 C == '\v'; 136 } 137 138 /// Returns the corresponding lowercase character if \p x is uppercase. 139 inline char toLower(char x) { 140 if (x >= 'A' && x <= 'Z') 141 return x - 'A' + 'a'; 142 return x; 143 } 144 145 /// Returns the corresponding uppercase character if \p x is lowercase. 146 inline char toUpper(char x) { 147 if (x >= 'a' && x <= 'z') 148 return x - 'a' + 'A'; 149 return x; 150 } 151 152 inline std::string utohexstr(uint64_t X, bool LowerCase = false, 153 unsigned Width = 0) { 154 char Buffer[17]; 155 char *BufPtr = std::end(Buffer); 156 157 if (X == 0) *--BufPtr = '0'; 158 159 for (unsigned i = 0; Width ? (i < Width) : X; ++i) { 160 unsigned char Mod = static_cast<unsigned char>(X) & 15; 161 *--BufPtr = hexdigit(Mod, LowerCase); 162 X >>= 4; 163 } 164 165 return std::string(BufPtr, std::end(Buffer)); 166 } 167 168 /// Convert buffer \p Input to its hexadecimal representation. 169 /// The returned string is double the size of \p Input. 170 inline void toHex(ArrayRef<uint8_t> Input, bool LowerCase, 171 SmallVectorImpl<char> &Output) { 172 const size_t Length = Input.size(); 173 Output.resize_for_overwrite(Length * 2); 174 175 for (size_t i = 0; i < Length; i++) { 176 const uint8_t c = Input[i]; 177 Output[i * 2 ] = hexdigit(c >> 4, LowerCase); 178 Output[i * 2 + 1] = hexdigit(c & 15, LowerCase); 179 } 180 } 181 182 inline std::string toHex(ArrayRef<uint8_t> Input, bool LowerCase = false) { 183 SmallString<16> Output; 184 toHex(Input, LowerCase, Output); 185 return std::string(Output); 186 } 187 188 inline std::string toHex(StringRef Input, bool LowerCase = false) { 189 return toHex(arrayRefFromStringRef(Input), LowerCase); 190 } 191 192 /// Store the binary representation of the two provided values, \p MSB and 193 /// \p LSB, that make up the nibbles of a hexadecimal digit. If \p MSB or \p LSB 194 /// do not correspond to proper nibbles of a hexadecimal digit, this method 195 /// returns false. Otherwise, returns true. 196 inline bool tryGetHexFromNibbles(char MSB, char LSB, uint8_t &Hex) { 197 unsigned U1 = hexDigitValue(MSB); 198 unsigned U2 = hexDigitValue(LSB); 199 if (U1 == ~0U || U2 == ~0U) 200 return false; 201 202 Hex = static_cast<uint8_t>((U1 << 4) | U2); 203 return true; 204 } 205 206 /// Return the binary representation of the two provided values, \p MSB and 207 /// \p LSB, that make up the nibbles of a hexadecimal digit. 208 inline uint8_t hexFromNibbles(char MSB, char LSB) { 209 uint8_t Hex = 0; 210 bool GotHex = tryGetHexFromNibbles(MSB, LSB, Hex); 211 (void)GotHex; 212 assert(GotHex && "MSB and/or LSB do not correspond to hex digits"); 213 return Hex; 214 } 215 216 /// Convert hexadecimal string \p Input to its binary representation and store 217 /// the result in \p Output. Returns true if the binary representation could be 218 /// converted from the hexadecimal string. Returns false if \p Input contains 219 /// non-hexadecimal digits. The output string is half the size of \p Input. 220 inline bool tryGetFromHex(StringRef Input, std::string &Output) { 221 if (Input.empty()) 222 return true; 223 224 // If the input string is not properly aligned on 2 nibbles we pad out the 225 // front with a 0 prefix; e.g. `ABC` -> `0ABC`. 226 Output.resize((Input.size() + 1) / 2); 227 char *OutputPtr = const_cast<char *>(Output.data()); 228 if (Input.size() % 2 == 1) { 229 uint8_t Hex = 0; 230 if (!tryGetHexFromNibbles('0', Input.front(), Hex)) 231 return false; 232 *OutputPtr++ = Hex; 233 Input = Input.drop_front(); 234 } 235 236 // Convert the nibble pairs (e.g. `9C`) into bytes (0x9C). 237 // With the padding above we know the input is aligned and the output expects 238 // exactly half as many bytes as nibbles in the input. 239 size_t InputSize = Input.size(); 240 assert(InputSize % 2 == 0); 241 const char *InputPtr = Input.data(); 242 for (size_t OutputIndex = 0; OutputIndex < InputSize / 2; ++OutputIndex) { 243 uint8_t Hex = 0; 244 if (!tryGetHexFromNibbles(InputPtr[OutputIndex * 2 + 0], // MSB 245 InputPtr[OutputIndex * 2 + 1], // LSB 246 Hex)) 247 return false; 248 OutputPtr[OutputIndex] = Hex; 249 } 250 return true; 251 } 252 253 /// Convert hexadecimal string \p Input to its binary representation. 