StringExtras.h revision 7462b5dd99163785480a11a7c999569b349a137c
1//===-- llvm/ADT/StringExtras.h - Useful string functions -------*- 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// This file contains some functions that are useful when dealing with strings. 11// 12//===----------------------------------------------------------------------===// 13 14#ifndef LLVM_ADT_STRINGEXTRAS_H 15#define LLVM_ADT_STRINGEXTRAS_H 16 17#include "llvm/System/DataTypes.h" 18#include "llvm/ADT/APFloat.h" 19#include "llvm/ADT/StringRef.h" 20#include <cctype> 21#include <cstdio> 22#include <string> 23#include <vector> 24 25namespace llvm { 26template<typename T> class SmallVectorImpl; 27 28/// hexdigit - Return the (uppercase) hexadecimal character for the 29/// given number \arg X (which should be less than 16). 30static inline char hexdigit(unsigned X) { 31 return X < 10 ? '0' + X : 'A' + X - 10; 32} 33 34/// utohex_buffer - Emit the specified number into the buffer specified by 35/// BufferEnd, returning a pointer to the start of the string. This can be used 36/// like this: (note that the buffer must be large enough to handle any number): 37/// char Buffer[40]; 38/// printf("0x%s", utohex_buffer(X, Buffer+40)); 39/// 40/// This should only be used with unsigned types. 41/// 42template<typename IntTy> 43static inline char *utohex_buffer(IntTy X, char *BufferEnd) { 44 char *BufPtr = BufferEnd; 45 *--BufPtr = 0; // Null terminate buffer. 46 if (X == 0) { 47 *--BufPtr = '0'; // Handle special case. 48 return BufPtr; 49 } 50 51 while (X) { 52 unsigned char Mod = static_cast<unsigned char>(X) & 15; 53 *--BufPtr = hexdigit(Mod); 54 X >>= 4; 55 } 56 return BufPtr; 57} 58 59static inline std::string utohexstr(uint64_t X) { 60 char Buffer[40]; 61 return utohex_buffer(X, Buffer+40); 62} 63 64static inline std::string utostr_32(uint32_t X, bool isNeg = false) { 65 char Buffer[20]; 66 char *BufPtr = Buffer+19; 67 68 *BufPtr = 0; // Null terminate buffer... 69 if (X == 0) *--BufPtr = '0'; // Handle special case... 70 71 while (X) { 72 *--BufPtr = '0' + char(X % 10); 73 X /= 10; 74 } 75 76 if (isNeg) *--BufPtr = '-'; // Add negative sign... 77 78 return std::string(BufPtr); 79} 80 81static inline std::string utostr(uint64_t X, bool isNeg = false) { 82 if (X == uint32_t(X)) 83 return utostr_32(uint32_t(X), isNeg); 84 85 char Buffer[40]; 86 char *BufPtr = Buffer+39; 87 88 *BufPtr = 0; // Null terminate buffer... 89 if (X == 0) *--BufPtr = '0'; // Handle special case... 90 91 while (X) { 92 *--BufPtr = '0' + char(X % 10); 93 X /= 10; 94 } 95 96 if (isNeg) *--BufPtr = '-'; // Add negative sign... 97 return std::string(BufPtr); 98} 99 100 101static inline std::string itostr(int64_t X) { 102 if (X < 0) 103 return utostr(static_cast<uint64_t>(-X), true); 104 else 105 return utostr(static_cast<uint64_t>(X)); 106} 107 108static inline std::string ftostr(double V) { 109 char Buffer[200]; 110 sprintf(Buffer, "%20.6e", V); 111 char *B = Buffer; 112 while (*B == ' ') ++B; 113 return B; 114} 115 116static inline std::string ftostr(const APFloat& V) { 117 if (&V.getSemantics() == &APFloat::IEEEdouble) 118 return ftostr(V.convertToDouble()); 119 else if (&V.getSemantics() == &APFloat::IEEEsingle) 120 return ftostr((double)V.convertToFloat()); 121 return "<unknown format in ftostr>"; // error 122} 123 124static inline std::string LowercaseString(const std::string &S) { 125 std::string result(S); 126 for (unsigned i = 0; i < S.length(); ++i) 127 if (isupper(result[i])) 128 result[i] = char(tolower(result[i])); 129 return result; 130} 131 132static inline std::string UppercaseString(const std::string &S) { 133 std::string result(S); 134 for (unsigned i = 0; i < S.length(); ++i) 135 if (islower(result[i])) 136 result[i] = char(toupper(result[i])); 137 return result; 138} 139 140/// StringsEqualNoCase - Return true if the two strings are equal, ignoring 