package ppx_yojson_conv

README.org

#+TITLE: ppx\_yojson\_conv

* [@@deriving yojson]

=ppx_yojson_conv= is a PPX syntax extension that generates code for
converting OCaml types to and from Yojson.Safe, as defined in the
[[https://github.com/ocaml-community/yojson][=yojson=]] library.  Yojson.Safe are defined by the following type:

#+begin_src ocaml
type t =
    [ `Null
    | `Bool of bool
    | `Int of int
    | `Intlit of string
    | `Float of float
    | `String of string
    | `Assoc of (string * t) list
    | `List of t list
    | `Tuple of t list
    | `Variant of string * t option
]
#+end_src

and are rendered as normal json format.

=ppx_yojson_conv= fits into the [[https://github.com/whitequark/ppx_deriving][=ppx_deriving=]] framework, so you can
invoke it the same way you invoke any other deriving plug-in.  Thus,
we can write

#+begin_src ocaml
type int_pair = (int * int) [@@deriving yojson]
#+end_src

to get two values defined automatically, =yojson_of_int_pair= and
=int_pair_of_yojson=.  If we only want one direction, we can write one
of the following.

#+begin_src ocaml
type int_pair = (int * int) [@@deriving yojson_of]
type int_pair = (int * int) [@@deriving of_yojson]
#+end_src

Note that Yojson-converters for primitive types like =int= need to be brought
into scope in order for the ppx to use them, for example by adding =open
Ppx_yojson_conv_lib.Yojson_conv.Primitives= at the beginning of the file.

It's also possible to construct converters based on type expressions,
/i.e./:

#+begin_src ocaml
  [%yojson_of: (int * string) list] [1,"one"; 2,"two"]
  |> Yojson.Safe.to_string;;
  => {|[[1,"one"],[2,"two"]]|}

  [%yojson_of: (int * string) list] [1,"one"; 2,"two"]
  |> [%of_yojson: (int * string) list];;
  => [1,"one"; 2,"two"]
#+end_src

For =%yojson_of=, we can also omit the conversion of some types by
putting underscores for that type name.

#+begin_src ocaml
  [%yojson_of: (int * _) list] [1,"one"; 2,"two"]
  |> Yojson.Safe.to_string;;
  => {|[[1,"_"],[2,"_"]]|}
#+end_src

** Conversion rules

In the following, we'll review the serialization rules for different
OCaml types.

*** Basic types

For numbers like =int=,
=int32=, =int64=, the value is stored as =`Int value=.
For =float= number, the value is stored as =`Float value=.
For the types =char= or =string=, the value is stored as =`String str= where =str= is
respectively a one character string or the string itself.

*** Lists and arrays

OCaml-lists and arrays are represented as Yojson.Safe lists.

*** Tuples and unit

OCaml tuples are treated as lists of values in the same order as in
the tuple.  The type =unit= is treated as Yojson =`Null=.  /e.g./:

#+begin_src ocaml
  (3.14, "foo", "bar bla", 27)  =>  [3.14, "foo", "bar bla", 27]
#+end_src

*** Options

With options, =None= is treated as Yojson =`Null=, and =Some= is
treated as the value contained, as shown below.

#+begin_src ocaml
None        =>  `Null
Some value  =>  value
#+end_src

The rules for variants are described below.

*** Records

Records are represented as Yojson =`Assoc (string * t) list=, where item of the list is a
key-value pair. Each pair consists of the name of the record field
(first element), and its value (second element).  /e.g./:

#+begin_src ocaml
  { foo = (3,4);
    bar = "some string"; }
  =>  {"foo":[3,4],"bar":"some string"}
#+end_src

Type specifications of records allow the use of several attributes. The
attribute =yojson.option= indicates that a record field should be optional.
/e.g./:

#+begin_src ocaml
  type t =
    { x : int option;
      y : int option [@yojson.option];
    } [@@deriving yojson]
#+end_src

The following examples show how this works.

