1//===- StringRef.h - Constant String Reference Wrapper ----------*- C++ -*-===//
2//
3// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4// See https://llvm.org/LICENSE.txt for license information.
5// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6//
7//===----------------------------------------------------------------------===//
8
9#ifndef LLVM_ADT_STRINGREF_H
10#define LLVM_ADT_STRINGREF_H
11
12#include "llvm/ADT/DenseMapInfo.h"
13#include "llvm/ADT/STLFunctionalExtras.h"
14#include "llvm/ADT/iterator_range.h"
15#include "llvm/Support/Compiler.h"
16#include <algorithm>
17#include <cassert>
18#include <cstddef>
19#include <cstring>
20#include <iterator>
21#include <limits>
22#include <string>
23#include <string_view>
24#include <type_traits>
25#include <utility>
26
27namespace llvm {
28
29 class APInt;
30 class hash_code;
31 template <typename T> class SmallVectorImpl;
32 class StringRef;
33
34 /// Helper functions for StringRef::getAsInteger.
35 bool getAsUnsignedInteger(StringRef Str, unsigned Radix,
36 unsigned long long &Result);
37
38 bool getAsSignedInteger(StringRef Str, unsigned Radix, long long &Result);
39
40 bool consumeUnsignedInteger(StringRef &Str, unsigned Radix,
41 unsigned long long &Result);
42 bool consumeSignedInteger(StringRef &Str, unsigned Radix, long long &Result);
43
44 /// StringRef - Represent a constant reference to a string, i.e. a character
45 /// array and a length, which need not be null terminated.
46 ///
47 /// This class does not own the string data, it is expected to be used in
48 /// situations where the character data resides in some other buffer, whose
49 /// lifetime extends past that of the StringRef. For this reason, it is not in
50 /// general safe to store a StringRef.
51 class LLVM_GSL_POINTER StringRef {
52 public:
53 static constexpr size_t npos = ~size_t(0);
54
55 using iterator = const char *;
56 using const_iterator = const char *;
57 using size_type = size_t;
58 using value_type = char;
59 using reverse_iterator = std::reverse_iterator<iterator>;
60 using const_reverse_iterator = std::reverse_iterator<const_iterator>;
61
62 private:
63 /// The start of the string, in an external buffer.
64 const char *Data = nullptr;
65
66 /// The length of the string.
67 size_t Length = 0;
68
69 // Workaround memcmp issue with null pointers (undefined behavior)
70 // by providing a specialized version
71 static int compareMemory(const char *Lhs, const char *Rhs, size_t Length) {
72 if (Length == 0) { return 0; }
73 return ::memcmp(s1: Lhs,s2: Rhs,n: Length);
74 }
75
76 public:
77 /// @name Constructors
78 /// @{
79
80 /// Construct an empty string ref.
81 /*implicit*/ StringRef() = default;
82
83 /// Disable conversion from nullptr. This prevents things like
84 /// if (S == nullptr)
85 StringRef(std::nullptr_t) = delete;
86
87 /// Construct a string ref from a cstring.
88 /*implicit*/ constexpr StringRef(const char *Str LLVM_LIFETIME_BOUND)
89 : Data(Str), Length(Str ?
90 // GCC 7 doesn't have constexpr char_traits. Fall back to __builtin_strlen.
91#if defined(_GLIBCXX_RELEASE) && _GLIBCXX_RELEASE < 8
92 __builtin_strlen(Str)
93#else
94 std::char_traits<char>::length(s: Str)
95#endif
96 : 0) {
97 }
98
99 /// Construct a string ref from a pointer and length.
100 /*implicit*/ constexpr StringRef(const char *data LLVM_LIFETIME_BOUND,
101 size_t length)
102 : Data(data), Length(length) {}
103
104 /// Construct a string ref from an std::string.
105 /*implicit*/ StringRef(const std::string &Str)
106 : Data(Str.data()), Length(Str.length()) {}
107
108 /// Construct a string ref from an std::string_view.
109 /*implicit*/ constexpr StringRef(std::string_view Str)
110 : Data(Str.data()), Length(Str.size()) {}
111
112 /// @}
113 /// @name Iterators
114 /// @{
115
116 iterator begin() const { return data(); }
117
118 iterator end() const { return data() + size(); }
119
120 reverse_iterator rbegin() const {
121 return std::make_reverse_iterator(i: end());
122 }
123
124 reverse_iterator rend() const {
125 return std::make_reverse_iterator(i: begin());
126 }
127
128 const unsigned char *bytes_begin() const {
129 return reinterpret_cast<const unsigned char *>(begin());
130 }
131 const unsigned char *bytes_end() const {
132 return reinterpret_cast<const unsigned char *>(end());
133 }
134 iterator_range<const unsigned char *> bytes() const {
135 return make_range(x: bytes_begin(), y: bytes_end());
136 }
137
138 /// @}
139 /// @name String Operations
140 /// @{
141
142 /// data - Get a pointer to the start of the string (which may not be null
143 /// terminated).
