Generate functions to convert OCaml values to/from Python values.
ppx_python is a PPX syntax extension that generates code for
converting OCaml types to and from Python. This uses
the pyml OCaml bindings for Python
to start a Python runtime, create the Python objects or analyze
them.
Annotate the type with [@@deriving python] as in the following example.
let () =
if not (Pyml.Py.is_initialized ())
then Pyml.Py.initialize ~version:3 ()
;;
type int_pair = (int * int) [@@deriving python];;This results in two functions being created automatically, python_of_int_pair and int_pair_of_python
with the following types.
val python_of_int_pair: int_pair -> pyobject
val int_pair_of_python: pyobject -> int_pairIf only one direction is needed it is possible to write one of the following.
type int_pair = (int * int) [@@deriving python_of]
type int_pair = (int * int) [@@deriving of_python]Python converters for primitive types such as int, float, bool,
or string can be brought into scope by opening Python_lib.
It is also possible to construct converters based on type expressions as in the following example.
let pyobject = [%python_of: (int * string) list] [ 1, "one"; 2, "two" ];;
Stdio.printf "pyobject: %s\n" (Pyml.Py.Object.to_string pyobject);;
let list = [%of_python: (int * string) list] pyobject;;The conversion is straightforward for basic types such as int, float, bool, or string.
unit is converted to None.
OCaml tuples are converted into Python tuples. OCaml lists and arrays are converted in Python lists.
For options, None is used on the Python side to represent the None case. Otherwise the value is
directly available. Note that this makes ocaml values Some None and None indistinguishable on the
Python side.
Records are represented using Python dictionaries which keys are strings. The [@python.default]
attribute can be used on some of the fields: these fields are then optional on the Python side
and if not present the default value gets used.
type t =
{ foo : int [@python.default 42]
; bar : float
} [@@deriving python]Variants don't have an idiomatic Python representation. They get converted to a pair where the first
element is the constructor as a string and the second element is the content of the variant or None
if this variant case does not embed any data.
Below are some more involved usage examples taken from the test suite.
type t =
{ field_a : int
; field_b : string
}
[@@deriving python]
type u =
{ foo : int * int
; bar : t
}
[@@deriving python]
type v =
| A
| B of string
| C of int
| D of t * string
| E of
{ x : int
; y : string
}
[@@deriving python]
type 'a w =
| One of 'a
| Multi of 'a list
[@@deriving python]
type 'a tree =
| Leaf of 'a
| Node of 'a tree * 'a tree
[@@deriving python]