forked from mpusz/mp-units
implementing units::ratio_add and tests
not yet used in src/include/units/bits/dim_consolidate.h because it breaks and cascades (due to new exponent we think)
This commit is contained in:
committed by
Mateusz Pusz
parent
dfb9042e9d
commit
0566cc631d
@@ -30,10 +30,10 @@ namespace units::detail {
|
||||
|
||||
/**
|
||||
* @brief Consolidates contiguous ranges of exponents of the same dimension
|
||||
*
|
||||
*
|
||||
* If there is more than one exponent with the same dimension they are aggregated into one exponent by adding
|
||||
* their exponents. If this accumulation will result with 0, such a dimension is removed from the list.
|
||||
*
|
||||
*
|
||||
* @tparam D derived dimension to consolidate
|
||||
*/
|
||||
template<typename ExpList>
|
||||
@@ -56,7 +56,7 @@ struct dim_consolidate<exp_list<E1, ERest...>> {
|
||||
|
||||
template<BaseDimension Dim, std::intmax_t Num1, std::intmax_t Den1, std::intmax_t Num2, std::intmax_t Den2, typename... ERest>
|
||||
struct dim_consolidate<exp_list<exp<Dim, Num1, Den1>, exp<Dim, Num2, Den2>, ERest...>> {
|
||||
// TODO: provide custom implementation for ratio_add
|
||||
// TODO: we have ration_add now, but dim_consolidate etc, now need to cope with our new ratio
|
||||
using r1 = std::ratio<Num1, Den1>;
|
||||
using r2 = std::ratio<Num2, Den2>;
|
||||
using r = std::ratio_add<r1, r2>;
|
||||
|
||||
Reference in New Issue
Block a user