144 [[nodiscard]] constexpr const char *data() const { return Data; }
145
146 /// empty - Check if the string is empty.
147 [[nodiscard]] constexpr bool empty() const { return size() == 0; }
148
149 /// size - Get the string size.
150 [[nodiscard]] constexpr size_t size() const { return Length; }
151
152 /// front - Get the first character in the string.
153 [[nodiscard]] char front() const {
154 assert(!empty());
155 return data()[0];
156 }
157
158 /// back - Get the last character in the string.
159 [[nodiscard]] char back() const {
160 assert(!empty());
161 return data()[size() - 1];
162 }
163
164 // copy - Allocate copy in Allocator and return StringRef to it.
165 template <typename Allocator>
166 [[nodiscard]] StringRef copy(Allocator &A) const {
167 // Don't request a length 0 copy from the allocator.
168 if (empty())
169 return StringRef();
170 char *S = A.template Allocate<char>(size());
171 std::copy(begin(), end(), S);
172 return StringRef(S, size());
173 }
174
175 /// Check for string equality, ignoring case.
176 [[nodiscard]] bool equals_insensitive(StringRef RHS) const {
177 return size() == RHS.size() && compare_insensitive(RHS) == 0;
178 }
179
180 /// compare - Compare two strings; the result is negative, zero, or positive
181 /// if this string is lexicographically less than, equal to, or greater than
182 /// the \p RHS.
183 [[nodiscard]] int compare(StringRef RHS) const {
184 // Check the prefix for a mismatch.
185 if (int Res =
186 compareMemory(Lhs: data(), Rhs: RHS.data(), Length: std::min(a: size(), b: RHS.size())))
187 return Res < 0 ? -1 : 1;
188
189 // Otherwise the prefixes match, so we only need to check the lengths.
190 if (size() == RHS.size())
191 return 0;
192 return size() < RHS.size() ? -1 : 1;
193 }
194
195 /// Compare two strings, ignoring case.
196 [[nodiscard]] int compare_insensitive(StringRef RHS) const;
197
198 /// compare_numeric - Compare two strings, treating sequences of digits as
199 /// numbers.
200 [[nodiscard]] int compare_numeric(StringRef RHS) const;
201
202 /// Determine the edit distance between this string and another
203 /// string.
204 ///
205 /// \param Other the string to compare this string against.
206 ///
207 /// \param AllowReplacements whether to allow character
208 /// replacements (change one character into another) as a single
209 /// operation, rather than as two operations (an insertion and a
210 /// removal).
211 ///
212 /// \param MaxEditDistance If non-zero, the maximum edit distance that
213 /// this routine is allowed to compute. If the edit distance will exceed
214 /// that maximum, returns \c MaxEditDistance+1.
215 ///
216 /// \returns the minimum number of character insertions, removals,
217 /// or (if \p AllowReplacements is \c true) replacements needed to
218 /// transform one of the given strings into the other. If zero,
219 /// the strings are identical.
220 [[nodiscard]] unsigned edit_distance(StringRef Other,
221 bool AllowReplacements = true,
222 unsigned MaxEditDistance = 0) const;
223
224 [[nodiscard]] unsigned
225 edit_distance_insensitive(StringRef Other, bool AllowReplacements = true,
226 unsigned MaxEditDistance = 0) const;
227
228 /// str - Get the contents as an std::string.
229 [[nodiscard]] std::string str() const {
230 if (!data())
231 return std::string();
232 return std::string(data(), size());
233 }
234
235 /// @}
236 /// @name Operator Overloads
237 /// @{
238
239 [[nodiscard]] char operator[](size_t Index) const {
240 assert(Index < size() && "Invalid index!");
241 return data()[Index];
242 }
243
244 /// Disallow accidental assignment from a temporary std::string.
245 ///
246 /// The declaration here is extra complicated so that `stringRef = {}`
247 /// and `stringRef = "abc"` continue to select the move assignment operator.
