text_light.hpp

Go to the documentation of this file.

#pragma once
 
/// @file userver/utils/text_light.hpp
/// @brief Text utilities that do not use locales
/// @ingroup userver_universal
 
#include <string>
#include <string_view>
#include <vector>
 
USERVER_NAMESPACE_BEGIN
 
/// @brief Text utilities
namespace utils::text {
 
/// Return trimmed copy of string.
[[nodiscard]]
std::string Trim(std::string_view str);
 
/// Trim string in-place.
[[nodiscard]]
std::string Trim(std::string&& str);
 
/// Return trimmed view of a string.
[[nodiscard]]
std::string_view TrimView(std::string_view str);
 
enum class SplitFlags {
    kNone = 0,
    kCompressAdjacentSeparators = 1 << 0,
};
 
/// Split string by separators
///
/// @snippet utils/text_light_test.cpp  SplitMultiple
std::vector<std::string> Split(
    std::string_view str,
    std::string_view separators,
    SplitFlags split_flags = SplitFlags::kCompressAdjacentSeparators
);
 
/// Split string by separators and return a non-owning container of chunks.
///
/// @warning Initial `str` should outlive the result of the function
///
/// @snippet utils/text_light_test.cpp  SplitStringViewMultiple
std::vector<std::string_view> SplitIntoStringViewVector(std::string_view str, std::string_view separators);
 
/// Join string
std::string Join(const std::vector<std::string>& strs, std::string_view sep);
 
/// Return number formatted
std::string Format(double value, int ndigits);
 
/// Return true if `hay` starts with `needle`, false otherwise.
/// @deprecated Use std::string_view::starts_with() instead.
constexpr bool StartsWith(std::string_view hay, std::string_view needle) noexcept { return hay.starts_with(needle); }
 
/// Return true if `hay` ends with `needle`, false otherwise.
/// @deprecated Use std::string_view::ends_with() instead.
constexpr bool EndsWith(std::string_view hay, std::string_view needle) noexcept { return hay.ends_with(needle); }
 
/// Case insensitive (ASCII only) variant of StartsWith()
bool ICaseStartsWith(std::string_view hay, std::string_view needle) noexcept;
 
/// Case insensitive (ASCII only) variant of EndsWith()
bool ICaseEndsWith(std::string_view hay, std::string_view needle) noexcept;
 
/// Removes double quotes from front and back of string.
///
/// Examples:
/// @code
///    RemoveQuotes("\"test\"")      // returns "test"
///    RemoveQuotes("\"test")        // returns "\"test"
///    RemoveQuotes("'test'")        // returns "'test'"
///    RemoveQuotes("\"\"test\"\"")  // returns "\"test\""
/// @endcode
std::string RemoveQuotes(std::string_view str);
 
/// Checks whether the character is an ASCII character
bool IsAscii(char ch) noexcept;
 
/// Checks whether the character is a whitespace character in C locale
bool IsAsciiSpace(char ch) noexcept;
 
/// Checks if text contains only ASCII characters
bool IsAscii(std::string_view text) noexcept;
 
/// @brief UTF8 text utilities
namespace utf8 {
 
/// Returns the length in bytes of the UTF-8 code point by the first byte.
unsigned CodePointLengthByFirstByte(unsigned char c) noexcept;
 
/// `bytes` must not be a nullptr, `length` must not be 0.
bool IsWellFormedCodePoint(const unsigned char* bytes, std::size_t length) noexcept;
 
/// `bytes` must not be a nullptr, `length` must not be 0.
bool IsValid(const unsigned char* bytes, std::size_t length) noexcept;
 
/// returns number of utf-8 code points, text must be in utf-8 encoding
/// @throws std::runtime_error if not a valid UTF8 text
std::size_t GetCodePointsCount(std::string_view text);
 
/// Removes the longest (possible empty) suffix of `str` which is a proper
/// prefix of some utf-8 multibyte character. If `str` is not in utf-8 it may
/// remove some suffix of length up to 3.
void TrimTruncatedEnding(std::string& str);
 
/// @see void TrimTruncatedEnding(std::string& str)
/// @warning this **does not** change the original string
void TrimViewTruncatedEnding(std::string_view& view);
 
/// Returns position in `text` where utf-8 code point with position `pos` starts
/// OR `text.length()` if `text` contains less than or equal to `pos` points
/// @warning this **does not** check if `text` is valid utf-8 text
std::size_t GetTextPosByCodePointPos(std::string_view text, std::size_t pos) noexcept;
 
/// Removes the first `count` utf-8 code points from `text`
/// @warning this **does not** check if `text` is valid utf-8 text
void RemovePrefix(std::string& text, std::size_t count) noexcept;
 
/// @see void RemovePrefix(std::string& text, std::size_t count)
/// @warning this **does not** change the original string
void RemoveViewPrefix(std::string_view& text, std::size_t count) noexcept;
 
/// Takes the first `count` utf-8 code points from `text`
/// @warning this **does not** check if `text` is valid utf-8 text
void TakePrefix(std::string& text, std::size_t count) noexcept;
 
/// @see void TakePrefix(std::string& text, std::size_t count)
/// @warning this **does not** change the original string
void TakeViewPrefix(std::string_view& text, std::size_t count) noexcept;
 
}  // namespace utf8
 
/// Checks if text is in utf-8 encoding
bool IsUtf8(std::string_view text) noexcept;
 
/// Checks text on matching to the following conditions:
/// 1. text is in utf-8 encoding
/// 2. text does not contain any of control ascii characters
/// 3. if flag ascii is true than text contains only ascii characters
bool IsPrintable(std::string_view text, bool ascii_only = true) noexcept;
 
/// Checks if there are no embedded null ('\0') characters in text
bool IsCString(std::string_view text) noexcept;
 
/// convert CamelCase to snake_case(underscore)
std::string CamelCaseToSnake(std::string_view camel);
 
/// convert snake_case or SCREAMING_SNAKE_CASE to CamelCase
std::string SnakeCaseToCamel(std::string_view snake);
 
}  // namespace utils::text
 
USERVER_NAMESPACE_END