1 //===--- StringRef.h - Constant String Reference Wrapper --------*- C++ -*-===// 2 // 3 // The LLVM Compiler Infrastructure 4 // 5 // This file is distributed under the University of Illinois Open Source 6 // License. See LICENSE.TXT for details. 7 // 8 //===----------------------------------------------------------------------===// 9 10 #ifndef LLVM_ADT_STRINGREF_H 11 #define LLVM_ADT_STRINGREF_H 12 13 #include "llvm/Support/type_traits.h" 14 #include <algorithm> 15 #include <cassert> 16 #include <cstring> 17 #include <limits> 18 #include <string> 19 #include <utility> 20 21 namespace llvm { 22 template<typename T> 23 class SmallVectorImpl; 24 class APInt; 25 class hash_code; 26 class StringRef; 27 28 /// Helper functions for StringRef::getAsInteger. 29 bool getAsUnsignedInteger(StringRef Str, unsigned Radix, 30 unsigned long long &Result); 31 32 bool getAsSignedInteger(StringRef Str, unsigned Radix, long long &Result); 33 34 /// StringRef - Represent a constant reference to a string, i.e. a character 35 /// array and a length, which need not be null terminated. 36 /// 37 /// This class does not own the string data, it is expected to be used in 38 /// situations where the character data resides in some other buffer, whose 39 /// lifetime extends past that of the StringRef. For this reason, it is not in 40 /// general safe to store a StringRef. 41 class StringRef { 42 public: 43 typedef const char *iterator; 44 typedef const char *const_iterator; 45 static const size_t npos = ~size_t(0); 46 typedef size_t size_type; 47 48 private: 49 /// The start of the string, in an external buffer. 50 const char *Data; 51 52 /// The length of the string. 53 size_t Length; 54 55 // Workaround PR5482: nearly all gcc 4.x miscompile StringRef and std::min() 56 // Changing the arg of min to be an integer, instead of a reference to an 57 // integer works around this bug. min(size_t a,size_t b)58 static size_t min(size_t a, size_t b) { return a < b ? a : b; } max(size_t a,size_t b)59 static size_t max(size_t a, size_t b) { return a > b ? a : b; } 60 61 // Workaround memcmp issue with null pointers (undefined behavior) 62 // by providing a specialized version compareMemory(const char * Lhs,const char * Rhs,size_t Length)63 static int compareMemory(const char *Lhs, const char *Rhs, size_t Length) { 64 if (Length == 0) { return 0; } 65 return ::memcmp(Lhs,Rhs,Length); 66 } 67 68 public: 69 /// @name Constructors 70 /// @{ 71 72 /// Construct an empty string ref. StringRef()73 /*implicit*/ StringRef() : Data(0), Length(0) {} 74 75 /// Construct a string ref from a cstring. StringRef(const char * Str)76 /*implicit*/ StringRef(const char *Str) 77 : Data(Str) { 78 assert(Str && "StringRef cannot be built from a NULL argument"); 79 Length = ::strlen(Str); // invoking strlen(NULL) is undefined behavior 80 } 81 82 /// Construct a string ref from a pointer and length. StringRef(const char * data,size_t length)83 /*implicit*/ StringRef(const char *data, size_t length) 84 : Data(data), Length(length) { 85 assert((data || length == 0) && 86 "StringRef cannot be built from a NULL argument with non-null length"); 87 } 88 89 /// Construct a string ref from an std::string. StringRef(const std::string & Str)90 /*implicit*/ StringRef(const std::string &Str) 91 : Data(Str.data()), Length(Str.length()) {} 92 93 /// @} 94 /// @name Iterators 95 /// @{ 96 begin()97 iterator begin() const { return Data; } 98 end()99 iterator end() const { return Data + Length; } 100 101 /// @} 102 /// @name String Operations 103 /// @{ 104 105 /// data - Get a pointer to the start of the string (which may not be null 106 /// terminated). data()107 const char *data() const { return Data; } 108 109 /// empty - Check if the string is empty. empty()110 bool empty() const { return Length == 0; } 111 112 /// size - Get the string size. size()113 size_t size() const { return Length; } 114 115 /// front - Get the first character