xref: /freebsd/contrib/llvm-project/llvm/lib/Support/StringRef.cpp (revision 924226fba12cc9a228c73b956e1b7fa24c60b055)
1 //===-- StringRef.cpp - Lightweight String References ---------------------===//
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 #include "llvm/ADT/StringRef.h"
10 #include "llvm/ADT/APFloat.h"
11 #include "llvm/ADT/APInt.h"
12 #include "llvm/ADT/Hashing.h"
13 #include "llvm/ADT/StringExtras.h"
14 #include "llvm/ADT/edit_distance.h"
15 #include "llvm/Support/Error.h"
16 #include <bitset>
17 
18 using namespace llvm;
19 
20 // MSVC emits references to this into the translation units which reference it.
21 #ifndef _MSC_VER
22 constexpr size_t StringRef::npos;
23 #endif
24 
25 // strncasecmp() is not available on non-POSIX systems, so define an
26 // alternative function here.
27 static int ascii_strncasecmp(const char *LHS, const char *RHS, size_t Length) {
28   for (size_t I = 0; I < Length; ++I) {
29     unsigned char LHC = toLower(LHS[I]);
30     unsigned char RHC = toLower(RHS[I]);
31     if (LHC != RHC)
32       return LHC < RHC ? -1 : 1;
33   }
34   return 0;
35 }
36 
37 int StringRef::compare_insensitive(StringRef RHS) const {
38   if (int Res = ascii_strncasecmp(Data, RHS.Data, std::min(Length, RHS.Length)))
39     return Res;
40   if (Length == RHS.Length)
41     return 0;
42   return Length < RHS.Length ? -1 : 1;
43 }
44 
45 bool StringRef::startswith_insensitive(StringRef Prefix) const {
46   return Length >= Prefix.Length &&
47       ascii_strncasecmp(Data, Prefix.Data, Prefix.Length) == 0;
48 }
49 
50 bool StringRef::endswith_insensitive(StringRef Suffix) const {
51   return Length >= Suffix.Length &&
52       ascii_strncasecmp(end() - Suffix.Length, Suffix.Data, Suffix.Length) == 0;
53 }
54 
55 size_t StringRef::find_insensitive(char C, size_t From) const {
56   char L = toLower(C);
57   return find_if([L](char D) { return toLower(D) == L; }, From);
58 }
59 
60 /// compare_numeric - Compare strings, handle embedded numbers.
61 int StringRef::compare_numeric(StringRef RHS) const {
62   for (size_t I = 0, E = std::min(Length, RHS.Length); I != E; ++I) {
63     // Check for sequences of digits.
64     if (isDigit(Data[I]) && isDigit(RHS.Data[I])) {
65       // The longer sequence of numbers is considered larger.
66       // This doesn't really handle prefixed zeros well.
67       size_t J;
68       for (J = I + 1; J != E + 1; ++J) {
69         bool ld = J < Length && isDigit(Data[J]);
70         bool rd = J < RHS.Length && isDigit(RHS.Data[J]);
71         if (ld != rd)
72           return rd ? -1 : 1;
73         if (!rd)
74           break;
75       }
76       // The two number sequences have the same length (J-I), just memcmp them.
77       if (int Res = compareMemory(Data + I, RHS.Data + I, J - I))
78         return Res < 0 ? -1 : 1;
79       // Identical number sequences, continue search after the numbers.
80       I = J - 1;
81       continue;
82     }
83     if (Data[I] != RHS.Data[I])
84       return (unsigned char)Data[I] < (unsigned char)RHS.Data[I] ? -1 : 1;
85   }
86   if (Length == RHS.Length)
87     return 0;
88   return Length < RHS.Length ? -1 : 1;
89 }
90 
91 // Compute the edit distance between the two given strings.
92 unsigned StringRef::edit_distance(llvm::StringRef Other,
93                                   bool AllowReplacements,
94                                   unsigned MaxEditDistance) const {
95   return llvm::ComputeEditDistance(
96       makeArrayRef(data(), size()),
97       makeArrayRef(Other.data(), Other.size()),
98       AllowReplacements, MaxEditDistance);
99 }
100 
101 //===----------------------------------------------------------------------===//
102 // String Operations
103 //===----------------------------------------------------------------------===//
104 
105 std::string StringRef::lower() const {
106   return std::string(map_iterator(begin(), toLower),
107                      map_iterator(end(), toLower));
108 }
109 
110 std::string StringRef::upper() const {
111   return std::string(map_iterator(begin(), toUpper),
112                      map_iterator(end(), toUpper));
113 }
114 
115 //===----------------------------------------------------------------------===//
116 // String Searching
117 //===----------------------------------------------------------------------===//
118 
119 
120 /// find - Search for the first string \arg Str in the string.
