AZ
и это бест практис
Size: a a a
2020 January 24
E
ок, согласен
AZ
Although I can’t give it the full treatment it deserves right now, I’d still like to touch on the idea briefly so that interested readers may be able to find other resources on the subject should they wish to do so. The key idea is that many dynamically typed languages idiomatically reuse simple data structures like hashmaps to represent what in statically-typed languages are often represented by bespoke datatypes (usually defined as classes or structs).
These two styles facilitate very different flavors of programming. A JavaScript or Clojure program may represent a record as a hashmap from string or symbol keys to values, written using object or hash literals and manipulated using ordinary functions from the standard library that manipulate keys and values in a generic way. This makes it straightforward to take two records and union their fields or to take an arbitrary (or even dynamic) subselection of fields from an existing record.
In contrast, most static type systems do not allow such free-form manipulation of records because records are not maps at all but unique types distinct from all other types. These types are uniquely identified by their (fully-qualified) name, hence the term nominal typing. If you wish to take a subselection of a struct’s fields, you must define an entirely new struct; doing this often creates an explosion of awkward boilerplate.
These two styles facilitate very different flavors of programming. A JavaScript or Clojure program may represent a record as a hashmap from string or symbol keys to values, written using object or hash literals and manipulated using ordinary functions from the standard library that manipulate keys and values in a generic way. This makes it straightforward to take two records and union their fields or to take an arbitrary (or even dynamic) subselection of fields from an existing record.
In contrast, most static type systems do not allow such free-form manipulation of records because records are not maps at all but unique types distinct from all other types. These types are uniquely identified by their (fully-qualified) name, hence the term nominal typing. If you wish to take a subselection of a struct’s fields, you must define an entirely new struct; doing this often creates an explosion of awkward boilerplate.
SP
в тс все неплохо с этим
SP
есть keyof есть &
VS
😂 жжешь. На самом деле для типичного веба неплох после php, nodejs и прочего
Там где PHP, JS норм там уже все что угодно норм. Даже Go
SP
тс имхо стильно лучше го
AZ
есть keyof есть &
да, это приятно
AZ
но то что валидация типчиков прикручена где-то слева через io-ts это неудобно
MK
но самое неприятное в проекте на ТС - это коллеги, любители джаваскрипта
MK
которые делают все, чтобы отключить тс
DS
Although I can’t give it the full treatment it deserves right now, I’d still like to touch on the idea briefly so that interested readers may be able to find other resources on the subject should they wish to do so. The key idea is that many dynamically typed languages idiomatically reuse simple data structures like hashmaps to represent what in statically-typed languages are often represented by bespoke datatypes (usually defined as classes or structs).
These two styles facilitate very different flavors of programming. A JavaScript or Clojure program may represent a record as a hashmap from string or symbol keys to values, written using object or hash literals and manipulated using ordinary functions from the standard library that manipulate keys and values in a generic way. This makes it straightforward to take two records and union their fields or to take an arbitrary (or even dynamic) subselection of fields from an existing record.
In contrast, most static type systems do not allow such free-form manipulation of records because records are not maps at all but unique types distinct from all other types. These types are uniquely identified by their (fully-qualified) name, hence the term nominal typing. If you wish to take a subselection of a struct’s fields, you must define an entirely new struct; doing this often creates an explosion of awkward boilerplate.
These two styles facilitate very different flavors of programming. A JavaScript or Clojure program may represent a record as a hashmap from string or symbol keys to values, written using object or hash literals and manipulated using ordinary functions from the standard library that manipulate keys and values in a generic way. This makes it straightforward to take two records and union their fields or to take an arbitrary (or even dynamic) subselection of fields from an existing record.
In contrast, most static type systems do not allow such free-form manipulation of records because records are not maps at all but unique types distinct from all other types. These types are uniquely identified by their (fully-qualified) name, hence the term nominal typing. If you wish to take a subselection of a struct’s fields, you must define an entirely new struct; doing this often creates an explosion of awkward boilerplate.
На шейплессе в две строчки делается и статически типизированно.
AZ
На шейплессе в две строчки делается и статически типизированно.
я рад, но обычно под "статически типизированными япами" имеют ввиду джавы
SP
я рад, но обычно под "статически типизированными япами" имеют ввиду джавы
я на такое говорю что жава не статика но похоже это слишком смело )
MK
но то что валидация типчиков прикручена где-то слева через io-ts это неудобно
https://github.com/ts-type-makeup/superstruct-ts-transformer
вот такая штука еще есть (но еще не юзал)
вот такая штука еще есть (но еще не юзал)
MK
но пока в проде мало где используется
SP
ts-type-visitor
SP
если по статике генерить типы иотс то был бы топ
SP
деприкейшон
MK