Module Ltac2_plugin.Tac2ffi

Toplevel values
type closure
type valexpr =
| ValInt of int(*

Immediate integers

*)
| ValBlk of Tac2expr.tag * valexpr array(*

Structured blocks

*)
| ValStr of Stdlib.Bytes.t(*

Strings

*)
| ValCls of closure(*

Closures

*)
| ValOpn of Names.KerName.t * valexpr array(*

Open constructors

*)
| ValExt : 'a Tac2dyn.Val.tag * 'a -> valexpr(*

Arbitrary data

*)
| ValUint63 of Uint63.t(*

Primitive integers

*)
| ValFloat of Float64.t(*

Primitive floats

*)
type 'a arity
val arity_one : (valexpr -> valexpr Proofview.tactic) arity
val arity_suc : 'a arity -> (valexpr -> 'a) arity
val mk_closure : 'v arity -> 'v -> closure
module Valexpr : sig ... end
Ltac2 FFI
type 'a repr
val repr_of : 'a repr -> 'a -> valexpr
val repr_to : 'a repr -> valexpr -> 'a
val make_repr : ('a -> valexpr) -> (valexpr -> 'a) -> 'a repr
val map_repr : ('a -> 'b) -> ('b -> 'a) -> 'a repr -> 'b repr

These functions allow to convert back and forth between OCaml and Ltac2 data representation. The to_* functions raise an anomaly whenever the data has not expected shape.

val of_unit : unit -> valexpr
val to_unit : valexpr -> unit
val unit : unit repr
val of_int : int -> valexpr
val to_int : valexpr -> int
val int : int repr
val of_bool : bool -> valexpr
val to_bool : valexpr -> bool
val bool : bool repr
val of_char : char -> valexpr
val to_char : valexpr -> char
val char : char repr
val of_bytes : Stdlib.Bytes.t -> valexpr
val to_bytes : valexpr -> Stdlib.Bytes.t
val bytes : Stdlib.Bytes.t repr
val of_string : string -> valexpr

WARNING these functions copy

val to_string : valexpr -> string
val string : string repr
val of_list : ('a -> valexpr) -> 'a list -> valexpr
val to_list : (valexpr -> 'a) -> valexpr -> 'a list
val list : 'a repr -> 'a list repr
val of_constr : EConstr.t -> valexpr
val to_constr : valexpr -> EConstr.t
val constr : EConstr.t repr
val of_exn : Exninfo.iexn -> valexpr
val to_exn : valexpr -> Exninfo.iexn
val exn : Exninfo.iexn repr
val of_ident : Names.Id.t -> valexpr
val to_ident : valexpr -> Names.Id.t
val ident : Names.Id.t repr
val of_closure : closure -> valexpr
val to_closure : valexpr -> closure
val closure : closure repr
val of_block : (int * valexpr array) -> valexpr
val to_block : valexpr -> int * valexpr array
val block : (int * valexpr array) repr
val of_array : ('a -> valexpr) -> 'a array -> valexpr
val to_array : (valexpr -> 'a) -> valexpr -> 'a array
val array : 'a repr -> 'a array repr
val of_tuple : valexpr array -> valexpr
val to_tuple : valexpr -> valexpr array
val of_pair : ('a -> valexpr) -> ('b -> valexpr) -> ('a * 'b) -> valexpr
val to_pair : (valexpr -> 'a) -> (valexpr -> 'b) -> valexpr -> 'a * 'b
val pair : 'a repr -> 'b repr -> ('a * 'b) repr
val of_triple : ('a -> valexpr) -> ('b -> valexpr) -> ('c -> valexpr) -> ('a * 'b * 'c) -> valexpr
val to_triple : (valexpr -> 'a) -> (valexpr -> 'b) -> (valexpr -> 'c) -> valexpr -> 'a * 'b * 'c
val triple : 'a repr -> 'b repr -> 'c repr -> ('a * 'b * 'c) repr
val of_option : ('a -> valexpr) -> 'a option -> valexpr
val to_option : (valexpr -> 'a) -> valexpr -> 'a option
val option : 'a repr -> 'a option repr
val of_pattern : Pattern.constr_pattern -> valexpr
val to_pattern : valexpr -> Pattern.constr_pattern
val of_evar : Evar.t -> valexpr
val to_evar : valexpr -> Evar.t
val evar : Evar.t repr
val of_pp : Pp.t -> valexpr
val to_pp : valexpr -> Pp.t
val pp : Pp.t repr
val of_constant : Names.Constant.t -> valexpr
val to_constant : valexpr -> Names.Constant.t
val constant : Names.Constant.t repr
val of_reference : Names.GlobRef.t -> valexpr
val to_reference : valexpr -> Names.GlobRef.t
val reference : Names.GlobRef.t repr
val of_ext : 'a Tac2dyn.Val.tag -> 'a -> valexpr
val to_ext : 'a Tac2dyn.Val.tag -> valexpr -> 'a
val repr_ext : 'a Tac2dyn.Val.tag -> 'a repr
val of_open : (Names.KerName.t * valexpr array) -> valexpr
val to_open : valexpr -> Names.KerName.t * valexpr array
val open_ : (Names.KerName.t * valexpr array) repr
val of_uint63 : Uint63.t -> valexpr
val to_uint63 : valexpr -> Uint63.t
val uint63 : Uint63.t repr
val of_float : Float64.t -> valexpr
val to_float : valexpr -> Float64.t
val float : Float64.t repr
type ('a, 'b) fun1
val app_fun1 : ('a, 'b) fun1 -> 'a repr -> 'b repr -> 'a -> 'b Proofview.tactic
val to_fun1 : 'a repr -> 'b repr -> valexpr -> ('a, 'b) fun1
val fun1 : 'a repr -> 'b repr -> ('a, 'b) fun1 repr
val valexpr : valexpr repr
Dynamic tags
val val_constr : EConstr.t Tac2dyn.Val.tag
val val_ident : Names.Id.t Tac2dyn.Val.tag
val val_matching_context : Constr_matching.context Tac2dyn.Val.tag
val val_evar : Evar.t Tac2dyn.Val.tag
val val_pp : Pp.t Tac2dyn.Val.tag
val val_inductive : Names.inductive Tac2dyn.Val.tag
val val_constructor : Names.constructor Tac2dyn.Val.tag
val val_projection : Names.Projection.t Tac2dyn.Val.tag
val val_free : Names.Id.Set.t Tac2dyn.Val.tag

Toplevel representation of OCaml exceptions. Invariant: no LtacError should be put into a value with tag val_exn.

Closures

val apply : closure -> valexpr list -> valexpr Proofview.tactic

Given a closure, apply it to some arguments. Handling of argument mismatches is done automatically, i.e. in case of over or under-application.

val abstract : int -> (valexpr list -> valexpr Proofview.tactic) -> closure

Turn a fixed-arity function into a closure. The inner function is guaranteed to be applied to a list whose size is the integer argument.

Exception

exception LtacError of Names.KerName.t * valexpr array

Ltac2-defined exceptions seen from OCaml side