248 template <typename T>
249 std::enable_if_t<std::is_same<T, std::string>::value, StringRef> &
250 operator=(T &&Str) = delete;
251
252 /// @}
253 /// @name Type Conversions
254 /// @{
255
256 constexpr operator std::string_view() const {
257 return std::string_view(data(), size());
258 }
259
260 /// @}
261 /// @name String Predicates
262 /// @{
263
264 /// Check if this string starts with the given \p Prefix.
265 [[nodiscard]] bool starts_with(StringRef Prefix) const {
266 return size() >= Prefix.size() &&
267 compareMemory(Lhs: data(), Rhs: Prefix.data(), Length: Prefix.size()) == 0;
268 }
269 [[nodiscard]] bool starts_with(char Prefix) const {
270 return !empty() && front() == Prefix;
271 }
272
273 /// Check if this string starts with the given \p Prefix, ignoring case.
274 [[nodiscard]] bool starts_with_insensitive(StringRef Prefix) const;
275
276 /// Check if this string ends with the given \p Suffix.
277 [[nodiscard]] bool ends_with(StringRef Suffix) const {
278 return size() >= Suffix.size() &&
279 compareMemory(Lhs: end() - Suffix.size(), Rhs: Suffix.data(),
280 Length: Suffix.size()) == 0;
281 }
282 [[nodiscard]] bool ends_with(char Suffix) const {
283 return !empty() && back() == Suffix;
284 }
285
286 /// Check if this string ends with the given \p Suffix, ignoring case.
287 [[nodiscard]] bool ends_with_insensitive(StringRef Suffix) const;
288
289 /// @}
290 /// @name String Searching
291 /// @{
292
293 /// Search for the first character \p C in the string.
294 ///
295 /// \returns The index of the first occurrence of \p C, or npos if not
296 /// found.
297 [[nodiscard]] size_t find(char C, size_t From = 0) const {
298 return std::string_view(*this).find(c: C, pos: From);
299 }
300
301 /// Search for the first character \p C in the string, ignoring case.
302 ///
303 /// \returns The index of the first occurrence of \p C, or npos if not
304 /// found.
305 [[nodiscard]] size_t find_insensitive(char C, size_t From = 0) const;
306
307 /// Search for the first character satisfying the predicate \p F
308 ///
309 /// \returns The index of the first character satisfying \p F starting from
310 /// \p From, or npos if not found.
311 [[nodiscard]] size_t find_if(function_ref<bool(char)> F,
312 size_t From = 0) const {
313 StringRef S = drop_front(N: From);
314 while (!S.empty()) {
315 if (F(S.front()))
316 return size() - S.size();
317 S = S.drop_front();
318 }
319 return npos;
320 }
321
322 /// Search for the first character not satisfying the predicate \p F
323 ///
324 /// \returns The index of the first character not satisfying \p F starting
325 /// from \p From, or npos if not found.
326 [[nodiscard]] size_t find_if_not(function_ref<bool(char)> F,
327 size_t From = 0) const {
328 return find_if(F: [F](char c) { return !F(c); }, From);
329 }
330
331 /// Search for the first string \p Str in the string.
332 ///
333 /// \returns The index of the first occurrence of \p Str, or npos if not
334 /// found.
335 [[nodiscard]] size_t find(StringRef Str, size_t From = 0) const;
336
337 /// Search for the first string \p Str in the string, ignoring case.
338 ///
339 /// \returns The index of the first occurrence of \p Str, or npos if not
340 /// found.
341 [[nodiscard]] size_t find_insensitive(StringRef Str, size_t From = 0) const;
342
343 /// Search for the last character \p C in the string.
344 ///
345 /// \returns The index of the last occurrence of \p C, or npos if not
346 /// found.
347 [[nodiscard]] size_t rfind(char C, size_t From = npos) const {
348 size_t I = std::min(a: From, b: size());
349 while (I) {
350 --I;
351 if (data()[I] == C)
352 return I;
353 }
354 return npos;
355 }
356
357 /// Search for the last character \p C in the string, ignoring case.
358 ///
359 /// \returns The index of the last occurrence of \p C, or npos if not
360 /// found.
361 [[nodiscard]] size_t rfind_insensitive(char C, size_t From = npos) const;
362
363 /// Search for the last string \p Str in the string.
364 ///
365 /// \returns The index of the last occurrence of \p Str, or npos if not
366 /// found.
367 [[nodiscard]] size_t rfind(StringRef Str) const;
368
369 /// Search for the last string \p Str in the string, ignoring case.
370 ///
371 /// \returns The index of the last occurrence of \p Str, or npos if not
372 /// found.