254 /// The return string is half the size of \p Input. 255 inline std::string fromHex(StringRef Input) { 256 std::string Hex; 257 bool GotHex = tryGetFromHex(Input, Hex); 258 (void)GotHex; 259 assert(GotHex && "Input contains non hex digits"); 260 return Hex; 261 } 262 263 /// Convert the string \p S to an integer of the specified type using 264 /// the radix \p Base. If \p Base is 0, auto-detects the radix. 265 /// Returns true if the number was successfully converted, false otherwise. 266 template <typename N> bool to_integer(StringRef S, N &Num, unsigned Base = 0) { 267 return !S.getAsInteger(Base, Num); 268 } 269 270 namespace detail { 271 template <typename N> 272 inline bool to_float(const Twine &T, N &Num, N (*StrTo)(const char *, char **)) { 273 SmallString<32> Storage; 274 StringRef S = T.toNullTerminatedStringRef(Storage); 275 char *End; 276 N Temp = StrTo(S.data(), &End); 277 if (*End != '\0') 278 return false; 279 Num = Temp; 280 return true; 281 } 282 } 283 284 inline bool to_float(const Twine &T, float &Num) { 285 return detail::to_float(T, Num, strtof); 286 } 287 288 inline bool to_float(const Twine &T, double &Num) { 289 return detail::to_float(T, Num, strtod); 290 } 291 292 inline bool to_float(const Twine &T, long double &Num) { 293 return detail::to_float(T, Num, strtold); 294 } 295 296 inline std::string utostr(uint64_t X, bool isNeg = false) { 297 char Buffer[21]; 298 char *BufPtr = std::end(Buffer); 299 300 if (X == 0) *--BufPtr = '0'; // Handle special case... 301 302 while (X) { 303 *--BufPtr = '0' + char(X % 10); 304 X /= 10; 305 } 306 307 if (isNeg) *--BufPtr = '-'; // Add negative sign... 308 return std::string(BufPtr, std::end(Buffer)); 309 } 310 311 inline std::string itostr(int64_t X) { 312 if (X < 0) 313 return utostr(static_cast<uint64_t>(1) + ~static_cast<uint64_t>(X), true); 314 else 315 return utostr(static_cast<uint64_t>(X)); 316 } 317 318 inline std::string toString(const APInt &I, unsigned Radix, bool Signed, 319 bool formatAsCLiteral = false) { 320 SmallString<40> S; 321 I.toString(S, Radix, Signed, formatAsCLiteral); 322 return std::string(S.str()); 323 } 324 325 inline std::string toString(const APSInt &I, unsigned Radix) { 326 return toString(I, Radix, I.isSigned()); 327 } 328 329 /// StrInStrNoCase - Portable version of strcasestr. Locates the first 330 /// occurrence of string 's1' in string 's2', ignoring case. Returns 331 /// the offset of s2 in s1 or npos if s2 cannot be found. 332 StringRef::size_type StrInStrNoCase(StringRef s1, StringRef s2); 333 334 /// getToken - This function extracts one token from source, ignoring any 335 /// leading characters that appear in the Delimiters string, and ending the 336 /// token at any of the characters that appear in the Delimiters string. If 337 /// there are no tokens in the source string, an empty string is returned. 338 /// The function returns a pair containing the extracted token and the 339 /// remaining tail string. 340 std::pair<StringRef, StringRef> getToken(StringRef Source, 341 StringRef Delimiters = " \t\n\v\f\r"); 342 343 /// SplitString - Split up the specified string according to the specified 344 /// delimiters, appending the result fragments to the output list. 345 void SplitString(StringRef Source, 346 SmallVectorImpl<StringRef> &OutFragments, 347 StringRef Delimiters = " \t\n\v\f\r"); 348 349 /// Returns the English suffix for an ordinal integer (-st, -nd, -rd, -th). 350 inline StringRef getOrdinalSuffix(unsigned Val) { 351 // It is critically important that we do this perfectly for 352 // user-written sequences with over 100 elements. 