[PATCH] libcamera: yaml_parser: Use std::from_chars()

Laurent Pinchart laurent.pinchart at ideasonboard.com
Wed Nov 13 12:43:12 CET 2024


On Wed, Nov 13, 2024 at 12:39:01PM +0100, Jacopo Mondi wrote:
> Hi Laurent
> 
> On Tue, Nov 12, 2024 at 08:56:37PM +0200, Laurent Pinchart wrote:
> > std::from_chars(), introduced in C++17, is a fast, locale-independent
> > string-to-arithmetic conversion function. The C++ standard library
> > provides overloads for all integer types, making it a prime candidate to
> > replace the manual handling of integer sizes in the YamlParser string to
> > integer conversion.
> >
> > Compared to std::strtol(), std::from_chars() doesn't recognize the '0x'
> 
> only if the base is 16, or do I read
> https://en.cppreference.com/w/cpp/utility/from_chars
> wrong

You're correct. If the base is 10, there should be no '0x' prefix
anyway.

> > prefix or '+' prefix, and doesn't ignore leading white space. As the
> > YamlParser doesn't require those features, std::from_chars() can be used
> > safely, reducing the amount of code.
> 
> Ok, I presume we won't have "0x.." entries in config files ?

I wouldn't assume so, as we explicitly specify base 10 in the strtol()
call.

> > C++17 also requires the standard C++ library to provide overloads for
> > floating-point types, but libc++ does not implement those. The float and
> 
> Do you know why ?

I believe it's more complicated than it may seem. The C++ standard
requires that the roundtrip std::from_chars(std::to_chars(value))
produces the exact same value, and apparently it's not trivial.

