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
20 class SmallVectorImpl;
23 /// StringRef - Represent a constant reference to a string, i.e. a character
24 /// array and a length, which need not be null terminated.
26 /// This class does not own the string data, it is expected to be used in
27 /// situations where the character data resides in some other buffer, whose
28 /// lifetime extends past that of the StringRef. For this reason, it is not in
29 /// general safe to store a StringRef.
32 typedef const char *iterator;
33 typedef const char *const_iterator;
34 static const size_t npos = ~size_t(0);
35 typedef size_t size_type;
38 /// The start of the string, in an external buffer.
41 /// The length of the string.
44 // Workaround PR5482: nearly all gcc 4.x miscompile StringRef and std::min()
45 // Changing the arg of min to be an integer, instead of a reference to an
46 // integer works around this bug.
47 static size_t min(size_t a, size_t b) { return a < b ? a : b; }
48 static size_t max(size_t a, size_t b) { return a > b ? a : b; }
50 // Workaround memcmp issue with null pointers (undefined behavior)
51 // by providing a specialized version
52 static int compareMemory(const char *Lhs, const char *Rhs, size_t Length) {
53 if (Length == 0) { return 0; }
54 return ::memcmp(Lhs,Rhs,Length);
58 /// @name Constructors
61 /// Construct an empty string ref.
62 /*implicit*/ StringRef() : Data(0), Length(0) {}
64 /// Construct a string ref from a cstring.
65 /*implicit*/ StringRef(const char *Str)
67 assert(Str && "StringRef cannot be built from a NULL argument");
68 Length = ::strlen(Str); // invoking strlen(NULL) is undefined behavior
71 /// Construct a string ref from a pointer and length.
72 /*implicit*/ StringRef(const char *data, size_t length)
73 : Data(data), Length(length) {
74 assert((data || length == 0) &&
75 "StringRef cannot be built from a NULL argument with non-null length");
78 /// Construct a string ref from an std::string.
79 /*implicit*/ StringRef(const std::string &Str)
80 : Data(Str.data()), Length(Str.length()) {}
86 iterator begin() const { return Data; }
88 iterator end() const { return Data + Length; }
91 /// @name String Operations
94 /// data - Get a pointer to the start of the string (which may not be null
96 const char *data() const { return Data; }
98 /// empty - Check if the string is empty.
99 bool empty() const { return Length == 0; }
101 /// size - Get the string size.
102 size_t size() const { return Length; }
104 /// front - Get the first character in the string.
110 /// back - Get the last character in the string.
113 return Data[Length-1];
116 /// equals - Check for string equality, this is more efficient than
117 /// compare() when the relative ordering of inequal strings isn't needed.
118 bool equals(StringRef RHS) const {
119 return (Length == RHS.Length &&
120 compareMemory(Data, RHS.Data, RHS.Length) == 0);
123 /// equals_lower - Check for string equality, ignoring case.
124 bool equals_lower(StringRef RHS) const {
125 return Length == RHS.Length && compare_lower(RHS) == 0;
128 /// compare - Compare two strings; the result is -1, 0, or 1 if this string
129 /// is lexicographically less than, equal to, or greater than the \arg RHS.
130 int compare(StringRef RHS) const {
131 // Check the prefix for a mismatch.
132 if (int Res = compareMemory(Data, RHS.Data, min(Length, RHS.Length)))
133 return Res < 0 ? -1 : 1;
135 // Otherwise the prefixes match, so we only need to check the lengths.
136 if (Length == RHS.Length)
138 return Length < RHS.Length ? -1 : 1;
141 /// compare_lower - Compare two strings, ignoring case.
142 int compare_lower(StringRef RHS) const;
144 /// compare_numeric - Compare two strings, treating sequences of digits as
146 int compare_numeric(StringRef RHS) const;
148 /// \brief Determine the edit distance between this string and another
151 /// \param Other the string to compare this string against.
153 /// \param AllowReplacements whether to allow character
154 /// replacements (change one character into another) as a single
155 /// operation, rather than as two operations (an insertion and a
158 /// \param MaxEditDistance If non-zero, the maximum edit distance that
159 /// this routine is allowed to compute. If the edit distance will exceed
160 /// that maximum, returns \c MaxEditDistance+1.
162 /// \returns the minimum number of character insertions, removals,
163 /// or (if \p AllowReplacements is \c true) replacements needed to
164 /// transform one of the given strings into the other. If zero,
165 /// the strings are identical.
166 unsigned edit_distance(StringRef Other, bool AllowReplacements = true,
167 unsigned MaxEditDistance = 0);
169 /// str - Get the contents as an std::string.
