string_func.h

Go to the documentation of this file.

/*
 * This file is part of OpenTTD.
 * OpenTTD is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2.
 * OpenTTD is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with OpenTTD. If not, see <http://www.gnu.org/licenses/>.
 */
 
#ifndef STRING_FUNC_H
#define STRING_FUNC_H
 
#include <iosfwd>
 
#include "core/bitmath_func.hpp"
#include "string_type.h"
 
void strecpy(std::span<char> dst, std::string_view src);
 
std::string FormatArrayAsHex(std::span<const uint8_t> data);
 
void StrMakeValidInPlace(char *str, const char *last, StringValidationSettings settings = SVS_REPLACE_WITH_QUESTION_MARK) NOACCESS(2);
[[nodiscard]] std::string StrMakeValid(std::string_view str, StringValidationSettings settings = SVS_REPLACE_WITH_QUESTION_MARK);
void StrMakeValidInPlace(char *str, StringValidationSettings settings = SVS_REPLACE_WITH_QUESTION_MARK);
 
bool strtolower(std::string &str, std::string::size_type offs = 0);
 
[[nodiscard]] bool StrValid(std::span<const char> str);
void StrTrimInPlace(std::string &str);
std::string_view StrTrimView(std::string_view str);
 
[[nodiscard]] bool StrStartsWithIgnoreCase(std::string_view str, const std::string_view prefix);
[[nodiscard]] bool StrEndsWithIgnoreCase(std::string_view str, const std::string_view suffix);
 
[[nodiscard]] int StrCompareIgnoreCase(const std::string_view str1, const std::string_view str2);
[[nodiscard]] bool StrEqualsIgnoreCase(const std::string_view str1, const std::string_view str2);
[[nodiscard]] int StrNaturalCompare(std::string_view s1, std::string_view s2, bool ignore_garbage_at_front = false);
[[nodiscard]] bool StrNaturalContains(const std::string_view str, const std::string_view value);
[[nodiscard]] bool StrNaturalContainsIgnoreCase(const std::string_view str, const std::string_view value);
 
bool ConvertHexToBytes(std::string_view hex, std::span<uint8_t> bytes);
 
struct CaseInsensitiveComparator {
  bool operator()(const std::string_view s1, const std::string_view s2) const { return StrCompareIgnoreCase(s1, s2) < 0; }
};
 
inline bool StrEmpty(const char *s)
{
  return s == nullptr || s[0] == '\0';
}
 
inline size_t ttd_strnlen(const char *str, size_t maxlen)
{
  const char *t;
  for (t = str; static_cast<size_t>(t - str) < maxlen && *t != '\0'; t++) {}
  return t - str;
}
 
bool IsValidChar(char32_t key, CharSetFilter afilter);
 
size_t Utf8Decode(char32_t *c, const char *s);
/* std::string_view::iterator might be char *, in which case we do not want this templated variant to be taken. */
template <typename T> requires (!std::is_same_v<T, char *> && (std::is_same_v<std::string_view::iterator, T> || std::is_same_v<std::string::iterator, T>))
inline size_t Utf8Decode(char32_t *c, T &s) { return Utf8Decode(c, &*s); }
size_t Utf8Encode(char *buf, char32_t c);
size_t Utf8Encode(std::ostreambuf_iterator<char> &buf, char32_t c);
size_t Utf8Encode(std::back_insert_iterator<std::string> &buf, char32_t c);
inline size_t Utf8Encode(std::string::iterator &s, char32_t c) { return Utf8Encode(&*s, c); }
size_t Utf8TrimString(char *s, size_t maxlen);
 
 
inline char32_t Utf8Consume(const char **s)
{
  char32_t c;
  *s += Utf8Decode(&c, *s);
  return c;
}
 
template <class Titr>
inline char32_t Utf8Consume(Titr &s)
{
  char32_t c;
  s += Utf8Decode(&c, &*s);
  return c;
}
 
inline int8_t Utf8CharLen(char32_t c)
{
  if (c < 0x80)       return 1;
  if (c < 0x800)      return 2;
  if (c < 0x10000)    return 3;
  if (c < 0x110000)   return 4;
 
  /* Invalid valid, we encode as a '?' */
  return 1;
}
 
 
inline int8_t Utf8EncodedCharLen(char c)
{
  if (GB(c, 3, 5) == 0x1E) return 4;
  if (GB(c, 4, 4) == 0x0E) return 3;
  if (GB(c, 5, 3) == 0x06) return 2;
  if (GB(c, 7, 1) == 0x00) return 1;
 
  /* Invalid UTF8 start encoding */
  return 0;
}
 
 
/* Check if the given character is part of a UTF8 sequence */
inline bool IsUtf8Part(char c)
{
  return GB(c, 6, 2) == 2;
}
 
inline char *Utf8PrevChar(char *s)
{
  char *ret = s;
  while (IsUtf8Part(*--ret)) {}
  return ret;
}
 
inline const char *Utf8PrevChar(const char *s)
{
  const char *ret = s;
  while (IsUtf8Part(*--ret)) {}
  return ret;
}
 
inline std::string::iterator Utf8PrevChar(std::string::iterator &s)
{
  auto cur = s;
  do {
    cur = std::prev(cur);
  } while (IsUtf8Part(*cur));
  return cur;
}
 
size_t Utf8StringLength(const char *s);
size_t Utf8StringLength(const std::string &str);
 
inline bool Utf16IsLeadSurrogate(uint c)
{
  return c >= 0xD800 && c <= 0xDBFF;
}
 
inline bool Utf16IsTrailSurrogate(uint c)
{
  return c >= 0xDC00 && c <= 0xDFFF;
}
 
inline char32_t Utf16DecodeSurrogate(uint lead, uint trail)
{
  return 0x10000 + (((lead - 0xD800) << 10) | (trail - 0xDC00));
}
 
inline char32_t Utf16DecodeChar(const uint16_t *c)
{
  if (Utf16IsLeadSurrogate(c[0])) {
    return Utf16DecodeSurrogate(c[0], c[1]);
  } else {
    return *c;
  }
}
 
inline bool IsTextDirectionChar(char32_t c)
{
  switch (c) {
    case CHAR_TD_LRM:
    case CHAR_TD_RLM:
    case CHAR_TD_LRE:
    case CHAR_TD_RLE:
    case CHAR_TD_LRO:
    case CHAR_TD_RLO:
    case CHAR_TD_PDF:
      return true;
 
    default:
      return false;
  }
}
 
inline bool IsPrintable(char32_t c)
{
  if (c < 0x20)   return false;
  if (c < 0xE000) return true;
  if (c < 0xE200) return false;
  return true;
}
 
inline bool IsWhitespace(char32_t c)
{
  return c == 0x0020 /* SPACE */ || c == 0x3000; /* IDEOGRAPHIC SPACE */
}
 
/* Needed for NetBSD version (so feature) testing */
#if defined(__NetBSD__) || defined(__FreeBSD__)
#include <sys/param.h>
#endif
 
/* strcasestr is available for _GNU_SOURCE, BSD and some Apple */
#if defined(_GNU_SOURCE) || (defined(__BSD_VISIBLE) && __BSD_VISIBLE) || (defined(__APPLE__) && (!defined(_POSIX_C_SOURCE) || defined(_DARWIN_C_SOURCE))) || defined(_NETBSD_SOURCE)
# undef DEFINE_STRCASESTR
#else
# define DEFINE_STRCASESTR
char *strcasestr(const char *haystack, const char *needle);
#endif /* strcasestr is available */
 
#endif /* STRING_FUNC_H */