Line data Source code
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 <algorithm>
14 : #include <cassert>
15 : #include <cstring>
16 : #include <limits>
17 : #include <string>
18 : #include <utility>
19 :
20 : namespace llvm {
21 : template <typename T>
22 : class SmallVectorImpl;
23 : class APInt;
24 : class hash_code;
25 : class StringRef;
26 :
27 : /// Helper functions for StringRef::getAsInteger.
28 : bool getAsUnsignedInteger(StringRef Str, unsigned Radix,
29 : unsigned long long &Result);
30 :
31 : bool getAsSignedInteger(StringRef Str, unsigned Radix, long long &Result);
32 :
33 : /// StringRef - Represent a constant reference to a string, i.e. a character
34 : /// array and a length, which need not be null terminated.
35 : ///
36 : /// This class does not own the string data, it is expected to be used in
37 : /// situations where the character data resides in some other buffer, whose
38 : /// lifetime extends past that of the StringRef. For this reason, it is not in
39 : /// general safe to store a StringRef.
40 : class StringRef {
41 : public:
42 : typedef const char *iterator;
43 : typedef const char *const_iterator;
44 : static const size_t npos = ~size_t(0);
45 : typedef size_t size_type;
46 :
47 : private:
48 : /// The start of the string, in an external buffer.
49 : const char *Data;
50 :
51 : /// The length of the string.
52 : size_t Length;
53 :
54 : // Workaround memcmp issue with null pointers (undefined behavior)
55 : // by providing a specialized version
56 : static int compareMemory(const char *Lhs, const char *Rhs, size_t Length) {
57 : if (Length == 0) { return 0; }
58 : return ::memcmp(Lhs,Rhs,Length);
59 : }
60 :
61 : public:
62 : /// @name Constructors
63 : /// @{
64 :
65 : /// Construct an empty string ref.
66 : /*implicit*/ StringRef() : Data(nullptr), Length(0) {}
67 :
68 : /// Construct a string ref from a cstring.
69 : /*implicit*/ StringRef(const char *Str)
70 : : Data(Str) {
71 : assert(Str && "StringRef cannot be built from a NULL argument");
72 : Length = ::strlen(Str); // invoking strlen(NULL) is undefined behavior
73 : }
74 :
75 : /// Construct a string ref from a pointer and length.
76 : /*implicit*/ StringRef(const char *data, size_t length)
77 : : Data(data), Length(length) {
78 : assert((data || length == 0) &&
79 : "StringRef cannot be built from a NULL argument with non-null length");
80 : }
81 :
82 : /// Construct a string ref from an std::string.
83 : /*implicit*/ StringRef(const std::string &Str)
84 41 : : Data(Str.data()), Length(Str.length()) {}
85 :
86 : /// @}
87 : /// @name Iterators
88 : /// @{
89 :
90 : iterator begin() const { return Data; }
91 :
92 : iterator end() const { return Data + Length; }
93 :
94 : const unsigned char *bytes_begin() const {
95 : return reinterpret_cast<const unsigned char *>(begin());
96 : }
97 : const unsigned char *bytes_end() const {
98 : return reinterpret_cast<const unsigned char *>(end());
99 : }
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).
107 : const char *data() const { return Data; }
108 :
109 : /// empty - Check if the string is empty.
110 : bool empty() const { return Length == 0; }
111 :
112 : /// size - Get the string size.
113 : size_t size() const { return Length; }
114 :
115 : /// front - Get the first character in the string.
116 : char front() const {
117 : assert(!empty());
118 : return Data[0];
119 : }
120 :
121 : /// back - Get the last character in the string.
122 : char back() const {
123 : assert(!empty());
124 : return Data[Length-1];
125 : }
126 :
127 : // copy - Allocate copy in Allocator and return StringRef to it.
128 : template <typename Allocator> StringRef copy(Allocator &A) const {
129 : char *S = A.template Allocate<char>(Length);
130 : std::copy(begin(), end(), S);
131 : return StringRef(S, Length);
132 : }
133 :
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);
139 : }
140 :
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;
144 : }
145 :
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, std::min(Length, RHS.Length)))
151 : return Res < 0 ? -1 : 1;
152 :
153 : // Otherwise the prefixes match, so we only need to check the lengths.
154 : if (Length == RHS.Length)
155 : return 0;
156 : return Length < RHS.Length ? -1 : 1;
157 : }
158 :
159 : /// compare_lower - Compare two strings, ignoring case.
160 : int compare_lower(StringRef RHS) const;
161 :
162 : /// compare_numeric - Compare two strings, treating sequences of digits as
163 : /// numbers.
164 : int compare_numeric(StringRef RHS) const;
165 :
166 : /// \brief Determine the edit distance between this string and another
167 : /// string.
168 : ///
169 : /// \param Other the string to compare this string against.