121 ///
122 /// \return - The index of the first occurrence of \arg Str, or npos if not
123 /// found.
124 size_t StringRef::find(StringRef Str, size_t From) const {
125   if (From > Length)
126     return npos;
127 
128   const char *Start = Data + From;
129   size_t Size = Length - From;
130 
131   const char *Needle = Str.data();
132   size_t N = Str.size();
133   if (N == 0)
134     return From;
135   if (Size < N)
136     return npos;
137   if (N == 1) {
138     const char *Ptr = (const char *)::memchr(Start, Needle[0], Size);
139     return Ptr == nullptr ? npos : Ptr - Data;
140   }
141 
142   const char *Stop = Start + (Size - N + 1);
143 
144   // For short haystacks or unsupported needles fall back to the naive algorithm
145   if (Size < 16 || N > 255) {
146     do {
147       if (std::memcmp(Start, Needle, N) == 0)
148         return Start - Data;
149       ++Start;
150     } while (Start < Stop);
151     return npos;
152   }
153 
154   // Build the bad char heuristic table, with uint8_t to reduce cache thrashing.
155   uint8_t BadCharSkip[256];
156   std::memset(BadCharSkip, N, 256);
157   for (unsigned i = 0; i != N-1; ++i)
158     BadCharSkip[(uint8_t)Str[i]] = N-1-i;
159 
160   do {
161     uint8_t Last = Start[N - 1];
162     if (LLVM_UNLIKELY(Last == (uint8_t)Needle[N - 1]))
163       if (std::memcmp(Start, Needle, N - 1) == 0)
164         return Start - Data;
165 
166     // Otherwise skip the appropriate number of bytes.
167     Start += BadCharSkip[Last];
168   } while (Start < Stop);
169 
170   return npos;
171 }
172 
173 size_t StringRef::find_insensitive(StringRef Str, size_t From) const {
174   StringRef This = substr(From);
175   while (This.size() >= Str.size()) {
176     if (This.startswith_insensitive(Str))
177       return From;
178     This = This.drop_front();
179     ++From;
180   }
181   return npos;
182 }
183 
184 size_t StringRef::rfind_insensitive(char C, size_t From) const {
185   From = std::min(From, Length);
186   size_t i = From;
187   while (i != 0) {
188     --i;
189     if (toLower(Data[i]) == toLower(C))
190       return i;
191   }
192   return npos;
193 }
194 
195 /// rfind - Search for the last string \arg Str in the string.
196 ///
197 /// \return - The index of the last occurrence of \arg Str, or npos if not
198 /// found.
199 size_t StringRef::rfind(StringRef Str) const {
200   size_t N = Str.size();
201   if (N > Length)
202     return npos;
203   for (size_t i = Length - N + 1, e = 0; i != e;) {
204     --i;
205     if (substr(i, N).equals(Str))
206       return i;
207   }
208   return npos;
209 }
210 
211 size_t StringRef::rfind_insensitive(StringRef Str) const {
212   size_t N = Str.size();
213   if (N > Length)
214     return npos;
215   for (size_t i = Length - N + 1, e = 0; i != e;) {
216     --i;
217     if (substr(i, N).equals_insensitive(Str))
218       return i;
219   }
220   return npos;
221 }
222 
223 /// find_first_of - Find the first character in the string that is in \arg
224 /// Chars, or npos if not found.
225 ///
226 /// Note: O(size() + Chars.size())
227 StringRef::size_type StringRef::find_first_of(StringRef Chars,
228                                               size_t From) const {
229   std::bitset<1 << CHAR_BIT> CharBits;
230   for (char C : Chars)
231     CharBits.set((unsigned char)C);
232 
233   for (size_type i = std::min(From, Length), e = Length; i != e; ++i)
234     if (CharBits.test((unsigned char)Data[i]))
235       return i;
236   return npos;
237 }
238 
239 /// find_first_not_of - Find the first character in the string that is not
240 /// \arg C or npos if not found.