373 [[nodiscard]] size_t rfind_insensitive(StringRef Str) const;
374
375 /// Find the first character in the string that is \p C, or npos if not
376 /// found. Same as find.
377 [[nodiscard]] size_t find_first_of(char C, size_t From = 0) const {
378 return find(C, From);
379 }
380
381 /// Find the first character in the string that is in \p Chars, or npos if
382 /// not found.
383 ///
384 /// Complexity: O(size() + Chars.size())
385 [[nodiscard]] size_t find_first_of(StringRef Chars, size_t From = 0) const;
386
387 /// Find the first character in the string that is not \p C or npos if not
388 /// found.
389 [[nodiscard]] size_t find_first_not_of(char C, size_t From = 0) const;
390
391 /// Find the first character in the string that is not in the string
392 /// \p Chars, or npos if not found.
393 ///
394 /// Complexity: O(size() + Chars.size())
395 [[nodiscard]] size_t find_first_not_of(StringRef Chars,
396 size_t From = 0) const;
397
398 /// Find the last character in the string that is \p C, or npos if not
399 /// found.
400 [[nodiscard]] size_t find_last_of(char C, size_t From = npos) const {
401 return rfind(C, From);
402 }
403
404 /// Find the last character in the string that is in \p C, or npos if not
405 /// found.
406 ///
407 /// Complexity: O(size() + Chars.size())
408 [[nodiscard]] size_t find_last_of(StringRef Chars,
409 size_t From = npos) const;
410
411 /// Find the last character in the string that is not \p C, or npos if not
412 /// found.
413 [[nodiscard]] size_t find_last_not_of(char C, size_t From = npos) const;
414
415 /// Find the last character in the string that is not in \p Chars, or
416 /// npos if not found.
417 ///
418 /// Complexity: O(size() + Chars.size())
419 [[nodiscard]] size_t find_last_not_of(StringRef Chars,
420 size_t From = npos) const;
421
422 /// Return true if the given string is a substring of *this, and false
423 /// otherwise.
424 [[nodiscard]] bool contains(StringRef Other) const {
425 return find(Str: Other) != npos;
426 }
427
428 /// Return true if the given character is contained in *this, and false
429 /// otherwise.
430 [[nodiscard]] bool contains(char C) const {
431 return find_first_of(C) != npos;
432 }
433
434 /// Return true if the given string is a substring of *this, and false
435 /// otherwise.
436 [[nodiscard]] bool contains_insensitive(StringRef Other) const {
437 return find_insensitive(Str: Other) != npos;
438 }
439
440 /// Return true if the given character is contained in *this, and false
441 /// otherwise.
442 [[nodiscard]] bool contains_insensitive(char C) const {
443 return find_insensitive(C) != npos;
444 }
445
446 /// @}
447 /// @name Helpful Algorithms
448 /// @{
449
450 /// Return the number of occurrences of \p C in the string.
451 [[nodiscard]] size_t count(char C) const {
452 size_t Count = 0;
453 for (size_t I = 0; I != size(); ++I)
454 if (data()[I] == C)
455 ++Count;
456 return Count;
457 }
458
459 /// Return the number of non-overlapped occurrences of \p Str in
460 /// the string.
461 size_t count(StringRef Str) const;
462
463 /// Parse the current string as an integer of the specified radix. If
464 /// \p Radix is specified as zero, this does radix autosensing using
465 /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
466 ///
467 /// If the string is invalid or if only a subset of the string is valid,
468 /// this returns true to signify the error. The string is considered
469 /// erroneous if empty or if it overflows T.
470 template <typename T> bool getAsInteger(unsigned Radix, T &Result) const {
471 if constexpr (std::numeric_limits<T>::is_signed) {
472 long long LLVal;
473 if (getAsSignedInteger(Str: *this, Radix, Result&: LLVal) ||
474 static_cast<T>(LLVal) != LLVal)
475 return true;
476 Result = LLVal;
477 } else {
478 unsigned long long ULLVal;
479 // The additional cast to unsigned long long is required to avoid the
480 // Visual C++ warning C4805: '!=' : unsafe mix of type 'bool' and type
481 // 'unsigned __int64' when instantiating getAsInteger with T = bool.
482 if (getAsUnsignedInteger(Str: *this, Radix, Result&: ULLVal) ||
483 static_cast<unsigned long long>(static_cast<T>(ULLVal)) != ULLVal)
484 return true;
485 Result = ULLVal;
486 }
487 return false;
488 }
489
490 /// Parse the current string as an integer of the specified radix. If
491 /// \p Radix is specified as zero, this does radix autosensing using
492 /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
493 ///
494 /// If the string does not begin with a number of the specified radix,
495 /// this returns true to signify the error. The string is considered
496 /// erroneous if empty or if it overflows T.