in the string. front()116 char front() const { 117 assert(!empty()); 118 return Data[0]; 119 } 120 121 /// back - Get the last character in the string. back()122 char back() const { 123 assert(!empty()); 124 return Data[Length-1]; 125 } 126 127 /// equals - Check for string equality, this is more efficient than 128 /// compare() when the relative ordering of inequal strings isn't needed. equals(StringRef RHS)129 bool equals(StringRef RHS) const { 130 return (Length == RHS.Length && 131 compareMemory(Data, RHS.Data, RHS.Length) == 0); 132 } 133 134 /// equals_lower - Check for string equality, ignoring case. equals_lower(StringRef RHS)135 bool equals_lower(StringRef RHS) const { 136 return Length == RHS.Length && compare_lower(RHS) == 0; 137 } 138 139 /// compare - Compare two strings; the result is -1, 0, or 1 if this string 140 /// is lexicographically less than, equal to, or greater than the \p RHS. compare(StringRef RHS)141 int compare(StringRef RHS) const { 142 // Check the prefix for a mismatch. 143 if (int Res = compareMemory(Data, RHS.Data, min(Length, RHS.Length))) 144 return Res < 0 ? -1 : 1; 145 146 // Otherwise the prefixes match, so we only need to check the lengths. 147 if (Length == RHS.Length) 148 return 0; 149 return Length < RHS.Length ? -1 : 1; 150 } 151 152 /// compare_lower - Compare two strings, ignoring case. 153 int compare_lower(StringRef RHS) const; 154 155 /// compare_numeric - Compare two strings, treating sequences of digits as 156 /// numbers. 157 int compare_numeric(StringRef RHS) const; 158 159 /// \brief Determine the edit distance between this string and another 160 /// string. 161 /// 162 /// \param Other the string to compare this string against. 163 /// 164 /// \param AllowReplacements whether to allow character 165 /// replacements (change one character into another) as a single 166 /// operation, rather than as two operations (an insertion and a 167 /// removal). 168 /// 169 /// \param MaxEditDistance If non-zero, the maximum edit distance that 170 /// this routine is allowed to compute. If the edit distance will exceed 171 /// that maximum, returns \c MaxEditDistance+1. 172 /// 173 /// \returns the minimum number of character insertions, removals, 174 /// or (if \p AllowReplacements is \c true) replacements needed to 175 /// transform one of the given strings into the other. If zero, 176 /// the strings are identical. 177 unsigned edit_distance(StringRef Other, bool AllowReplacements = true, 178 unsigned MaxEditDistance = 0); 179 180 /// str - Get the contents as an std::string. str()181 std::string str() const { 182 if (Data == 0) return std::string(); 183 return std::string(Data, Length); 184 } 185 186 /// @} 187 /// @name Operator Overloads 188 /// @{ 189 190 char operator[](size_t Index) const { 191 assert(Index < Length && "Invalid index!"); 192 return Data[Index]; 193 } 194 195 /// @} 196 /// @name Type Conversions 197 /// @{ 198 string()199 operator std::string() const { 200 return str(); 201 } 202 203 /// @} 204 /// @name String Predicates 205 /// @{ 206 207 /// Check if this string starts with the given \p Prefix. startswith(StringRef Prefix)208 bool startswith(StringRef Prefix) const { 209 return Length >= Prefix.Length && 210 compareMemory(Data, Prefix.Data, Prefix.Length) == 0; 211 } 212 213 /// Check if this string ends with the given \p Suffix. endswith(StringRef Suffix)214 bool endswith(StringRef Suffix) const { 215 return Length >= Suffix.Length && 216 compareMemory(end() - Suffix.Length, Suffix.Data, Suffix.Length) == 0; 217 } 218 219 /// @} 220 /// @name String Searching 221 /// @{ 222 223 /// Search for the first character \p C in the string. 224 /// 225 /// \returns The index of the first occurrence of \p C, or npos if not 226 /// found. 227 size_t find(char C, size_t From = 0) const { 228 for (size_t i = min(From, Length), e = Length; i != e; ++i) 229 if (Data[i] == C) 230 return i; 231 return npos; 232 } 233 234 /// Search for the first string \p Str in the string. 