170 : ///
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
174 : /// removal).
175 : ///
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.
179 : ///
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;
186 :
187 : /// str - Get the contents as an std::string.
188 : std::string str() const {
189 : if (!Data) return std::string();
190 : return std::string(Data, Length);
191 : }
192 :
193 : /// @}
194 : /// @name Operator Overloads
195 : /// @{
196 :
197 : char operator[](size_t Index) const {
198 : assert(Index < Length && "Invalid index!");
199 : return Data[Index];
200 : }
201 :
202 : /// @}
203 : /// @name Type Conversions
204 : /// @{
205 :
206 : operator std::string() const {
207 : return str();
208 : }
209 :
210 : /// @}
211 : /// @name String Predicates
212 : /// @{
213 :
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;
218 : }
219 :
220 : /// Check if this string starts with the given \p Prefix, ignoring case.
221 : bool startswith_lower(StringRef Prefix) const;
222 :
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;
227 : }
228 :
229 : /// Check if this string ends with the given \p Suffix, ignoring case.
230 : bool endswith_lower(StringRef Suffix) const;
231 :
232 : /// @}
233 : /// @name String Searching
234 : /// @{
235 :
236 : /// Search for the first character \p C in the string.
237 : ///
238 : /// \returns The index of the first occurrence of \p C, or npos if not
239 : /// found.
240 : size_t find(char C, size_t From = 0) const {
241 : size_t FindBegin = std::min(From, Length);
242 : if (FindBegin < Length) { // Avoid calling memchr with nullptr.
243 : // Just forward to memchr, which is faster than a hand-rolled loop.
244 : if (const void *P = ::memchr(Data + FindBegin, C, Length - FindBegin))
245 : return static_cast<const char *>(P) - Data;
246 : }
247 : return npos;
248 : }
249 :
250 : /// Search for the first string \p Str in the string.
251 : ///
252 : /// \returns The index of the first occurrence of \p Str, or npos if not
253 : /// found.
254 : size_t find(StringRef Str, size_t From = 0) const;
255 :
256 : /// Search for the last character \p C in the string.
257 : ///
258 : /// \returns The index of the last occurrence of \p C, or npos if not
259 : /// found.
260 : size_t rfind(char C, size_t From = npos) const {
261 : From = std::min(From, Length);
262 : size_t i = From;
263 : while (i != 0) {
264 : --i;
265 : if (Data[i] == C)
266 : return i;
267 : }
268 : return npos;
269 : }
270 :
271 : /// Search for the last string \p Str in the string.
272 : ///
273 : /// \returns The index of the last occurrence of \p Str, or npos if not
274 : /// found.
275 : size_t rfind(StringRef Str) const;
276 :
277 : /// Find the first character in the string that is \p C, or npos if not
278 : /// found. Same as find.
279 : size_t find_first_of(char C, size_t From = 0) const {
280 : return find(C, From);
281 : }
282 :
283 : /// Find the first character in the string that is in \p Chars, or npos if
284 : /// not found.
285 : ///
286 : /// Complexity: O(size() + Chars.size())
287 : size_t find_first_of(StringRef Chars, size_t From = 0) const;
288 :
289 : /// Find the first character in the string that is not \p C or npos if not
290 : /// found.
291 : size_t find_first_not_of(char C, size_t From = 0) const;
292 :
293 : /// Find the first character in the string that is not in the string
294 : /// \p Chars, or npos if not found.
295 : ///
296 : /// Complexity: O(size() + Chars.size())
297 : size_t find_first_not_of(StringRef Chars, size_t From = 0) const;
298 :
299 : /// Find the last character in the string that is \p C, or npos if not
300 : /// found.
301 : size_t find_last_of(char C, size_t From = npos) const {
302 : return rfind(C, From);
303 : }
304 :
305 : /// Find the last character in the string that is in \p C, or npos if not
306 : /// found.
307 : ///
308 : /// Complexity: O(size() + Chars.size())
309 : size_t find_last_of(StringRef Chars, size_t From = npos) const;
310 :
311 : /// Find the last character in the string that is not \p C, or npos if not
312 : /// found.
313 : size_t find_last_not_of(char C, size_t From = npos) const;
314 :
315 : /// Find the last character in the string that is not in \p Chars, or
316 : /// npos if not found.
317 : ///
318 : /// Complexity: O(size() + Chars.size())
319 : size_t find_last_not_of(StringRef Chars, size_t From = npos) const;
320 :
321 : /// @}
322 : /// @name Helpful Algorithms
323 : /// @{
324 :
325 : /// Return the number of occurrences of \p C in the string.
326 : size_t count(char C) const {
327 : size_t Count = 0;
328 : for (size_t i = 0, e = Length; i != e; ++i)
329 : if (Data[i] == C)
330 : ++Count;
331 : return Count;
332 : }
333 :
334 : /// Return the number of non-overlapped occurrences of \p Str in
335 : /// the string.