497 /// The portion of the string representing the discovered numeric value
498 /// is removed from the beginning of the string.
499 template <typename T> bool consumeInteger(unsigned Radix, T &Result) {
500 if constexpr (std::numeric_limits<T>::is_signed) {
501 long long LLVal;
502 if (consumeSignedInteger(Str&: *this, Radix, Result&: LLVal) ||
503 static_cast<long long>(static_cast<T>(LLVal)) != LLVal)
504 return true;
505 Result = LLVal;
506 } else {
507 unsigned long long ULLVal;
508 if (consumeUnsignedInteger(Str&: *this, Radix, Result&: ULLVal) ||
509 static_cast<unsigned long long>(static_cast<T>(ULLVal)) != ULLVal)
510 return true;
511 Result = ULLVal;
512 }
513 return false;
514 }
515
516 /// Parse the current string as an integer of the specified \p Radix, or of
517 /// an autosensed radix if the \p Radix given is 0. The current value in
518 /// \p Result is discarded, and the storage is changed to be wide enough to
519 /// store the parsed integer.
520 ///
521 /// \returns true if the string does not solely consist of a valid
522 /// non-empty number in the appropriate base.
523 ///
524 /// APInt::fromString is superficially similar but assumes the
525 /// string is well-formed in the given radix.
526 bool getAsInteger(unsigned Radix, APInt &Result) const;
527
528 /// Parse the current string as an integer of the specified \p Radix. If
529 /// \p Radix is specified as zero, this does radix autosensing using
530 /// extended C rules: 0 is octal, 0x is hex, 0b is binary.
531 ///
532 /// If the string does not begin with a number of the specified radix,
533 /// this returns true to signify the error. The string is considered
534 /// erroneous if empty.
535 /// The portion of the string representing the discovered numeric value
536 /// is removed from the beginning of the string.
537 bool consumeInteger(unsigned Radix, APInt &Result);
538
539 /// Parse the current string as an IEEE double-precision floating
540 /// point value. The string must be a well-formed double.
541 ///
542 /// If \p AllowInexact is false, the function will fail if the string
543 /// cannot be represented exactly. Otherwise, the function only fails
544 /// in case of an overflow or underflow, or an invalid floating point
545 /// representation.
546 bool getAsDouble(double &Result, bool AllowInexact = true) const;
547
548 /// @}
549 /// @name String Operations
550 /// @{
551
552 // Convert the given ASCII string to lowercase.
553 [[nodiscard]] std::string lower() const;
554
555 /// Convert the given ASCII string to uppercase.
556 [[nodiscard]] std::string upper() const;
557
558 /// @}
559 /// @name Substring Operations
560 /// @{
561
562 /// Return a reference to the substring from [Start, Start + N).
563 ///
564 /// \param Start The index of the starting character in the substring; if
565 /// the index is npos or greater than the length of the string then the
566 /// empty substring will be returned.
567 ///
568 /// \param N The number of characters to included in the substring. If N
569 /// exceeds the number of characters remaining in the string, the string
570 /// suffix (starting with \p Start) will be returned.
571 [[nodiscard]] constexpr StringRef substr(size_t Start,
572 size_t N = npos) const {
573 Start = std::min(a: Start, b: size());
574 return StringRef(data() + Start, std::min(a: N, b: size() - Start));
575 }
576
577 /// Return a StringRef equal to 'this' but with only the first \p N
578 /// elements remaining. If \p N is greater than the length of the
579 /// string, the entire string is returned.
580 [[nodiscard]] StringRef take_front(size_t N = 1) const {
581 if (N >= size())
582 return *this;
583 return drop_back(N: size() - N);
584 }
585
586 /// Return a StringRef equal to 'this' but with only the last \p N
587 /// elements remaining. If \p N is greater than the length of the
588 /// string, the entire string is returned.
589 [[nodiscard]] StringRef take_back(size_t N = 1) const {
590 if (N >= size())
591 return *this;
592 return drop_front(N: size() - N);
593 }
594
595 /// Return the longest prefix of 'this' such that every character
596 /// in the prefix satisfies the given predicate.
597 [[nodiscard]] StringRef take_while(function_ref<bool(char)> F) const {
598 return substr(Start: 0, N: find_if_not(F));
599 }
600
601 /// Return the longest prefix of 'this' such that no character in
602 /// the prefix satisfies the given predicate.