241 StringRef::size_type StringRef::find_first_not_of(char C, size_t From) const {
242   for (size_type i = std::min(From, Length), e = Length; i != e; ++i)
243     if (Data[i] != C)
244       return i;
245   return npos;
246 }
247 
248 /// find_first_not_of - Find the first character in the string that is not
249 /// in the string \arg Chars, or npos if not found.
250 ///
251 /// Note: O(size() + Chars.size())
252 StringRef::size_type StringRef::find_first_not_of(StringRef Chars,
253                                                   size_t From) const {
254   std::bitset<1 << CHAR_BIT> CharBits;
255   for (char C : Chars)
256     CharBits.set((unsigned char)C);
257 
258   for (size_type i = std::min(From, Length), e = Length; i != e; ++i)
259     if (!CharBits.test((unsigned char)Data[i]))
260       return i;
261   return npos;
262 }
263 
264 /// find_last_of - Find the last character in the string that is in \arg C,
265 /// or npos if not found.
266 ///
267 /// Note: O(size() + Chars.size())
268 StringRef::size_type StringRef::find_last_of(StringRef Chars,
269                                              size_t From) const {
270   std::bitset<1 << CHAR_BIT> CharBits;
271   for (char C : Chars)
272     CharBits.set((unsigned char)C);
273 
274   for (size_type i = std::min(From, Length) - 1, e = -1; i != e; --i)
275     if (CharBits.test((unsigned char)Data[i]))
276       return i;
277   return npos;
278 }
279 
280 /// find_last_not_of - Find the last character in the string that is not
281 /// \arg C, or npos if not found.
282 StringRef::size_type StringRef::find_last_not_of(char C, size_t From) const {
283   for (size_type i = std::min(From, Length) - 1, e = -1; i != e; --i)
284     if (Data[i] != C)
285       return i;
286   return npos;
287 }
288 
289 /// find_last_not_of - Find the last character in the string that is not in
290 /// \arg Chars, or npos if not found.
291 ///
292 /// Note: O(size() + Chars.size())
293 StringRef::size_type StringRef::find_last_not_of(StringRef Chars,
294                                                  size_t From) const {
295   std::bitset<1 << CHAR_BIT> CharBits;
296   for (char C : Chars)
297     CharBits.set((unsigned char)C);
298 
299   for (size_type i = std::min(From, Length) - 1, e = -1; i != e; --i)
300     if (!CharBits.test((unsigned char)Data[i]))
301       return i;
302   return npos;
303 }
304 
305 void StringRef::split(SmallVectorImpl<StringRef> &A,
306                       StringRef Separator, int MaxSplit,
307                       bool KeepEmpty) const {
308   StringRef S = *this;
309 
310   // Count down from MaxSplit. When MaxSplit is -1, this will just split
311   // "forever". This doesn't support splitting more than 2^31 times
312   // intentionally; if we ever want that we can make MaxSplit a 64-bit integer
313   // but that seems unlikely to be useful.
314   while (MaxSplit-- != 0) {
315     size_t Idx = S.find(Separator);
316     if (Idx == npos)
317       break;
318 
319     // Push this split.
320     if (KeepEmpty || Idx > 0)
321       A.push_back(S.slice(0, Idx));
322 
323     // Jump forward.
324     S = S.slice(Idx + Separator.size(), npos);
325   }
326 
327   // Push the tail.
328   if (KeepEmpty || !S.empty())
329     A.push_back(S);
330 }
331 
332 void StringRef::split(SmallVectorImpl<StringRef> &A, char Separator,
333                       int MaxSplit, bool KeepEmpty) const {
334   StringRef S = *this;
335 
336   // Count down from MaxSplit. When MaxSplit is -1, this will just split
337   // "forever". This doesn't support splitting more than 2^31 times
338   // intentionally; if we ever want that we can make MaxSplit a 64-bit integer
339   // but that seems unlikely to be useful.
340   while (MaxSplit-- != 0) {
341     size_t Idx = S.find(Separator);
342     if (Idx == npos)
343       break;
344 
345     // Push this split.
346     if (KeepEmpty || Idx > 0)
347       A.push_back(S.slice(0, Idx));
348 
349     // Jump forward.
350     S = S.slice(Idx + 1, npos);
351   }
352 
353   // Push the tail.