#+begin_src ocaml
  { x = Some 1; y = Some 2; } => {"x":1,"y":2}
  { x = None  ; y = None;   } => {"x":null}
#+end_src


When the JSON object keys differ from the ocaml field names, users can
specify the corresponding JSON key implicitly using =[@key "field"]=,
for example:

#+begin_src ocaml
type t = {
  typ    : float [@key "type"];
  class_ : float [@key "CLASS"];
}
[@@deriving yojson, yojson_fields]
#+end_src

The =yojson_fields= attribute generates the list of JSON keys from a
record type, for example:
#+begin_src ocaml
type ty = {
  x : float [@key "a"];
  y : float [@key "b"];
  z : float
}
[@@deriving yojson_fields]
#+end_src
generates the list below, and the list will not be generated for the signature.
#+begin_src ocaml
yojson_fields_of_ty = ["a"; "b"; "z"]
#+end_src

Note that ppx_deriving_yojson support duplicated fields, while our library does not.

**** Defaults

More complex default values can be specified explicitly using several
constructs, /e.g./:

#+begin_src ocaml
  type t =
    { a : int [@default 42];
      b : int [@default 3] [@yojson_drop_default (=)];
      c : int [@default 3] [@yojson_drop_if fun x -> x = 3];
      d : int list
    } [@@deriving yojson]
#+end_src

The =@default= annotation lets one specify a default value to be
selected if the field is not specified, when converting from
Yojson.Safe.  The =@yojson_drop_default= annotation implies that the
field will be dropped when generating the Yojson.Safe if the value
being serialized is equal to the default according to the specified equality
function. =@yojson_drop_if= is like =@yojson_drop_default=, except that
it lets you specify the condition under which the field is dropped.

***** Specifying equality for [@yojson_drop_default]

The equality used by [@yojson_drop_default] is customizable. There
are several ways to specify the equality function:

#+begin_src ocaml
  type t =
    { a : u [@default u0] [@yojson_drop_default (=)]; (* explicit user-provided function *)
      b : u [@default u0] [@yojson_drop_default.compare]; (* uses [%compare.equal: u] *)
      c : u [@default u0] [@yojson_drop_default.equal]; (* uses [%equal: u] *)
      d : u [@default u0] [@yojson_drop_default.yojson]; (* compares yojson representations *)
      e : u [@default u0] [@yojson_drop_default]; (* deprecated. uses polymorphic equality. *)
    } [@@deriving yojson]
#+end_src

**** Allowing extra fields

The =@yojson.allow_extra_fields= annotation lets one specify that the
yojson-converters should silently ignore extra fields, instead of
raising.  This applies only to the record to which the annotation is
attached, and not to deeper yojson converters that may be called during
conversion of a yojson to the record.

#+begin_src ocaml
  type t = { a: int } [@@deriving yojson]
  {"a":1,"b":2} => exception

  type t = { a: int } [@@deriving yojson] [@@yojson.allow_extra_fields]
  {"a":1,"b":2} => {a = 1}

  type t = A of { a : int } [@yojson.allow_extra_fields] [@@deriving yojson]
  ["A", {"a":1,"b":2}] => A {a = 0}
#+end_src

*** Variants
Constant constructors in variants are represented as a list with one
string, which is the name of the contructor.
Constructors with arguments are represented as lists, the
first element being the constructor name, the rest being its
arguments.
For example:

#+begin_src ocaml
  type t = A | B of int * float * t [@@deriving yojson]
  B (42, 3.14, B (-1, 2.72, A))  =>  ["B",42,3.14,["B",-1,2.72,["A"]]]
#+end_src

The above example also demonstrates recursion in data structures.

if the JSON variant names differ from OCaml conventions, users can specify the
corresponding JSON string explicitly using =[@name "constr"]=, for example:

#+begin_src ocaml
type t =
| Typ   [@name "type"]
| Class [@name "class"]
[@@deriving yojson]
#+end_src

*** Polymorphic variants

Polymorphic variants behave almost the same as ordinary variants.  The
notable difference is that polymorphic variant constructors must
always start with an either lower- or uppercase character, matching
the way it was specified in the type definition.  This is because
OCaml distinguishes between upper and lowercase variant
constructors. Note that type specifications containing unions of
variant types are also supported by the Yojson converter, for
example as in:

#+begin_src ocaml
  type ab = [ `A | `B ] [@@deriving yojson]
  type cd = [ `C | `D ] [@@deriving yojson]
  type abcd = [ ab | cd ] [@@deriving yojson]
#+end_src