603 [[nodiscard]] StringRef take_until(function_ref<bool(char)> F) const {
604 return substr(Start: 0, N: find_if(F));
605 }
606
607 /// Return a StringRef equal to 'this' but with the first \p N elements
608 /// dropped.
609 [[nodiscard]] StringRef drop_front(size_t N = 1) const {
610 assert(size() >= N && "Dropping more elements than exist");
611 return substr(Start: N);
612 }
613
614 /// Return a StringRef equal to 'this' but with the last \p N elements
615 /// dropped.
616 [[nodiscard]] StringRef drop_back(size_t N = 1) const {
617 assert(size() >= N && "Dropping more elements than exist");
618 return substr(Start: 0, N: size()-N);
619 }
620
621 /// Return a StringRef equal to 'this', but with all characters satisfying
622 /// the given predicate dropped from the beginning of the string.
623 [[nodiscard]] StringRef drop_while(function_ref<bool(char)> F) const {
624 return substr(Start: find_if_not(F));
625 }
626
627 /// Return a StringRef equal to 'this', but with all characters not
628 /// satisfying the given predicate dropped from the beginning of the string.
629 [[nodiscard]] StringRef drop_until(function_ref<bool(char)> F) const {
630 return substr(Start: find_if(F));
631 }
632
633 /// Returns true if this StringRef has the given prefix and removes that
634 /// prefix.
635 bool consume_front(StringRef Prefix) {
636 if (!starts_with(Prefix))
637 return false;
638
639 *this = substr(Start: Prefix.size());
640 return true;
641 }
642
643 /// Returns true if this StringRef has the given prefix, ignoring case,
644 /// and removes that prefix.
645 bool consume_front_insensitive(StringRef Prefix) {
646 if (!starts_with_insensitive(Prefix))
647 return false;
648
649 *this = substr(Start: Prefix.size());
650 return true;
651 }
652
653 /// Returns true if this StringRef has the given suffix and removes that
654 /// suffix.
655 bool consume_back(StringRef Suffix) {
656 if (!ends_with(Suffix))
657 return false;
658
659 *this = substr(Start: 0, N: size() - Suffix.size());
660 return true;
661 }
662
663 /// Returns true if this StringRef has the given suffix, ignoring case,
664 /// and removes that suffix.
665 bool consume_back_insensitive(StringRef Suffix) {
666 if (!ends_with_insensitive(Suffix))
667 return false;
668
669 *this = substr(Start: 0, N: size() - Suffix.size());
670 return true;
671 }
672
673 /// Return a reference to the substring from [Start, End).
674 ///
675 /// \param Start The index of the starting character in the substring; if
676 /// the index is npos or greater than the length of the string then the
677 /// empty substring will be returned.
678 ///
679 /// \param End The index following the last character to include in the
680 /// substring. If this is npos or exceeds the number of characters
681 /// remaining in the string, the string suffix (starting with \p Start)
682 /// will be returned. If this is less than \p Start, an empty string will
683 /// be returned.
684 [[nodiscard]] StringRef slice(size_t Start, size_t End) const {
685 Start = std::min(a: Start, b: size());
686 End = std::clamp(val: End, lo: Start, hi: size());
687 return StringRef(data() + Start, End - Start);
688 }
689
690 /// Split into two substrings around the first occurrence of a separator
691 /// character.
692 ///
693 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
694 /// such that (*this == LHS + Separator + RHS) is true and RHS is
695 /// maximal. If \p Separator is not in the string, then the result is a
696 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
697 ///
698 /// \param Separator The character to split on.
699 /// \returns The split substrings.
700 [[nodiscard]] std::pair<StringRef, StringRef> split(char Separator) const {
701 return split(Separator: StringRef(&Separator, 1));
702 }
703
704 /// Split into two substrings around the first occurrence of a separator
705 /// string.
706 ///
707 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
708 /// such that (*this == LHS + Separator + RHS) is true and RHS is
709 /// maximal. If \p Separator is not in the string, then the result is a
710 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
711 ///
712 /// \param Separator - The string to split on.
713 /// \return - The split substrings.
714 [[nodiscard]] std::pair<StringRef, StringRef>
715 split(StringRef Separator) const {
716 size_t Idx = find(Str: Separator);
717 if (Idx == npos)
718 return std::make_pair(x: *this, y: StringRef());
719 return std::make_pair(x: slice(Start: 0, End: Idx), y: substr(Start: Idx + Separator.size()));
720 }
721
722 /// Split into two substrings around the last occurrence of a separator
723 /// string.
