src/llvm/include/llvm/ADT/StringRef.h

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