336 : size_t count(StringRef Str) const;
337 :
338 : /// Parse the current string as an integer of the specified radix. If
339 : /// \p Radix is specified as zero, this does radix autosensing using
340 : /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
341 : ///
342 : /// If the string is invalid or if only a subset of the string is valid,
343 : /// this returns true to signify the error. The string is considered
344 : /// erroneous if empty or if it overflows T.
345 : template <typename T>
346 : typename std::enable_if<std::numeric_limits<T>::is_signed, bool>::type
347 : getAsInteger(unsigned Radix, T &Result) const {
348 : long long LLVal;
349 : if (getAsSignedInteger(*this, Radix, LLVal) ||
350 : static_cast<T>(LLVal) != LLVal)
351 : return true;
352 : Result = LLVal;
353 : return false;
354 : }
355 :
356 : template <typename T>
357 : typename std::enable_if<!std::numeric_limits<T>::is_signed, bool>::type
358 : getAsInteger(unsigned Radix, T &Result) const {
359 : unsigned long long ULLVal;
360 : // The additional cast to unsigned long long is required to avoid the
361 : // Visual C++ warning C4805: '!=' : unsafe mix of type 'bool' and type
362 : // 'unsigned __int64' when instantiating getAsInteger with T = bool.
363 : if (getAsUnsignedInteger(*this, Radix, ULLVal) ||
364 : static_cast<unsigned long long>(static_cast<T>(ULLVal)) != ULLVal)
365 : return true;
366 : Result = ULLVal;
367 : return false;
368 : }
369 :
370 : /// Parse the current string as an integer of the specified \p Radix, or of
371 : /// an autosensed radix if the \p Radix given is 0. The current value in
372 : /// \p Result is discarded, and the storage is changed to be wide enough to
373 : /// store the parsed integer.
374 : ///
375 : /// \returns true if the string does not solely consist of a valid
376 : /// non-empty number in the appropriate base.
377 : ///
378 : /// APInt::fromString is superficially similar but assumes the
379 : /// string is well-formed in the given radix.
380 : bool getAsInteger(unsigned Radix, APInt &Result) const;
381 :
382 : /// @}
383 : /// @name String Operations
384 : /// @{
385 :
386 : // Convert the given ASCII string to lowercase.
387 : std::string lower() const;
388 :
389 : /// Convert the given ASCII string to uppercase.
390 : std::string upper() const;
391 :
392 : /// @}
393 : /// @name Substring Operations
394 : /// @{
395 :
396 : /// Return a reference to the substring from [Start, Start + N).
397 : ///
398 : /// \param Start The index of the starting character in the substring; if
399 : /// the index is npos or greater than the length of the string then the
400 : /// empty substring will be returned.
401 : ///
402 : /// \param N The number of characters to included in the substring. If N
403 : /// exceeds the number of characters remaining in the string, the string
404 : /// suffix (starting with \p Start) will be returned.
405 : StringRef substr(size_t Start, size_t N = npos) const {
406 : Start = std::min(Start, Length);
407 : return StringRef(Data + Start, std::min(N, Length - Start));
408 : }
409 :
410 : /// Return a StringRef equal to 'this' but with the first \p N elements
411 : /// dropped.
412 : StringRef drop_front(size_t N = 1) const {
413 : assert(size() >= N && "Dropping more elements than exist");
414 : return substr(N);
415 : }
416 :
417 : /// Return a StringRef equal to 'this' but with the last \p N elements
418 : /// dropped.
419 : StringRef drop_back(size_t N = 1) const {
420 : assert(size() >= N && "Dropping more elements than exist");
421 : return substr(0, size()-N);
422 : }
423 :
424 : /// Return a reference to the substring from [Start, End).
425 : ///
426 : /// \param Start The index of the starting character in the substring; if
427 : /// the index is npos or greater than the length of the string then the
428 : /// empty substring will be returned.
429 : ///
430 : /// \param End The index following the last character to include in the
431 : /// substring. If this is npos, or less than \p Start, or exceeds the
432 : /// number of characters remaining in the string, the string suffix
433 : /// (starting with \p Start) will be returned.
434 : StringRef slice(size_t Start, size_t End) const {
435 : Start = std::min(Start, Length);
436 : End = std::min(std::max(Start, End), Length);
437 : return StringRef(Data + Start, End - Start);
438 : }
439 :
440 : /// Split into two substrings around the first occurrence of a separator
441 : /// character.
442 : ///
443 : /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
444 : /// such that (*this == LHS + Separator + RHS) is true and RHS is
445 : /// maximal. If \p Separator is not in the string, then the result is a
446 : /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
447 : ///
448 : /// \param Separator The character to split on.
449 : /// \returns The split substrings.