724 ///
725 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
726 /// such that (*this == LHS + Separator + RHS) is true and RHS is
727 /// minimal. If \p Separator is not in the string, then the result is a
728 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
729 ///
730 /// \param Separator - The string to split on.
731 /// \return - The split substrings.
732 [[nodiscard]] std::pair<StringRef, StringRef>
733 rsplit(StringRef Separator) const {
734 size_t Idx = rfind(Str: Separator);
735 if (Idx == npos)
736 return std::make_pair(x: *this, y: StringRef());
737 return std::make_pair(x: slice(Start: 0, End: Idx), y: substr(Start: Idx + Separator.size()));
738 }
739
740 /// Split into substrings around the occurrences of a separator string.
741 ///
742 /// Each substring is stored in \p A. If \p MaxSplit is >= 0, at most
743 /// \p MaxSplit splits are done and consequently <= \p MaxSplit + 1
744 /// elements are added to A.
745 /// If \p KeepEmpty is false, empty strings are not added to \p A. They
746 /// still count when considering \p MaxSplit
747 /// An useful invariant is that
748 /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true
749 ///
750 /// \param A - Where to put the substrings.
751 /// \param Separator - The string to split on.
752 /// \param MaxSplit - The maximum number of times the string is split.
753 /// \param KeepEmpty - True if empty substring should be added.
754 void split(SmallVectorImpl<StringRef> &A,
755 StringRef Separator, int MaxSplit = -1,
756 bool KeepEmpty = true) const;
757
758 /// Split into substrings around the occurrences of a separator character.
759 ///
760 /// Each substring is stored in \p A. If \p MaxSplit is >= 0, at most
761 /// \p MaxSplit splits are done and consequently <= \p MaxSplit + 1
762 /// elements are added to A.
763 /// If \p KeepEmpty is false, empty strings are not added to \p A. They
764 /// still count when considering \p MaxSplit
765 /// An useful invariant is that
766 /// Separator.join(A) == *this if MaxSplit == -1 and KeepEmpty == true
767 ///
768 /// \param A - Where to put the substrings.
769 /// \param Separator - The string to split on.
770 /// \param MaxSplit - The maximum number of times the string is split.
771 /// \param KeepEmpty - True if empty substring should be added.
772 void split(SmallVectorImpl<StringRef> &A, char Separator, int MaxSplit = -1,
773 bool KeepEmpty = true) const;
774
775 /// Split into two substrings around the last occurrence of a separator
776 /// character.
777 ///
778 /// If \p Separator is in the string, then the result is a pair (LHS, RHS)
779 /// such that (*this == LHS + Separator + RHS) is true and RHS is
780 /// minimal. If \p Separator is not in the string, then the result is a
781 /// pair (LHS, RHS) where (*this == LHS) and (RHS == "").
782 ///
783 /// \param Separator - The character to split on.
784 /// \return - The split substrings.
785 [[nodiscard]] std::pair<StringRef, StringRef> rsplit(char Separator) const {
786 return rsplit(Separator: StringRef(&Separator, 1));
787 }
788
789 /// Return string with consecutive \p Char characters starting from the
790 /// the left removed.
791 [[nodiscard]] StringRef ltrim(char Char) const {
792 return drop_front(N: std::min(a: size(), b: find_first_not_of(C: Char)));
793 }
794
795 /// Return string with consecutive characters in \p Chars starting from
796 /// the left removed.
797 [[nodiscard]] StringRef ltrim(StringRef Chars = " \t\n\v\f\r") const {
798 return drop_front(N: std::min(a: size(), b: find_first_not_of(Chars)));
799 }
800
801 /// Return string with consecutive \p Char characters starting from the
802 /// right removed.
803 [[nodiscard]] StringRef rtrim(char Char) const {
804 return drop_back(N: size() - std::min(a: size(), b: find_last_not_of(C: Char) + 1));
805 }
806
807 /// Return string with consecutive characters in \p Chars starting from
808 /// the right removed.
809 [[nodiscard]] StringRef rtrim(StringRef Chars = " \t\n\v\f\r") const {
810 return drop_back(N: size() - std::min(a: size(), b: find_last_not_of(Chars) + 1));
811 }
812
813 /// Return string with consecutive \p Char characters starting from the
814 /// left and right removed.
815 [[nodiscard]] StringRef trim(char Char) const {
816 return ltrim(Char).rtrim(Char);
817 }
818
819 /// Return string with consecutive characters in \p Chars starting from
820 /// the left and right removed.