353 switch (Val % 100) { 354 case 11: 355 case 12: 356 case 13: 357 return "th"; 358 default: 359 switch (Val % 10) { 360 case 1: return "st"; 361 case 2: return "nd"; 362 case 3: return "rd"; 363 default: return "th"; 364 } 365 } 366 } 367 368 /// Print each character of the specified string, escaping it if it is not 369 /// printable or if it is an escape char. 370 void printEscapedString(StringRef Name, raw_ostream &Out); 371 372 /// Print each character of the specified string, escaping HTML special 373 /// characters. 374 void printHTMLEscaped(StringRef String, raw_ostream &Out); 375 376 /// printLowerCase - Print each character as lowercase if it is uppercase. 377 void printLowerCase(StringRef String, raw_ostream &Out); 378 379 /// Converts a string from camel-case to snake-case by replacing all uppercase 380 /// letters with '_' followed by the letter in lowercase, except if the 381 /// uppercase letter is the first character of the string. 382 std::string convertToSnakeFromCamelCase(StringRef input); 383 384 /// Converts a string from snake-case to camel-case by replacing all occurrences 385 /// of '_' followed by a lowercase letter with the letter in uppercase. 386 /// Optionally allow capitalization of the first letter (if it is a lowercase 387 /// letter) 388 std::string convertToCamelFromSnakeCase(StringRef input, 389 bool capitalizeFirst = false); 390 391 namespace detail { 392 393 template <typename IteratorT> 394 inline std::string join_impl(IteratorT Begin, IteratorT End, 395 StringRef Separator, std::input_iterator_tag) { 396 std::string S; 397 if (Begin == End) 398 return S; 399 400 S += (*Begin); 401 while (++Begin != End) { 402 S += Separator; 403 S += (*Begin); 404 } 405 return S; 406 } 407 408 template <typename IteratorT> 409 inline std::string join_impl(IteratorT Begin, IteratorT End, 410 StringRef Separator, std::forward_iterator_tag) { 411 std::string S; 412 if (Begin == End) 413 return S; 414 415 size_t Len = (std::distance(Begin, End) - 1) * Separator.size(); 416 for (IteratorT I = Begin; I != End; ++I) 417 Len += (*I).size(); 418 S.reserve(Len); 419 size_t PrevCapacity = S.capacity(); 420 (void)PrevCapacity; 421 S += (*Begin); 422 while (++Begin != End) { 423 S += Separator; 424 S += (*Begin); 425 } 426 assert(PrevCapacity == S.capacity() && "String grew during building"); 427 return S; 428 } 429 430 template <typename Sep> 431 inline void join_items_impl(std::string &Result, Sep Separator) {} 432 433 template <typename Sep, typename Arg> 434 inline void join_items_impl(std::string &Result, Sep Separator, 435 const Arg &Item) { 436 Result += Item; 437 } 438 439 template <typename Sep, typename Arg1, typename... Args> 440 inline void join_items_impl(std::string &Result, Sep Separator, const Arg1 &A1, 441 Args &&... Items) { 442 Result += A1; 443 Result += Separator; 444 join_items_impl(Result, Separator, std::forward<Args>(Items)...); 445 } 446 447 inline size_t join_one_item_size(char) { return 1; } 448 inline size_t join_one_item_size(const char *S) { return S ? ::strlen(S) : 0; } 449 450 template <typename T> inline size_t join_one_item_size(const T &Str) { 451 return Str.size(); 452 } 453 454 inline size_t join_items_size() { return 0; } 455 456 template <typename A1> inline size_t join_items_size(const A1 &A) { 457 return join_one_item_size(A); 458 } 459 template <typename A1, typename... Args> 460 inline size_t join_items_size(const A1 &A, Args &&... Items) { 461 return join_one_item_size(A) + join_items_size(std::forward<Args>(Items)...); 462 } 463 464 } // end namespace detail 465 466 /// Joins the strings in the range [Begin, End), adding Separator between 467 /// the elements. 468 template <typename IteratorT> 469 inline std::string join(IteratorT Begin, IteratorT End, StringRef Separator) { 470 using tag = typename std::iterator_traits<IteratorT>::iterator_category; 471 return detail::join_impl(Begin, End, Separator, tag()); 472 } 473 474 /// Joins the strings in the range [R.begin(), R.end()), adding Separator 475 /// between the elements. 