> > bool implementations of YamlParser::Getter::get() are therefore kept
> > as-is.
> >
> > Signed-off-by: Laurent Pinchart <laurent.pinchart at ideasonboard.com>
> > ---
> >  include/libcamera/internal/yaml_parser.h |   2 +-
> >  src/libcamera/yaml_parser.cpp            | 168 ++++-------------------
> >  2 files changed, 29 insertions(+), 141 deletions(-)
> >
> > diff --git a/include/libcamera/internal/yaml_parser.h b/include/libcamera/internal/yaml_parser.h
> > index de452844fe0a..8c7916565946 100644
> > --- a/include/libcamera/internal/yaml_parser.h
> > +++ b/include/libcamera/internal/yaml_parser.h
> > @@ -224,7 +224,7 @@ private:
> >  		Empty,
> >  	};
> >
> > -	template<typename T>
> > +	template<typename T, typename Enable = void>
> >  	struct Getter {
> >  		std::optional<T> get(const YamlObject &obj) const;
> >  	};
> > diff --git a/src/libcamera/yaml_parser.cpp b/src/libcamera/yaml_parser.cpp
> > index 7c0b341a32e7..db256ec5b04d 100644
> > --- a/src/libcamera/yaml_parser.cpp
> > +++ b/src/libcamera/yaml_parser.cpp
> > @@ -7,6 +7,7 @@
> >
> >  #include "libcamera/internal/yaml_parser.h"
> >
> > +#include <charconv>
> >  #include <cstdlib>
> >  #include <errno.h>
> >  #include <functional>
> > @@ -146,151 +147,38 @@ YamlObject::Getter<bool>::get(const YamlObject &obj) const
> >  	return std::nullopt;
> >  }
> >
> > -namespace {
> > -
> > -bool parseSignedInteger(const std::string &str, long min, long max,
> 
> out-of-range conversion is now handled by std::from_chars(), nice
> 
> > -			long *result)
> > +template<typename T>
> > +struct YamlObject::Getter<T, std::enable_if_t<
> > +	std::is_same_v<int8_t, T> ||
> > +	std::is_same_v<uint8_t, T> ||
> > +	std::is_same_v<int16_t, T> ||
> > +	std::is_same_v<uint16_t, T> ||
> > +	std::is_same_v<int32_t, T> ||
> > +	std::is_same_v<uint32_t, T>>>
> >  {
> > -	if (str == "")
> > -		return false;
> > +	std::optional<T> get(const YamlObject &obj) const
> > +	{
> > +		if (obj.type_ != Type::Value)
> > +			return std::nullopt;
> >
> > -	char *end;
> > +		const std::string &str = obj.value_;
> > +		T value;
> >
> > -	errno = 0;
> > -	long value = std::strtol(str.c_str(), &end, 10);
> > +		auto [ptr, ec] = std::from_chars(str.data(), str.data() + str.size(),
> > +						 value);
> > +		if (ptr != str.data() + str.size() || ec != std::errc())
> > +			return std::nullopt;
> >
> > -	if ('\0' != *end || errno == ERANGE || value < min || value > max)
> > -		return false;
> > +		return value;
> > +	}
> > +};
> >
> > -	*result = value;
> > -	return true;
> > -}
> > -
> > -bool parseUnsignedInteger(const std::string &str, unsigned long max,
> > -			  unsigned long *result)
> > -{
> > -	if (str == "")
> > -		return false;
> > -
> > -	/*
> > -	 * strtoul() accepts strings representing a negative number, in which
> > -	 * case it negates the converted value. We don't want to silently accept
> > -	 * negative values and return a large positive number, so check for a
> > -	 * minus sign (after optional whitespace) and return an error.
> > -	 */
> > -	std::size_t found = str.find_first_not_of(" \t");
> > -	if (found != std::string::npos && str[found] == '-')
> > -		return false;
> > -
> > -	char *end;
> > -
> > -	errno = 0;
> > -	unsigned long value = std::strtoul(str.c_str(), &end, 10);
> > -
> > -	if ('\0' != *end || errno == ERANGE || value > max)
> > -		return false;
> > -
> > -	*result = value;
> > -	return true;
> > -}
> > -
> > -} /* namespace */
> > -
> > -template<>
> > -std::optional<int8_t>
> > -YamlObject::Getter<int8_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	long value;
> > -
> > -	if (!parseSignedInteger(obj.value_, std::numeric_limits<int8_t>::min(),
> > -				std::numeric_limits<int8_t>::max(), &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > -
> > -template<>
> > -std::optional<uint8_t>
> > -YamlObject::Getter<uint8_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	unsigned long value;
> > -
> > -	if (!parseUnsignedInteger(obj.value_, std::numeric_limits<uint8_t>::max(),
> > -				  &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > -
> > -template<>
> > -std::optional<int16_t>
> > -YamlObject::Getter<int16_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	long value;
> > -
> > -	if (!parseSignedInteger(obj.value_, std::numeric_limits<int16_t>::min(),
> > -				std::numeric_limits<int16_t>::max(), &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > -
> > -template<>
> > -std::optional<uint16_t>
> > -YamlObject::Getter<uint16_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	unsigned long value;
> > -
> > -	if (!parseUnsignedInteger(obj.value_, std::numeric_limits<uint16_t>::max(),
> > -				  &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > -
> > -template<>
> > -std::optional<int32_t>
> > -YamlObject::Getter<int32_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	long value;
> > -
> > -	if (!parseSignedInteger(obj.value_, std::numeric_limits<int32_t>::min(),
> > -				std::numeric_limits<int32_t>::max(), &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > -
> > -template<>
> > -std::optional<uint32_t>
> > -YamlObject::Getter<uint32_t>::get(const YamlObject &obj) const
> > -{
> > -	if (obj.type_ != Type::Value)
> > -		return std::nullopt;
> > -
> > -	unsigned long value;
> > -
> > -	if (!parseUnsignedInteger(obj.value_, std::numeric_limits<uint32_t>::max(),
> > -				  &value))
> > -		return std::nullopt;
> > -
> > -	return value;
> > -}
> > +template struct YamlObject::Getter<int8_t>;
> > +template struct YamlObject::Getter<uint8_t>;
> > +template struct YamlObject::Getter<int16_t>;
> > +template struct YamlObject::Getter<uint16_t>;
> > +template struct YamlObject::Getter<int32_t>;
> > +template struct YamlObject::Getter<uint32_t>;
> >
> >  template<>
> >  std::optional<float>
> >
> > base-commit: dcb90f13cf79d909d04dacf2cfbd256a4744347f
> 
> Looks good to me
> Reviewed-by: Jacopo Mondi <jacopo.mondi at ideasonboard.com>

-- 
Regards,

Laurent Pinchart


More information about the libcamera-devel mailing list