235 /// 236 /// \returns The index of the first occurrence of \p Str, or npos if not 237 /// found. 238 size_t find(StringRef Str, size_t From = 0) const; 239 240 /// Search for the last character \p C in the string. 241 /// 242 /// \returns The index of the last occurrence of \p C, or npos if not 243 /// found. 244 size_t rfind(char C, size_t From = npos) const { 245 From = min(From, Length); 246 size_t i = From; 247 while (i != 0) { 248 --i; 249 if (Data[i] == C) 250 return i; 251 } 252 return npos; 253 } 254 255 /// Search for the last string \p Str in the string. 256 /// 257 /// \returns The index of the last occurrence of \p Str, or npos if not 258 /// found. 259 size_t rfind(StringRef Str) const; 260 261 /// Find the first character in the string that is \p C, or npos if not 262 /// found. Same as find. 263 size_type find_first_of(char C, size_t From = 0) const { 264 return find(C, From); 265 } 266 267 /// Find the first character in the string that is in \p Chars, or npos if 268 /// not found. 269 /// 270 /// Complexity: O(size() + Chars.size()) 271 size_type find_first_of(StringRef Chars, size_t From = 0) const; 272 273 /// Find the first character in the string that is not \p C or npos if not 274 /// found. 275 size_type find_first_not_of(char C, size_t From = 0) const; 276 277 /// Find the first character in the string that is not in the string 278 /// \p Chars, or npos if not found. 279 /// 280 /// Complexity: O(size() + Chars.size()) 281 size_type find_first_not_of(StringRef Chars, size_t From = 0) const; 282 283 /// Find the last character in the string that is \p C, or npos if not 284 /// found. 285 size_type find_last_of(char C, size_t From = npos) const { 286 return rfind(C, From); 287 } 288 289 /// Find the last character in the string that is in \p C, or npos if not 290 /// found. 291 /// 292 /// Complexity: O(size() + Chars.size()) 293 size_type find_last_of(StringRef Chars, size_t From = npos) const; 294 295 /// Find the last character in the string that is not \p C, or npos if not 296 /// found. 297 size_type find_last_not_of(char C, size_t From = npos) const; 298 299 /// Find the last character in the string that is not in \p Chars, or 300 /// npos if not found. 301 /// 302 /// Complexity: O(size() + Chars.size()) 303 size_type find_last_not_of(StringRef Chars, size_t From = npos) const; 304 305 /// @} 306 /// @name Helpful Algorithms 307 /// @{ 308 309 /// Return the number of occurrences of \p C in the string. count(char C)310 size_t count(char C) const { 311 size_t Count = 0; 312 for (size_t i = 0, e = Length; i != e; ++i) 313 if (Data[i] == C) 314 ++Count; 315 return Count; 316 } 317 318 /// Return the number of non-overlapped occurrences of \p Str in 319 /// the string. 320 size_t count(StringRef Str) const; 321 322 /// Parse the current string as an integer of the specified radix. If 323 /// \p Radix is specified as zero, this does radix autosensing using 324 /// extended C rules: 0 is octal, 0x is hex, 0b is binary. 325 /// 326 /// If the string is invalid or if only a subset of the string is valid, 327 /// this returns true to signify the error. The string is considered 328 /// erroneous if empty or if it overflows T. 329 template <typename T> 330 typename enable_if_c<std::numeric_limits<T>::is_signed, bool>::type getAsInteger(unsigned Radix,T & Result)331 getAsInteger(unsigned Radix, T &Result) const { 332 long long LLVal; 333 if (getAsSignedInteger(*this, Radix, LLVal) || 334 static_cast<T>(LLVal) != LLVal) 335 return true; 336 Result = LLVal; 337 return false; 338 } 339 340 template <typename T> 341 typename enable_if_c<!std::numeric_limits<T>::is_signed, bool>::type getAsInteger(unsigned Radix,T & Result)342 getAsInteger(unsigned Radix, T &Result) const { 343 unsigned long long ULLVal; 344 if (getAsUnsignedInteger(*this, Radix, ULLVal) || 345 static_cast<T>(ULLVal) != ULLVal) 346 return true; 347 Result = ULLVal; 348 return false; 349 } 350 351 /// Parse the current string as an integer of the specified \p Radix, or of 352 /// an autosensed radix if the \p Radix given is 0. The current value in 353 /// \p Result is discarded, and the storage is changed to be wide enough to 354 /// store the parsed integer. 