1 //===--- StringRef.h - Constant String Reference Wrapper --------*- C++ -*-===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 #ifndef LLVM_ADT_STRINGREF_H
11 #define LLVM_ADT_STRINGREF_H
13 #include "llvm/Support/Allocator.h"
23 class SmallVectorImpl;
28 /// Helper functions for StringRef::getAsInteger.
29 bool getAsUnsignedInteger(StringRef Str, unsigned Radix,
30 unsigned long long &Result);
32 bool getAsSignedInteger(StringRef Str, unsigned Radix, long long &Result);
34 /// StringRef - Represent a constant reference to a string, i.e. a character
35 /// array and a length, which need not be null terminated.
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.
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;
49 /// The start of the string, in an external buffer.
52 /// The length of the string.
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.
58 static size_t min(size_t a, size_t b) { return a < b ? a : b; }
59 static size_t max(size_t a, size_t b) { return a > b ? a : b; }
61 // Workaround memcmp issue with null pointers (undefined behavior)
62 // by providing a specialized version
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);
69 /// @name Constructors
72 /// Construct an empty string ref.
73 /*implicit*/ StringRef() : Data(nullptr), Length(0) {}
75 /// Construct a string ref from a cstring.
76 /*implicit*/ StringRef(const char *Str)
78 assert(Str && "StringRef cannot be built from a NULL argument");
79 Length = ::strlen(Str); // invoking strlen(NULL) is undefined behavior
82 /// Construct a string ref from a pointer and 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");
89 /// Construct a string ref from an std::string.
90 /*implicit*/ StringRef(const std::string &Str)
91 : Data(Str.data()), Length(Str.length()) {}
97 iterator begin() const { return Data; }
99 iterator end() const { return Data + Length; }
102 /// @name String Operations
105 /// data - Get a pointer to the start of the string (which may not be null
107 const char *data() const { return Data; }
109 /// empty - Check if the string is empty.
110 bool empty() const { return Length == 0; }
112 /// size - Get the string size.
113 size_t size() const { return Length; }
115 /// front - Get the first character in the string.
121 /// back - Get the last character in the string.
124 return Data[Length-1];
127 // copy - Allocate copy in BumpPtrAllocator and return StringRef to it.
128 StringRef copy(BumpPtrAllocator &Allocator) {
129 char *S = Allocator.Allocate<char>(Length);
130 std::copy(begin(), end(), S);
131 return StringRef(S, Length);
134 /// equals - Check for string equality, this is more efficient than
135 /// compare() when the relative ordering of inequal strings isn't needed.
136 bool equals(StringRef RHS) const {
137 return (Length == RHS.Length &&
138 compareMemory(Data, RHS.Data, RHS.Length) == 0);
141 /// equals_lower - Check for string equality, ignoring case.
142 bool equals_lower(StringRef RHS) const {
143 return Length == RHS.Length && compare_lower(RHS) == 0;
146 /// compare - Compare two strings; the result is -1, 0, or 1 if this string
147 /// is lexicographically less than, equal to, or greater than the \p RHS.
148 int compare(StringRef RHS) const {
149 // Check the prefix for a mismatch.
150 if (int Res = compareMemory(Data, RHS.Data, min(Length, RHS.Length)))
151 return Res < 0 ? -1 : 1;
153 // Otherwise the prefixes match, so we only need to check the lengths.
154 if (Length == RHS.Length)
156 return Length < RHS.Length ? -1 : 1;
159 /// compare_lower - Compare two strings, ignoring case.
160 int compare_lower(StringRef RHS) const;
162 /// compare_numeric - Compare two strings, treating sequences of digits as
164 int compare_numeric(StringRef RHS) const;
166 /// \brief Determine the edit distance between this string and another
169 /// \param Other the string to compare this string against.