141/// case. 142static inline bool StringsEqualNoCase(const std::string &LHS, 143 const std::string &RHS) { 144 if (LHS.size() != RHS.size()) return false; 145 for (unsigned i = 0, e = static_cast<unsigned>(LHS.size()); i != e; ++i) 146 if (tolower(LHS[i]) != tolower(RHS[i])) return false; 147 return true; 148} 149 150/// StringsEqualNoCase - Return true if the two strings are equal, ignoring 151/// case. 152static inline bool StringsEqualNoCase(const std::string &LHS, 153 const char *RHS) { 154 for (unsigned i = 0, e = static_cast<unsigned>(LHS.size()); i != e; ++i) { 155 if (RHS[i] == 0) return false; // RHS too short. 156 if (tolower(LHS[i]) != tolower(RHS[i])) return false; 157 } 158 return RHS[LHS.size()] == 0; // Not too long? 159} 160 161/// StringsEqualNoCase - Return true if the two null-terminated C strings are 162/// equal, ignoring 163 164static inline bool StringsEqualNoCase(const char *LHS, const char *RHS, 165 unsigned len) { 166 167 for (unsigned i = 0; i < len; ++i) { 168 if (tolower(LHS[i]) != tolower(RHS[i])) 169 return false; 170 171 // If RHS[i] == 0 then LHS[i] == 0 or otherwise we would have returned 172 // at the previous branch as tolower('\0') == '\0'. 173 if (RHS[i] == 0) 174 return true; 175 } 176 177 return true; 178} 179 180/// CStrInCStrNoCase - Portable version of strcasestr. Locates the first 181/// occurance of c-string 's2' in string 's1', ignoring case. Returns 182/// NULL if 's2' cannot be found. 183static inline const char* CStrInCStrNoCase(const char *s1, const char *s2) { 184 185 // Are either strings NULL or empty? 186 if (!s1 || !s2 || s1[0] == '\0' || s2[0] == '\0') 187 return 0; 188 189 if (s1 == s2) 190 return s1; 191 192 const char *I1=s1, *I2=s2; 193 194 while (*I1 != '\0' && *I2 != '\0' ) 195 if (tolower(*I1) != tolower(*I2)) { // No match. Start over. 196 ++s1; I1 = s1; I2 = s2; 197 } 198 else { // Character match. Advance to the next character. 199 ++I1; ++I2; 200 } 201 202 // If we exhausted all of the characters in 's2', then 's2' appears in 's1'. 203 return *I2 == '\0' ? s1 : 0; 204} 205 206/// StrInStrNoCase - Portable version of strcasestr. Locates the first 207/// occurrence of string 's1' in string 's2', ignoring case. Returns 208/// the offset of s2 in s1 or npos if s2 cannot be found. 209StringRef::size_type StrInStrNoCase(StringRef s1, StringRef s2); 210 211/// getToken - This function extracts one token from source, ignoring any 212/// leading characters that appear in the Delimiters string, and ending the 213/// token at any of the characters that appear in the Delimiters string. If 214/// there are no tokens in the source string, an empty string is returned. 215/// The function returns a pair containing the extracted token and the 216/// remaining tail string. 217std::pair<StringRef, StringRef> getToken(StringRef Source, 218 StringRef Delimiters = " \t\n\v\f\r"); 219 220/// SplitString - Split up the specified string according to the specified 221/// delimiters, appending the result fragments to the output list. 222void SplitString(StringRef Source, 223 SmallVectorImpl<StringRef> &OutFragments, 224 StringRef Delimiters = " \t\n\v\f\r"); 225 226// FIXME: remove when llvm-gcc doesn't use this anymore 227void SplitString(StringRef Source, 228 std::vector<std::string> &OutFragments, 229 StringRef Delimiters = " \t\n\v\f\r"); 230 231/// HashString - Hash funtion for strings. 232/// 233/// This is the Bernstein hash function. 234// 235// FIXME: Investigate whether a modified bernstein hash function performs 236// better: http://eternallyconfuzzled.com/tuts/algorithms/jsw_tut_hashing.aspx 237// X*33+c -> X*33^c 238static inline unsigned HashString(StringRef Str, unsigned Result = 0) { 239 for (unsigned i = 0, e = Str.size(); i != e; ++i) 240 Result = Result * 33 + Str[i]; 241 return Result; 242} 243 244} // End llvm namespace 245 246#endif 247