170 std::string str() const {
171 if (Data == 0) return std::string();
172 return std::string(Data, Length);
176 /// @name Operator Overloads
179 char operator[](size_t Index) const {
180 assert(Index < Length && "Invalid index!");
185 /// @name Type Conversions
188 operator std::string() const {
193 /// @name String Predicates
196 /// startswith - Check if this string starts with the given \arg Prefix.
197 bool startswith(StringRef Prefix) const {
198 return Length >= Prefix.Length &&
199 compareMemory(Data, Prefix.Data, Prefix.Length) == 0;
202 /// endswith - Check if this string ends with the given \arg Suffix.
203 bool endswith(StringRef Suffix) const {
204 return Length >= Suffix.Length &&
205 compareMemory(end() - Suffix.Length, Suffix.Data, Suffix.Length) == 0;
209 /// @name String Searching
212 /// find - Search for the first character \arg C in the string.
214 /// \return - The index of the first occurrence of \arg C, or npos if not
216 size_t find(char C, size_t From = 0) const {
217 for (size_t i = min(From, Length), e = Length; i != e; ++i)
223 /// find - Search for the first string \arg Str in the string.
225 /// \return - The index of the first occurrence of \arg Str, or npos if not
227 size_t find(StringRef Str, size_t From = 0) const;
229 /// rfind - Search for the last character \arg C in the string.
231 /// \return - The index of the last occurrence of \arg C, or npos if not
233 size_t rfind(char C, size_t From = npos) const {
234 From = min(From, Length);
244 /// rfind - Search for the last string \arg Str in the string.
246 /// \return - The index of the last occurrence of \arg Str, or npos if not
248 size_t rfind(StringRef Str) const;
250 /// find_first_of - Find the first character in the string that is \arg C,
251 /// or npos if not found. Same as find.
252 size_type find_first_of(char C, size_t From = 0) const {
253 return find(C, From);
256 /// find_first_of - Find the first character in the string that is in \arg
257 /// Chars, or npos if not found.
259 /// Note: O(size() + Chars.size())
260 size_type find_first_of(StringRef Chars, size_t From = 0) const;
262 /// find_first_not_of - Find the first character in the string that is not
263 /// \arg C or npos if not found.
264 size_type find_first_not_of(char C, size_t From = 0) const;
266 /// find_first_not_of - Find the first character in the string that is not
267 /// in the string \arg Chars, or npos if not found.
269 /// Note: O(size() + Chars.size())
270 size_type find_first_not_of(StringRef Chars, size_t From = 0) const;
272 /// find_last_of - Find the last character in the string that is \arg C, or
273 /// npos if not found.
274 size_type find_last_of(char C, size_t From = npos) const {
275 return rfind(C, From);
278 /// find_last_of - Find the last character in the string that is in \arg C,
279 /// or npos if not found.
281 /// Note: O(size() + Chars.size())
282 size_type find_last_of(StringRef Chars, size_t From = npos) const;
285 /// @name Helpful Algorithms
288 /// count - Return the number of occurrences of \arg C in the string.
289 size_t count(char C) const {
291 for (size_t i = 0, e = Length; i != e; ++i)
297 /// count - Return the number of non-overlapped occurrences of \arg Str in
299 size_t count(StringRef Str) const;
301 /// getAsInteger - Parse the current string as an integer of the specified
302 /// radix. If Radix is specified as zero, this does radix autosensing using
303 /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
305 /// If the string is invalid or if only a subset of the string is valid,
306 /// this returns true to signify the error. The string is considered
307 /// erroneous if empty.
309 bool getAsInteger(unsigned Radix, long long &Result) const;
310 bool getAsInteger(unsigned Radix, unsigned long long &Result) const;
311 bool getAsInteger(unsigned Radix, int &Result) const;
312 bool getAsInteger(unsigned Radix, unsigned &Result) const;
314 // TODO: Provide overloads for int/unsigned that check for overflow.
316 /// getAsInteger - Parse the current string as an integer of the
317 /// specified radix, or of an autosensed radix if the radix given
318 /// is 0. The current value in Result is discarded, and the
319 /// storage is changed to be wide enough to store the parsed
322 /// Returns true if the string does not solely consist of a valid
323 /// non-empty number in the appropriate base.
325 /// APInt::fromString is superficially similar but assumes the
326 /// string is well-formed in the given radix.
327 bool getAsInteger(unsigned Radix, APInt &Result) const;
330 /// @name String Operations
333 // lower - Convert the given ASCII string to lowercase.
334 std::string lower() const;
336 /// upper - Convert the given ASCII string to uppercase.
337 std::string upper() const;
340 /// @name Substring Operations
343 /// substr - Return a reference to the substring from [Start, Start + N).