476 template <typename Range> 477 inline std::string join(Range &&R, StringRef Separator) { 478 return join(R.begin(), R.end(), Separator); 479 } 480 481 /// Joins the strings in the parameter pack \p Items, adding \p Separator 482 /// between the elements. All arguments must be implicitly convertible to 483 /// std::string, or there should be an overload of std::string::operator+=() 484 /// that accepts the argument explicitly. 485 template <typename Sep, typename... Args> 486 inline std::string join_items(Sep Separator, Args &&... Items) { 487 std::string Result; 488 if (sizeof...(Items) == 0) 489 return Result; 490 491 size_t NS = detail::join_one_item_size(Separator); 492 size_t NI = detail::join_items_size(std::forward<Args>(Items)...); 493 Result.reserve(NI + (sizeof...(Items) - 1) * NS + 1); 494 detail::join_items_impl(Result, Separator, std::forward<Args>(Items)...); 495 return Result; 496 } 497 498 /// A helper class to return the specified delimiter string after the first 499 /// invocation of operator StringRef(). Used to generate a comma-separated 500 /// list from a loop like so: 501 /// 502 /// \code 503 /// ListSeparator LS; 504 /// for (auto &I : C) 505 /// OS << LS << I.getName(); 506 /// \end 507 class ListSeparator { 508 bool First = true; 509 StringRef Separator; 510 511 public: 512 ListSeparator(StringRef Separator = ", ") : Separator(Separator) {} 513 operator StringRef() { 514 if (First) { 515 First = false; 516 return {}; 517 } 518 return Separator; 519 } 520 }; 521 522 /// A forward iterator over partitions of string over a separator. 523 class SplittingIterator 524 : public iterator_facade_base<SplittingIterator, std::forward_iterator_tag, 525 StringRef> { 526 char SeparatorStorage; 527 StringRef Current; 528 StringRef Next; 529 StringRef Separator; 530 531 public: 532 SplittingIterator(StringRef Str, StringRef Separator) 533 : Next(Str), Separator(Separator) { 534 ++*this; 535 } 536 537 SplittingIterator(StringRef Str, char Separator) 538 : SeparatorStorage(Separator), Next(Str), 539 Separator(&SeparatorStorage, 1) { 540 ++*this; 541 } 542 543 SplittingIterator(const SplittingIterator &R) 544 : SeparatorStorage(R.SeparatorStorage), Current(R.Current), Next(R.Next), 545 Separator(R.Separator) { 546 if (R.Separator.data() == &R.SeparatorStorage) 547 Separator = StringRef(&SeparatorStorage, 1); 548 } 549 550 SplittingIterator &operator=(const SplittingIterator &R) { 551 if (this == &R) 552 return *this; 553 554 SeparatorStorage = R.SeparatorStorage; 555 Current = R.Current; 556 Next = R.Next; 557 Separator = R.Separator; 558 if (R.Separator.data() == &R.SeparatorStorage) 559 Separator = StringRef(&SeparatorStorage, 1); 560 return *this; 561 } 562 563 bool operator==(const SplittingIterator &R) const { 564 assert(Separator == R.Separator); 565 return Current.data() == R.Current.data(); 566 } 567 568 const StringRef &operator*() const { return Current; } 569 570 StringRef &operator*() { return Current; } 571 572 SplittingIterator &operator++() { 573 std::tie(Current, Next) = Next.split(Separator); 574 return *this; 575 } 576 }; 577 578 /// Split the specified string over a separator and return a range-compatible 579 /// iterable over its partitions. Used to permit conveniently iterating 580 /// over separated strings like so: 581 /// 582 /// \code 583 /// for (StringRef x : llvm::split("foo,bar,baz", ",")) 584 /// ...; 585 /// \end 586 /// 587 /// Note that the passed string must remain valid throuhgout lifetime 588 /// of the iterators. 589 inline iterator_range<SplittingIterator> split(StringRef Str, StringRef Separator) { 590 return {SplittingIterator(Str, Separator), 591 SplittingIterator(StringRef(), Separator)}; 592 } 593 594 inline iterator_range<SplittingIterator> split(StringRef Str, char Separator) { 595 return {SplittingIterator(Str, Separator), 596 SplittingIterator(StringRef(), Separator)}; 597 } 598 599 } // end namespace llvm 600 601 #endif // LLVM_ADT_STRINGEXTRAS_H 602