171 /// \param AllowReplacements whether to allow character
172 /// replacements (change one character into another) as a single
173 /// operation, rather than as two operations (an insertion and a
176 /// \param MaxEditDistance If non-zero, the maximum edit distance that
177 /// this routine is allowed to compute. If the edit distance will exceed
178 /// that maximum, returns \c MaxEditDistance+1.
180 /// \returns the minimum number of character insertions, removals,
181 /// or (if \p AllowReplacements is \c true) replacements needed to
182 /// transform one of the given strings into the other. If zero,
183 /// the strings are identical.
184 unsigned edit_distance(StringRef Other, bool AllowReplacements = true,
185 unsigned MaxEditDistance = 0) const;
187 /// str - Get the contents as an std::string.
188 std::string str() const {
189 if (Data == nullptr) return std::string();
190 return std::string(Data, Length);
194 /// @name Operator Overloads
197 char operator[](size_t Index) const {
198 assert(Index < Length && "Invalid index!");
203 /// @name Type Conversions
206 operator std::string() const {
211 /// @name String Predicates
214 /// Check if this string starts with the given \p Prefix.
215 bool startswith(StringRef Prefix) const {
216 return Length >= Prefix.Length &&
217 compareMemory(Data, Prefix.Data, Prefix.Length) == 0;
220 /// Check if this string starts with the given \p Prefix, ignoring case.
221 bool startswith_lower(StringRef Prefix) const;
223 /// Check if this string ends with the given \p Suffix.
224 bool endswith(StringRef Suffix) const {
225 return Length >= Suffix.Length &&
226 compareMemory(end() - Suffix.Length, Suffix.Data, Suffix.Length) == 0;
229 /// Check if this string ends with the given \p Suffix, ignoring case.
230 bool endswith_lower(StringRef Suffix) const;
233 /// @name String Searching
236 /// Search for the first character \p C in the string.
238 /// \returns The index of the first occurrence of \p C, or npos if not
240 size_t find(char C, size_t From = 0) const {
241 for (size_t i = min(From, Length), e = Length; i != e; ++i)
247 /// Search for the first string \p Str in the string.
249 /// \returns The index of the first occurrence of \p Str, or npos if not
251 size_t find(StringRef Str, size_t From = 0) const;
253 /// Search for the last character \p C in the string.
255 /// \returns The index of the last occurrence of \p C, or npos if not
257 size_t rfind(char C, size_t From = npos) const {
258 From = min(From, Length);
268 /// Search for the last string \p Str in the string.
270 /// \returns The index of the last occurrence of \p Str, or npos if not
272 size_t rfind(StringRef Str) const;
274 /// Find the first character in the string that is \p C, or npos if not
275 /// found. Same as find.
276 size_t find_first_of(char C, size_t From = 0) const {
277 return find(C, From);
280 /// Find the first character in the string that is in \p Chars, or npos if
283 /// Complexity: O(size() + Chars.size())
284 size_t find_first_of(StringRef Chars, size_t From = 0) const;
286 /// Find the first character in the string that is not \p C or npos if not
288 size_t find_first_not_of(char C, size_t From = 0) const;
290 /// Find the first character in the string that is not in the string
291 /// \p Chars, or npos if not found.
293 /// Complexity: O(size() + Chars.size())
294 size_t find_first_not_of(StringRef Chars, size_t From = 0) const;
296 /// Find the last character in the string that is \p C, or npos if not
298 size_t find_last_of(char C, size_t From = npos) const {
299 return rfind(C, From);
302 /// Find the last character in the string that is in \p C, or npos if not
305 /// Complexity: O(size() + Chars.size())
306 size_t find_last_of(StringRef Chars, size_t From = npos) const;
308 /// Find the last character in the string that is not \p C, or npos if not
310 size_t find_last_not_of(char C, size_t From = npos) const;
312 /// Find the last character in the string that is not in \p Chars, or
313 /// npos if not found.