355 /// 356 /// \returns true if the string does not solely consist of a valid 357 /// non-empty number in the appropriate base. 358 /// 359 /// APInt::fromString is superficially similar but assumes the 360 /// string is well-formed in the given radix. 361 bool getAsInteger(unsigned Radix, APInt &Result) const; 362 363 /// @} 364 /// @name String Operations 365 /// @{ 366 367 // Convert the given ASCII string to lowercase. 368 std::string lower() const; 369 370 /// Convert the given ASCII string to uppercase. 371 std::string upper() const; 372 373 /// @} 374 /// @name Substring Operations 375 /// @{ 376 377 /// Return a reference to the substring from [Start, Start + N). 378 /// 379 /// \param Start The index of the starting character in the substring; if 380 /// the index is npos or greater than the length of the string then the 381 /// empty substring will be returned. 382 /// 383 /// \param N The number of characters to included in the substring. If N 384 /// exceeds the number of characters remaining in the string, the string 385 /// suffix (starting with \p Start) will be returned. 386 StringRef substr(size_t Start, size_t N = npos) const { 387 Start = min(Start, Length); 388 return StringRef(Data + Start, min(N, Length - Start)); 389 } 390 391 /// Return a StringRef equal to 'this' but with the first \p N elements 392 /// dropped. 393 StringRef drop_front(unsigned N = 1) const { 394 assert(size() >= N && "Dropping more elements than exist"); 395 return substr(N); 396 } 397 398 /// Return a StringRef equal to 'this' but with the last \p N elements 399 /// dropped. 400 StringRef drop_back(unsigned N = 1) const { 401 assert(size() >= N && "Dropping more elements than exist"); 402 return substr(0, size()-N); 403 } 404 405 /// Return a reference to the substring from [Start, End). 406 /// 407 /// \param Start The index of the starting character in the substring; if 408 /// the index is npos or greater than the length of the string then the 409 /// empty substring will be returned. 410 /// 411 /// \param End The index following the last character to include in the 412 /// substring. If this is npos, or less than \p Start, or exceeds the 413 /// number of characters remaining in the string, the string suffix 414 /// (starting with \p Start) will be returned. slice(size_t Start,size_t End)415 StringRef slice(size_t Start, size_t End) const { 416 Start = min(Start, Length); 417 End = min(max(Start, End), Length); 418 return StringRef(Data + Start, End - Start); 419 } 420 421 /// Split into two substrings around the first occurrence of a separator 422 /// character. 423 /// 424 /// If \p Separator is in the string, then the result is a pair (LHS, RHS) 425 /// such that (*this == LHS + Separator + RHS) is true and RHS is 426 /// maximal. If \p Separator is not in the string, then the result is a 427 /// pair (LHS, RHS) where (*this == LHS) and (RHS == ""). 428 /// 429 /// \param Separator The character to split on. 430 /// \returns The split substrings. split(char Separator)431 std::pair<StringRef, StringRef> split(char Separator) const { 432 size_t Idx = find(Separator); 433 if (Idx == npos) 434 return std::make_pair(*this, StringRef()); 435 return std::make_pair(slice(0, Idx), slice(Idx+1, npos)); 436 } 437 438 /// Split into two substrings around the first occurrence of a separator 439 /// string. 440 /// 441 /// If \p Separator is in the string, then the result is a pair (LHS, RHS) 442 /// such that (*this == LHS + Separator + RHS) is true and RHS is 443 /// maximal. If \p Separator is not in the string, then the result is a 444 /// pair (LHS, RHS) where (*this == LHS) and (RHS == ""). 