However, because `ppx_yojson_conv` needs to generate additional code to
support inclusions of polymorphic variants, `ppx_yojson_conv` needs to
know when processing a type definition whether it might be included in
a polymorphic variant. `ppx_yojson_conv` will only generate the extra
code automatically in the common case where the type definition is
syntactically a polymorphic variant like in the example
above. Otherwise, you will need to indicate it by using `[@@deriving
yojson_poly]` (resp `of_yosjon_poly`) instead of `[@@deriving yojson]` (resp
`of_yojson`):

#+begin_src ocaml
  type ab = [ `A | `B ] [@@deriving yojson]
  type alias_of_ab = ab [@@deriving yojson_poly]
  type abcd = [ ab | `C | `D ] [@@deriving yojson]
#+end_src

*** Polymorphic values

There is nothing special about polymorphic values as long as there are
conversion functions for the type parameters.  /e.g./:

#+begin_src ocaml
type 'a t = A | B of 'a [@@deriving yojson]
type foo = int t [@@deriving yojson]
#+end_src

In the above case the conversion functions will behave as if =foo= had
been defined as a monomorphic version of =t= with ='a= replaced by
=int= on the right hand side.

If a data structure is indeed polymorphic and you want to convert it,
you will have to supply the conversion functions for the type
parameters at runtime.  If you wanted to convert a value of type ='a
t= as in the above example, you would have to write something like
this:

#+begin_src ocaml
  yojson_of_t yojson_of_a v
#+end_src

where =yojson_of_a=, which may also be named differently in this
particular case, is a function that converts values of type ='a= to a
Yojson.  Types with more than one parameter require passing
conversion functions for those parameters in the order of their
appearance on the left hand side of the type definition.

*** Opaque values

Opaque values are ones for which we do not want to perform
conversions.  This may be, because we do not have Yojson
converters for them, or because we do not want to apply them in a
particular type context. /e.g./ to hide large, unimportant parts of
configurations.  To prevent the preprocessor from generating calls to
converters, simply apply the attribute =yojson.opaque= to the type, /e.g./:

#+begin_src ocaml
  type foo = int * (stuff [@yojson.opaque]) [@@deriving yojson]
#+end_src

Thus, there is no need to specify converters for type =stuff=, and if
there are any, they will not be used in this particular context.
Needless to say, it is not possible to convert such a Yojson
back to the original value.  Here is an example conversion:

#+begin_src ocaml
  (42, some_stuff)  =>  [42,"<opaque>"]
#+end_src

*** Exceptions

Unlike Sexp deriver, we are not handling exceptions in the yojson derivier.

*** Hash tables

The Stdlib's Hash tables, which are abstract values in OCaml, are
represented as association lists, /i.e./ lists of key-value pairs,
/e.g./:

#+begin_src scheme
  [["foo",3],["bar",4]]
#+end_src

Reading in the above Yojson as hash table mapping strings to
integers (=(string, int) Hashtbl.t=) will map =foo= to =42= and =bar=
to =3=.

Note that the order of elements in the list may matter, because the
OCaml-implementation of hash tables keeps duplicates.  Bindings will
be inserted into the hash table in the order of appearance. Therefore,
the last binding of a key will be the "visible" one, the others are
"hidden".  See the OCaml documentation on hash tables for details.

** A note about signatures

In signatures, =ppx_yojson_conv= tries to generate an include of a named
interface, instead of a list of value bindings.
That is:

#+begin_src ocaml
type 'a t [@@deriving yojson]
#+end_src

will generate:

#+begin_src ocaml
include Yojsonable.S1 with type 'a t := 'a t
#+end_src

instead of:

#+begin_src ocaml
val t_of_yojson : (Yojson.Safe.t -> 'a) -> Yojson.Safe.t -> 'a t
val yojson_of_t : ('a -> Yojson.Safe.t) -> 'a t -> Yojson.Safe.t
#+end_src

There are however a number of limitations:
- the type has to be named t
- the type can only have up to 3 parameters
- there shouldn't be any constraint on the type parameters

If these aren't met, then =ppx_yojson_conv= will simply generate a list of value
bindings.