315 /// Complexity: O(size() + Chars.size())
316 size_t find_last_not_of(StringRef Chars, size_t From = npos) const;
319 /// @name Helpful Algorithms
322 /// Return the number of occurrences of \p C in the string.
323 size_t count(char C) const {
325 for (size_t i = 0, e = Length; i != e; ++i)
331 /// Return the number of non-overlapped occurrences of \p Str in
333 size_t count(StringRef Str) const;
335 /// Parse the current string as an integer of the specified radix. If
336 /// \p Radix is specified as zero, this does radix autosensing using
337 /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
339 /// If the string is invalid or if only a subset of the string is valid,
340 /// this returns true to signify the error. The string is considered
341 /// erroneous if empty or if it overflows T.
342 template <typename T>
343 typename std::enable_if<std::numeric_limits<T>::is_signed, bool>::type
344 getAsInteger(unsigned Radix, T &Result) const {
346 if (getAsSignedInteger(*this, Radix, LLVal) ||
347 static_cast<T>(LLVal) != LLVal)
353 template <typename T>
354 typename std::enable_if<!std::numeric_limits<T>::is_signed, bool>::type
355 getAsInteger(unsigned Radix, T &Result) const {
356 unsigned long long ULLVal;
357 if (getAsUnsignedInteger(*this, Radix, ULLVal) ||
358 static_cast<T>(ULLVal) != ULLVal)
364 /// Parse the current string as an integer of the specified \p Radix, or of
365 /// an autosensed radix if the \p Radix given is 0. The current value in
366 /// \p Result is discarded, and the storage is changed to be wide enough to
367 /// store the parsed integer.
369 /// \returns true if the string does not solely consist of a valid
370 /// non-empty number in the appropriate base.
372 /// APInt::fromString is superficially similar but assumes the
373 /// string is well-formed in the given radix.
374 bool getAsInteger(unsigned Radix, APInt &Result) const;
377 /// @name String Operations
380 // Convert the given ASCII string to lowercase.
381 std::string lower() const;
383 /// Convert the given ASCII string to uppercase.
384 std::string upper() const;
387 /// @name Substring Operations
390 /// Return a reference to the substring from [Start, Start + N).
392 /// \param Start The index of the starting character in the substring; if
393 /// the index is npos or greater than the length of the string then the
394 /// empty substring will be returned.
396 /// \param N The number of characters to included in the substring. If N
397 /// exceeds the number of characters remaining in the string, the string
398 /// suffix (starting with \p Start) will be returned.
399 StringRef substr(size_t Start, size_t N = npos) const {
400 Start = min(Start, Length);
401 return StringRef(Data + Start, min(N, Length - Start));
404 /// Return a StringRef equal to 'this' but with the first \p N elements
406 StringRef drop_front(size_t N = 1) const {
407 assert(size() >= N && "Dropping more elements than exist");
411 /// Return a StringRef equal to 'this' but with the last \p N elements
413 StringRef drop_back(size_t N = 1) const {
414 assert(size() >= N && "Dropping more elements than exist");
415 return substr(0, size()-N);
418 /// Return a reference to the substring from [Start, End).
420 /// \param Start The index of the starting character in the substring; if
421 /// the index is npos or greater than the length of the string then the
422 /// empty substring will be returned.
424 /// \param End The index following the last character to include in the
425 /// substring. If this is npos, or less than \p Start, or exceeds the
426 /// number of characters remaining in the string, the string suffix
427 /// (starting with \p Start) will be returned.
428 StringRef slice(size_t Start, size_t End) const {
429 Start = min(Start, Length);
430 End = min(max(Start, End), Length);
431 return StringRef(Data + Start, End - Start);
434 /// Split into two substrings around the first occurrence of a separator
437 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
438 /// such that (*this == LHS + Separator + RHS) is true and RHS is
439 /// maximal. If \p Separator is not in the string, then the result is a
440 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
442 /// \param Separator The character to split on.