821 [[nodiscard]] StringRef trim(StringRef Chars = " \t\n\v\f\r") const {
822 return ltrim(Chars).rtrim(Chars);
823 }
824
825 /// Detect the line ending style of the string.
826 ///
827 /// If the string contains a line ending, return the line ending character
828 /// sequence that is detected. Otherwise return '\n' for unix line endings.
829 ///
830 /// \return - The line ending character sequence.
831 [[nodiscard]] StringRef detectEOL() const {
832 size_t Pos = find(C: '\r');
833 if (Pos == npos) {
834 // If there is no carriage return, assume unix
835 return "\n";
836 }
837 if (Pos + 1 < size() && data()[Pos + 1] == '\n')
838 return "\r\n"; // Windows
839 if (Pos > 0 && data()[Pos - 1] == '\n')
840 return "\n\r"; // You monster!
841 return "\r"; // Classic Mac
842 }
843 /// @}
844 };
845
846 /// A wrapper around a string literal that serves as a proxy for constructing
847 /// global tables of StringRefs with the length computed at compile time.
848 /// In order to avoid the invocation of a global constructor, StringLiteral
849 /// should *only* be used in a constexpr context, as such:
850 ///
851 /// constexpr StringLiteral S("test");
852 ///
853 class StringLiteral : public StringRef {
854 private:
855 constexpr StringLiteral(const char *Str, size_t N) : StringRef(Str, N) {
856 }
857
858 public:
859 template <size_t N>
860 constexpr StringLiteral(const char (&Str)[N])
861#if defined(__clang__) && __has_attribute(enable_if)
862#pragma clang diagnostic push
863#pragma clang diagnostic ignored "-Wgcc-compat"
864 __attribute((enable_if(__builtin_strlen(Str) == N - 1,
865 "invalid string literal")))
866#pragma clang diagnostic pop
867#endif
868 : StringRef(Str, N - 1) {
869 }
870
871 // Explicit construction for strings like "foo\0bar".
872 template <size_t N>
873 static constexpr StringLiteral withInnerNUL(const char (&Str)[N]) {
874 return StringLiteral(Str, N - 1);
875 }
876 };
877
878 /// @name StringRef Comparison Operators
879 /// @{
880
881 inline bool operator==(StringRef LHS, StringRef RHS) {
882 if (LHS.size() != RHS.size())
883 return false;
884 if (LHS.empty())
885 return true;
886 return ::memcmp(s1: LHS.data(), s2: RHS.data(), n: LHS.size()) == 0;
887 }
888
889 inline bool operator!=(StringRef LHS, StringRef RHS) { return !(LHS == RHS); }
890
891 inline bool operator<(StringRef LHS, StringRef RHS) {
892 return LHS.compare(RHS) < 0;
893 }
894
895 inline bool operator<=(StringRef LHS, StringRef RHS) {
896 return LHS.compare(RHS) <= 0;
897 }
898
899 inline bool operator>(StringRef LHS, StringRef RHS) {
900 return LHS.compare(RHS) > 0;
901 }
902
903 inline bool operator>=(StringRef LHS, StringRef RHS) {
904 return LHS.compare(RHS) >= 0;
905 }
906
907 inline std::string &operator+=(std::string &buffer, StringRef string) {
908 return buffer.append(s: string.data(), n: string.size());
909 }
910
911 /// @}
912
913 /// Compute a hash_code for a StringRef.
914 [[nodiscard]] hash_code hash_value(StringRef S);
915
916 // Provide DenseMapInfo for StringRefs.
917 template <> struct DenseMapInfo<StringRef, void> {
918 static inline StringRef getEmptyKey() {
919 return StringRef(
920 reinterpret_cast<const char *>(~static_cast<uintptr_t>(0)), 0);
921 }
922
923 static inline StringRef getTombstoneKey() {
924 return StringRef(
925 reinterpret_cast<const char *>(~static_cast<uintptr_t>(1)), 0);
926 }
927
928 static unsigned getHashValue(StringRef Val);
929
930 static bool isEqual(StringRef LHS, StringRef RHS) {
931 if (RHS.data() == getEmptyKey().data())
932 return LHS.data() == getEmptyKey().data();
933 if (RHS.data() == getTombstoneKey().data())
934 return LHS.data() == getTombstoneKey().data();
935 return LHS == RHS;
936 }
937 };
938
939} // end namespace llvm
940
941#endif // LLVM_ADT_STRINGREF_H
942