354   if (KeepEmpty || !S.empty())
355     A.push_back(S);
356 }
357 
358 //===----------------------------------------------------------------------===//
359 // Helpful Algorithms
360 //===----------------------------------------------------------------------===//
361 
362 /// count - Return the number of non-overlapped occurrences of \arg Str in
363 /// the string.
364 size_t StringRef::count(StringRef Str) const {
365   size_t Count = 0;
366   size_t N = Str.size();
367   if (!N || N > Length)
368     return 0;
369   for (size_t i = 0, e = Length - N + 1; i < e;) {
370     if (substr(i, N).equals(Str)) {
371       ++Count;
372       i += N;
373     }
374     else
375       ++i;
376   }
377   return Count;
378 }
379 
380 static unsigned GetAutoSenseRadix(StringRef &Str) {
381   if (Str.empty())
382     return 10;
383 
384   if (Str.startswith("0x") || Str.startswith("0X")) {
385     Str = Str.substr(2);
386     return 16;
387   }
388 
389   if (Str.startswith("0b") || Str.startswith("0B")) {
390     Str = Str.substr(2);
391     return 2;
392   }
393 
394   if (Str.startswith("0o")) {
395     Str = Str.substr(2);
396     return 8;
397   }
398 
399   if (Str[0] == '0' && Str.size() > 1 && isDigit(Str[1])) {
400     Str = Str.substr(1);
401     return 8;
402   }
403 
404   return 10;
405 }
406 
407 bool llvm::consumeUnsignedInteger(StringRef &Str, unsigned Radix,
408                                   unsigned long long &Result) {
409   // Autosense radix if not specified.
410   if (Radix == 0)
411     Radix = GetAutoSenseRadix(Str);
412 
413   // Empty strings (after the radix autosense) are invalid.
414   if (Str.empty()) return true;
415 
416   // Parse all the bytes of the string given this radix.  Watch for overflow.
417   StringRef Str2 = Str;
418   Result = 0;
419   while (!Str2.empty()) {
420     unsigned CharVal;
421     if (Str2[0] >= '0' && Str2[0] <= '9')
422       CharVal = Str2[0] - '0';
423     else if (Str2[0] >= 'a' && Str2[0] <= 'z')
424       CharVal = Str2[0] - 'a' + 10;
425     else if (Str2[0] >= 'A' && Str2[0] <= 'Z')
426       CharVal = Str2[0] - 'A' + 10;
427     else
428       break;
429 
430     // If the parsed value is larger than the integer radix, we cannot
431     // consume any more characters.
432     if (CharVal >= Radix)
433       break;
434 
435     // Add in this character.
436     unsigned long long PrevResult = Result;
437     Result = Result * Radix + CharVal;
438 
439     // Check for overflow by shifting back and seeing if bits were lost.
440     if (Result / Radix < PrevResult)
441       return true;
442 
443     Str2 = Str2.substr(1);
444   }
445 
446   // We consider the operation a failure if no characters were consumed
447   // successfully.
448   if (Str.size() == Str2.size())
449     return true;
450 
451   Str = Str2;
452   return false;
453 }
454 
455 bool llvm::consumeSignedInteger(StringRef &Str, unsigned Radix,
456                                 long long &Result) {
457   unsigned long long ULLVal;
458 
459   // Handle positive strings first.
460   if (Str.empty() || Str.front() != '-') {
461     if (consumeUnsignedInteger(Str, Radix, ULLVal) ||
462         // Check for value so large it overflows a signed value.
463         (long long)ULLVal < 0)
464       return true;
465     Result = ULLVal;
466     return false;
467   }
468 
469   // Get the positive part of the value.
470   StringRef Str2 = Str.drop_front(1);
471   if (consumeUnsignedInteger(Str2, Radix, ULLVal) ||
472       // Reject values so large they'd overflow as negative signed, but allow
473       // "-0".  This negates the unsigned so that the negative isn't undefined
474       // on signed overflow.
475       (long long)-ULLVal > 0)
476     return true;
477 
478   Str = Str2;
479   Result = -ULLVal;
480   return false;
481 }
482 
483 /// GetAsUnsignedInteger - Workhorse method that converts a integer character
484 /// sequence of radix up to 36 to an unsigned long long value.