450 : std::pair<StringRef, StringRef> split(char Separator) const {
451 : size_t Idx = find(Separator);
452 : if (Idx == npos)
453 : return std::make_pair(*this, StringRef());
454 : return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
455 : }
456 :
457 : /// Split into two substrings around the first occurrence of a separator
458 : /// string.
459 : ///
460 : /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
461 : /// such that (*this == LHS + Separator + RHS) is true and RHS is
462 : /// maximal. If \p Separator is not in the string, then the result is a
463 : /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
464 : ///
465 : /// \param Separator - The string to split on.
466 : /// \return - The split substrings.
467 : std::pair<StringRef, StringRef> split(StringRef Separator) const {
468 : size_t Idx = find(Separator);
469 : if (Idx == npos)
470 : return std::make_pair(*this, StringRef());
471 : return std::make_pair(slice(0, Idx), slice(Idx + Separator.size(), npos));
472 : }
473 :
474 : /// Split into substrings around the occurrences of a separator string.
475 : ///
476 : /// Each substring is stored in \p A. If \p MaxSplit is >= 0, at most
477 : /// \p MaxSplit splits are done and consequently <= \p MaxSplit
478 : /// elements are added to A.
479 : /// If \p KeepEmpty is false, empty strings are not added to \p A. They
480 : /// still count when considering \p MaxSplit
481 : /// An useful invariant is that
482 : /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true
483 : ///
484 : /// \param A - Where to put the substrings.
485 : /// \param Separator - The string to split on.
486 : /// \param MaxSplit - The maximum number of times the string is split.
487 : /// \param KeepEmpty - True if empty substring should be added.
488 : void split(SmallVectorImpl<StringRef> &A,
489 : StringRef Separator, int MaxSplit = -1,
490 : bool KeepEmpty = true) const;
491 :
492 : /// Split into two substrings around the last occurrence of a separator
493 : /// character.
494 : ///
495 : /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
496 : /// such that (*this == LHS + Separator + RHS) is true and RHS is
497 : /// minimal. If \p Separator is not in the string, then the result is a
498 : /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
499 : ///
500 : /// \param Separator - The character to split on.
501 : /// \return - The split substrings.
502 : std::pair<StringRef, StringRef> rsplit(char Separator) const {
503 : size_t Idx = rfind(Separator);
504 : if (Idx == npos)
505 : return std::make_pair(*this, StringRef());
506 : return std::make_pair(slice(0, Idx), slice(Idx+1, npos));
507 : }
508 :
509 : /// Return string with consecutive characters in \p Chars starting from
510 : /// the left removed.
511 : StringRef ltrim(StringRef Chars = " \t\n\v\f\r") const {
512 : return drop_front(std::min(Length, find_first_not_of(Chars)));
513 : }
514 :
515 : /// Return string with consecutive characters in \p Chars starting from
516 : /// the right removed.
517 : StringRef rtrim(StringRef Chars = " \t\n\v\f\r") const {
518 : return drop_back(Length - std::min(Length, find_last_not_of(Chars) + 1));
519 : }
520 :
521 : /// Return string with consecutive characters in \p Chars starting from
522 : /// the left and right removed.
523 : StringRef trim(StringRef Chars = " \t\n\v\f\r") const {
524 : return ltrim(Chars).rtrim(Chars);
525 : }
526 :
527 : /// @}
528 : };
529 :
530 : /// @name StringRef Comparison Operators
531 : /// @{
532 :
533 : inline bool operator==(StringRef LHS, StringRef RHS) {
534 : return LHS.equals(RHS);
535 : }
536 :
537 : inline bool operator!=(StringRef LHS, StringRef RHS) {
538 : return !(LHS == RHS);
539 : }
540 :
541 : inline bool operator<(StringRef LHS, StringRef RHS) {
542 : return LHS.compare(RHS) == -1;
543 : }
544 :
545 : inline bool operator<=(StringRef LHS, StringRef RHS) {
546 : return LHS.compare(RHS) != 1;
547 : }
548 :
549 : inline bool operator>(StringRef LHS, StringRef RHS) {
550 : return LHS.compare(RHS) == 1;
551 : }
552 :
553 : inline bool operator>=(StringRef LHS, StringRef RHS) {
554 : return LHS.compare(RHS) != -1;
555 : }
556 :
557 : inline std::string &operator+=(std::string &buffer, StringRef string) {
558 : return buffer.append(string.data(), string.size());
559 : }
560 :
561 : /// @}
562 :
563 : /// \brief Compute a hash_code for a StringRef.
564 : hash_code hash_value(StringRef S);
565 :
566 : // StringRefs can be treated like a POD type.
567 : template <typename T> struct isPodLike;
568 : template <> struct isPodLike<StringRef> { static const bool value = true; };
569 : }
570 :
571 : #endif
|
