libc/src/string/string_utils.h

   1 //===-- String utils --------------------------------------------*- 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 LIBC_SRC_STRING_STRING_UTILS_H
  10 #define LIBC_SRC_STRING_STRING_UTILS_H
  11
  12 #include "utils/CPP/Bitset.h"
  13 #include <stddef.h> // size_t
  14
  15 namespace __llvm_libc {
  16 namespace internal {
  17
  18 // Returns the length of a string, denoted by the first occurrence
  19 // of a null terminator.
  20 static inline size_t string_length(const char *src) {
  21   size_t length;
  22   for (length = 0; *src; ++src, ++length)
  23     ;
  24   return length;
  25 }
  26
  27 // Returns the first occurrence of 'ch' within the first 'n' characters of
  28 // 'src'. If 'ch' is not found, returns nullptr.
  29 static inline void *find_first_character(const unsigned char *src,
  30                                          unsigned char ch, size_t n) {
  31   for (; n && *src != ch; --n, ++src)
  32     ;
  33   return n ? const_cast<unsigned char *>(src) : nullptr;
  34 }
  35
  36 // Returns the maximum length span that contains only characters not found in
  37 // 'segment'. If no characters are found, returns the length of 'src'.
  38 static inline size_t complementary_span(const char *src, const char *segment) {
  39   const char *initial = src;
  40   cpp::Bitset<256> bitset;
  41
  42   for (; *segment; ++segment)
  43     bitset.set(*segment);
  44   for (; *src && !bitset.test(*src); ++src)
  45     ;
  46   return src - initial;
  47 }
  48
  49 // Given the similarities between strtok and strtok_r, we can implement both
  50 // using a utility function. On the first call, 'src' is scanned for the
  51 // first character not found in 'delimiter_string'. Once found, it scans until
  52 // the first character in the 'delimiter_string' or the null terminator is
  53 // found. We define this span as a token. The end of the token is appended with
  54 // a null terminator, and the token is returned. The point where the last token
  55 // is found is then stored within 'context' for subsequent calls. Subsequent
  56 // calls will use 'context' when a nullptr is passed in for 'src'. Once the null
  57 // terminating character is reached, returns a nullptr.
  58 static inline char *string_token(char *__restrict src,
  59                                  const char *__restrict delimiter_string,
  60                                  char **__restrict saveptr) {
  61   cpp::Bitset<256> delimiter_set;
  62   for (; *delimiter_string; ++delimiter_string)
  63     delimiter_set.set(*delimiter_string);
  64
  65   src = src ? src : *saveptr;
  66   for (; *src && delimiter_set.test(*src); ++src)
  67     ;
  68   if (!*src) {
  69     *saveptr = src;
  70     return nullptr;
  71   }
  72   char *token = src;
  73   for (; *src && !delimiter_set.test(*src); ++src)
  74     ;
  75   if (*src) {
  76     *src = '\0';
  77     ++src;
  78   }
  79   *saveptr = src;
  80   return token;
  81 }
  82
  83 } // namespace internal
  84 } // namespace __llvm_libc
  85
  86 #endif //  LIBC_SRC_STRING_STRING_UTILS_H