445 /// 446 /// \param Separator - The string to split on. 447 /// \return - The split substrings. split(StringRef Separator)448 std::pair<StringRef, StringRef> split(StringRef Separator) const { 449 size_t Idx = find(Separator); 450 if (Idx == npos) 451 return std::make_pair(*this, StringRef()); 452 return std::make_pair(slice(0, Idx), slice(Idx + Separator.size(), npos)); 453 } 454 455 /// Split into substrings around the occurrences of a separator string. 456 /// 457 /// Each substring is stored in \p A. If \p MaxSplit is >= 0, at most 458 /// \p MaxSplit splits are done and consequently <= \p MaxSplit 459 /// elements are added to A. 460 /// If \p KeepEmpty is false, empty strings are not added to \p A. They 461 /// still count when considering \p MaxSplit 462 /// An useful invariant is that 463 /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true 464 /// 465 /// \param A - Where to put the substrings. 466 /// \param Separator - The string to split on. 467 /// \param MaxSplit - The maximum number of times the string is split. 468 /// \param KeepEmpty - True if empty substring should be added. 469 void split(SmallVectorImpl<StringRef> &A, 470 StringRef Separator, int MaxSplit = -1, 471 bool KeepEmpty = true) const; 472 473 /// Split into two substrings around the last occurrence of a separator 474 /// character. 475 /// 476 /// If \p Separator is in the string, then the result is a pair (LHS, RHS) 477 /// such that (*this == LHS + Separator + RHS) is true and RHS is 478 /// minimal. If \p Separator is not in the string, then the result is a 479 /// pair (LHS, RHS) where (*this == LHS) and (RHS == ""). 480 /// 481 /// \param Separator - The character to split on. 482 /// \return - The split substrings. rsplit(char Separator)483 std::pair<StringRef, StringRef> rsplit(char Separator) const { 484 size_t Idx = rfind(Separator); 485 if (Idx == npos) 486 return std::make_pair(*this, StringRef()); 487 return std::make_pair(slice(0, Idx), slice(Idx+1, npos)); 488 } 489 490 /// Return string with consecutive characters in \p Chars starting from 491 /// the left removed. 492 StringRef ltrim(StringRef Chars = " \t\n\v\f\r") const { 493 return drop_front(std::min(Length, find_first_not_of(Chars))); 494 } 495 496 /// Return string with consecutive characters in \p Chars starting from 497 /// the right removed. 498 StringRef rtrim(StringRef Chars = " \t\n\v\f\r") const { 499 return drop_back(Length - std::min(Length, find_last_not_of(Chars) + 1)); 500 } 501 502 /// Return string with consecutive characters in \p Chars starting from 503 /// the left and right removed. 504 StringRef trim(StringRef Chars = " \t\n\v\f\r") const { 505 return ltrim(Chars).rtrim(Chars); 506 } 507 508 /// @} 509 }; 510 511 /// @name StringRef Comparison Operators 512 /// @{ 513 514 inline bool operator==(StringRef LHS, StringRef RHS) { 515 return LHS.equals(RHS); 516 } 517 518 inline bool operator!=(StringRef LHS, StringRef RHS) { 519 return !(LHS == RHS); 520 } 521 522 inline bool operator<(StringRef LHS, StringRef RHS) { 523 return LHS.compare(RHS) == -1; 524 } 525 526 inline bool operator<=(StringRef LHS, StringRef RHS) { 527 return LHS.compare(RHS) != 1; 528 } 529 530 inline bool operator>(StringRef LHS, StringRef RHS) { 531 return LHS.compare(RHS) == 1; 532 } 533 534 inline bool operator>=(StringRef LHS, StringRef RHS) { 535 return LHS.compare(RHS) != -1; 536 } 537 538 inline std::string &operator+=(std::string &buffer, StringRef string) { 539 return buffer.append(string.data(), string.size()); 540 } 541 542 /// @} 543 544 /// \brief Compute a hash_code for a StringRef. 545 hash_code hash_value(StringRef S); 546 547 // StringRefs can be treated like a POD type. 548 template <typename T> struct isPodLike; 549 template <> struct isPodLike<StringRef> { static const bool value = true; }; 550 551 } 552 553 #endif 554