485 bool llvm::getAsUnsignedInteger(StringRef Str, unsigned Radix,
486                                 unsigned long long &Result) {
487   if (consumeUnsignedInteger(Str, Radix, Result))
488     return true;
489 
490   // For getAsUnsignedInteger, we require the whole string to be consumed or
491   // else we consider it a failure.
492   return !Str.empty();
493 }
494 
495 bool llvm::getAsSignedInteger(StringRef Str, unsigned Radix,
496                               long long &Result) {
497   if (consumeSignedInteger(Str, Radix, Result))
498     return true;
499 
500   // For getAsSignedInteger, we require the whole string to be consumed or else
501   // we consider it a failure.
502   return !Str.empty();
503 }
504 
505 bool StringRef::getAsInteger(unsigned Radix, APInt &Result) const {
506   StringRef Str = *this;
507 
508   // Autosense radix if not specified.
509   if (Radix == 0)
510     Radix = GetAutoSenseRadix(Str);
511 
512   assert(Radix > 1 && Radix <= 36);
513 
514   // Empty strings (after the radix autosense) are invalid.
515   if (Str.empty()) return true;
516 
517   // Skip leading zeroes.  This can be a significant improvement if
518   // it means we don't need > 64 bits.
519   while (!Str.empty() && Str.front() == '0')
520     Str = Str.substr(1);
521 
522   // If it was nothing but zeroes....
523   if (Str.empty()) {
524     Result = APInt(64, 0);
525     return false;
526   }
527 
528   // (Over-)estimate the required number of bits.
529   unsigned Log2Radix = 0;
530   while ((1U << Log2Radix) < Radix) Log2Radix++;
531   bool IsPowerOf2Radix = ((1U << Log2Radix) == Radix);
532 
533   unsigned BitWidth = Log2Radix * Str.size();
534   if (BitWidth < Result.getBitWidth())
535     BitWidth = Result.getBitWidth(); // don't shrink the result
536   else if (BitWidth > Result.getBitWidth())
537     Result = Result.zext(BitWidth);
538 
539   APInt RadixAP, CharAP; // unused unless !IsPowerOf2Radix
540   if (!IsPowerOf2Radix) {
541     // These must have the same bit-width as Result.
542     RadixAP = APInt(BitWidth, Radix);
543     CharAP = APInt(BitWidth, 0);
544   }
545 
546   // Parse all the bytes of the string given this radix.
547   Result = 0;
548   while (!Str.empty()) {
549     unsigned CharVal;
550     if (Str[0] >= '0' && Str[0] <= '9')
551       CharVal = Str[0]-'0';
552     else if (Str[0] >= 'a' && Str[0] <= 'z')
553       CharVal = Str[0]-'a'+10;
554     else if (Str[0] >= 'A' && Str[0] <= 'Z')
555       CharVal = Str[0]-'A'+10;
556     else
557       return true;
558 
559     // If the parsed value is larger than the integer radix, the string is
560     // invalid.
561     if (CharVal >= Radix)
562       return true;
563 
564     // Add in this character.
565     if (IsPowerOf2Radix) {
566       Result <<= Log2Radix;
567       Result |= CharVal;
568     } else {
569       Result *= RadixAP;
570       CharAP = CharVal;
571       Result += CharAP;
572     }
573 
574     Str = Str.substr(1);
575   }
576 
577   return false;
578 }
579 
580 bool StringRef::getAsDouble(double &Result, bool AllowInexact) const {
581   APFloat F(0.0);
582   auto StatusOrErr = F.convertFromString(*this, APFloat::rmNearestTiesToEven);
583   if (errorToBool(StatusOrErr.takeError()))
584     return true;
585 
586   APFloat::opStatus Status = *StatusOrErr;
587   if (Status != APFloat::opOK) {
588     if (!AllowInexact || !(Status & APFloat::opInexact))
589       return true;
590   }
591 
592   Result = F.convertToDouble();
593   return false;
594 }
595 
596 // Implementation of StringRef hashing.
597 hash_code llvm::hash_value(StringRef S) {
598   return hash_combine_range(S.begin(), S.end());
599 }
600 
601 unsigned DenseMapInfo<StringRef, void>::getHashValue(StringRef Val) {
602   assert(Val.data() != getEmptyKey().data() &&
603          "Cannot hash the empty key!");
604   assert(Val.data() != getTombstoneKey().data() &&
605          "Cannot hash the tombstone key!");
606   return (unsigned)(hash_value(Val));
607 }
608