345 /// \param Start - The index of the starting character in the substring; if
346 /// the index is npos or greater than the length of the string then the
347 /// empty substring will be returned.
349 /// \param N - The number of characters to included in the substring. If N
350 /// exceeds the number of characters remaining in the string, the string
351 /// suffix (starting with \arg Start) will be returned.
352 StringRef substr(size_t Start, size_t N = npos) const {
353 Start = min(Start, Length);
354 return StringRef(Data + Start, min(N, Length - Start));
357 /// slice - Return a reference to the substring from [Start, End).
359 /// \param Start - The index of the starting character in the substring; if
360 /// the index is npos or greater than the length of the string then the
361 /// empty substring will be returned.
363 /// \param End - The index following the last character to include in the
364 /// substring. If this is npos, or less than \arg Start, or exceeds the
365 /// number of characters remaining in the string, the string suffix
366 /// (starting with \arg Start) will be returned.
367 StringRef slice(size_t Start, size_t End) const {
368 Start = min(Start, Length);
369 End = min(max(Start, End), Length);
370 return StringRef(Data + Start, End - Start);
373 /// split - Split into two substrings around the first occurrence of a
374 /// separator character.
376 /// If \arg Separator is in the string, then the result is a pair (LHS, RHS)
377 /// such that (*this == LHS + Separator + RHS) is true and RHS is
378 /// maximal. If \arg Separator is not in the string, then the result is a
379 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
381 /// \param Separator - The character to split on.
382 /// \return - The split substrings.
383 std::pair<StringRef, StringRef> split(char Separator) const {
384 size_t Idx = find(Separator);
386 return std::make_pair(*this, StringRef());
387 return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
390 /// split - Split into two substrings around the first occurrence of a
391 /// separator string.
393 /// If \arg Separator is in the string, then the result is a pair (LHS, RHS)
394 /// such that (*this == LHS + Separator + RHS) is true and RHS is
395 /// maximal. If \arg Separator is not in the string, then the result is a
396 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
398 /// \param Separator - The string to split on.
399 /// \return - The split substrings.
400 std::pair<StringRef, StringRef> split(StringRef Separator) const {
401 size_t Idx = find(Separator);
403 return std::make_pair(*this, StringRef());
404 return std::make_pair(slice(0, Idx), slice(Idx + Separator.size(), npos));
407 /// split - Split into substrings around the occurrences of a separator
410 /// Each substring is stored in \arg A. If \arg MaxSplit is >= 0, at most
411 /// \arg MaxSplit splits are done and consequently <= \arg MaxSplit
412 /// elements are added to A.
413 /// If \arg KeepEmpty is false, empty strings are not added to \arg A. They
414 /// still count when considering \arg MaxSplit
415 /// An useful invariant is that
416 /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true
418 /// \param A - Where to put the substrings.
419 /// \param Separator - The string to split on.
420 /// \param MaxSplit - The maximum number of times the string is split.
421 /// \param KeepEmpty - True if empty substring should be added.
422 void split(SmallVectorImpl<StringRef> &A,
423 StringRef Separator, int MaxSplit = -1,
424 bool KeepEmpty = true) const;
426 /// rsplit - Split into two substrings around the last occurrence of a
427 /// separator character.
429 /// If \arg Separator is in the string, then the result is a pair (LHS, RHS)
430 /// such that (*this == LHS + Separator + RHS) is true and RHS is
431 /// minimal. If \arg Separator is not in the string, then the result is a
432 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
434 /// \param Separator - The character to split on.
435 /// \return - The split substrings.
436 std::pair<StringRef, StringRef> rsplit(char Separator) const {
437 size_t Idx = rfind(Separator);
439 return std::make_pair(*this, StringRef());
440 return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
446 /// @name StringRef Comparison Operators
449 inline bool operator==(StringRef LHS, StringRef RHS) {
450 return LHS.equals(RHS);
453 inline bool operator!=(StringRef LHS, StringRef RHS) {
454 return !(LHS == RHS);
457 inline bool operator<(StringRef LHS, StringRef RHS) {
458 return LHS.compare(RHS) == -1;
461 inline bool operator<=(StringRef LHS, StringRef RHS) {
462 return LHS.compare(RHS) != 1;
465 inline bool operator>(StringRef LHS, StringRef RHS) {
466 return LHS.compare(RHS) == 1;
469 inline bool operator>=(StringRef LHS, StringRef RHS) {
470 return LHS.compare(RHS) != -1;
473 inline std::string &operator+=(std::string &buffer, llvm::StringRef string) {
474 return buffer.append(string.data(), string.size());
479 // StringRefs can be treated like a POD type.
480 template <typename T> struct isPodLike;
481 template <> struct isPodLike<StringRef> { static const bool value = true; };