443 /// \returns The split substrings.
444 std::pair<StringRef, StringRef> split(char Separator) const {
445 size_t Idx = find(Separator);
447 return std::make_pair(*this, StringRef());
448 return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
451 /// Split into two substrings around the first occurrence of a separator
454 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
455 /// such that (*this == LHS + Separator + RHS) is true and RHS is
456 /// maximal. If \p Separator is not in the string, then the result is a
457 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
459 /// \param Separator - The string to split on.
460 /// \return - The split substrings.
461 std::pair<StringRef, StringRef> split(StringRef Separator) const {
462 size_t Idx = find(Separator);
464 return std::make_pair(*this, StringRef());
465 return std::make_pair(slice(0, Idx), slice(Idx + Separator.size(), npos));
468 /// Split into substrings around the occurrences of a separator string.
470 /// Each substring is stored in \p A. If \p MaxSplit is >= 0, at most
471 /// \p MaxSplit splits are done and consequently <= \p MaxSplit
472 /// elements are added to A.
473 /// If \p KeepEmpty is false, empty strings are not added to \p A. They
474 /// still count when considering \p MaxSplit
475 /// An useful invariant is that
476 /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true
478 /// \param A - Where to put the substrings.
479 /// \param Separator - The string to split on.
480 /// \param MaxSplit - The maximum number of times the string is split.
481 /// \param KeepEmpty - True if empty substring should be added.
482 void split(SmallVectorImpl<StringRef> &A,
483 StringRef Separator, int MaxSplit = -1,
484 bool KeepEmpty = true) const;
486 /// Split into two substrings around the last occurrence of a separator
489 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
490 /// such that (*this == LHS + Separator + RHS) is true and RHS is
491 /// minimal. If \p Separator is not in the string, then the result is a
492 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
494 /// \param Separator - The character to split on.
495 /// \return - The split substrings.
496 std::pair<StringRef, StringRef> rsplit(char Separator) const {
497 size_t Idx = rfind(Separator);
499 return std::make_pair(*this, StringRef());
500 return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
503 /// Return string with consecutive characters in \p Chars starting from
504 /// the left removed.
505 StringRef ltrim(StringRef Chars = " \t\n\v\f\r") const {
506 return drop_front(std::min(Length, find_first_not_of(Chars)));
509 /// Return string with consecutive characters in \p Chars starting from
510 /// the right removed.
511 StringRef rtrim(StringRef Chars = " \t\n\v\f\r") const {
512 return drop_back(Length - std::min(Length, find_last_not_of(Chars) + 1));
515 /// Return string with consecutive characters in \p Chars starting from
516 /// the left and right removed.
517 StringRef trim(StringRef Chars = " \t\n\v\f\r") const {
518 return ltrim(Chars).rtrim(Chars);
524 /// @name StringRef Comparison Operators
527 inline bool operator==(StringRef LHS, StringRef RHS) {
528 return LHS.equals(RHS);
531 inline bool operator!=(StringRef LHS, StringRef RHS) {
532 return !(LHS == RHS);
535 inline bool operator<(StringRef LHS, StringRef RHS) {
536 return LHS.compare(RHS) == -1;
539 inline bool operator<=(StringRef LHS, StringRef RHS) {
540 return LHS.compare(RHS) != 1;
543 inline bool operator>(StringRef LHS, StringRef RHS) {
544 return LHS.compare(RHS) == 1;
547 inline bool operator>=(StringRef LHS, StringRef RHS) {
548 return LHS.compare(RHS) != -1;
551 inline std::string &operator+=(std::string &buffer, StringRef string) {
552 return buffer.append(string.data(), string.size());
557 /// \brief Compute a hash_code for a StringRef.
558 hash_code hash_value(StringRef S);
560 // StringRefs can be treated like a POD type.
561 template <typename T> struct isPodLike;
562 template <> struct isPodLike<StringRef> { static const bool value = true; };