Utility Module for S-expression Conversions
default_string_of_float
reference to the default function used
to convert floats to strings.
Initially set to fun n -> sprintf "%.20G" n
.
write_old_option_format
reference for the default option format
used to write option values. If set to true
, the old-style option
format will be used, the new-style one otherwise.
Initially set to true
.
read_old_option_format
reference for the default option format
used to read option values. Of_sexp_error
will be raised
with old-style option values if this reference is set to false
.
Reading new-style option values is always supported. Using a global
reference instead of changing the converter calling conventions is
the only way to avoid breaking old code with the standard macros.
Initially set to true
.
We re-export a tail recursive map function, because some modules
override the standard library functions (e.g. StdLabels
) which
wrecks havoc with the camlp4 extension.
sexp_of_unit ()
converts a value of type unit
to an S-expression.
sexp_of_bool b
converts the value x
of type bool
to an
S-expression.
sexp_of_bool str
converts the value str
of type string
to an
S-expression.
sexp_of_char c
converts the value c
of type char
to an
S-expression.
sexp_of_int n
converts the value n
of type int
to an
S-expression.
sexp_of_float n
converts the value n
of type float
to an
S-expression.
sexp_of_int32 n
converts the value n
of type int32
to an
S-expression.
sexp_of_int64 n
converts the value n
of type int64
to an
S-expression.
sexp_of_nativeint n
converts the value n
of type nativeint
to an
S-expression.
sexp_of_hashtbl conv_key conv_value htbl
converts the value htbl
of type ('a, 'b) Hashtbl.t
to an S-expression. Uses conv_key
to convert the hashtable keys of type 'a
, and conv_value
to
convert hashtable values of type 'b
to S-expressions.
sexp_of_float32_vec vec
converts the one-dimensional bigarray
vec
of 32-bit floats in Fortran-layout to an S-expression.
sexp_of_float64_vec vec
converts the one-dimensional bigarray
vec
of 64-bit floats in Fortran-layout to an S-expression.
sexp_of_vec vec
same as Conv.sexp_of_float64_vec.
sexp_of_float32_mat mat
converts the two-dimensional bigarray
mat
of 32-bit floats in Fortran-layout to an S-expression.
sexp_of_float64_mat mat
converts the two-dimensional bigarray
mat
of 64-bit floats in Fortran-layout to an S-expression.
sexp_of_mat mat
same as Conv.sexp_of_float64_mat.
sexp_of_opaque x
converts the value x
of opaque type to an
S-expression. This means the user need not provide converters,
but the result cannot be interpreted.
sexp_of_fun f
converts the value f
of function type to a
dummy S-expression. Functions cannot be serialized as S-expressions,
but at least a placeholder can be generated for pretty-printing.
string_of__of__sexp_of conv x
converts the OCaml-value x
to
an S-expression represented as a string by using conversion function
conv
.
Of_sexp_error (exn, sexp)
the exception raised when an S-expression
could not be successfully converted to an OCaml-value.
record_check_extra_fields
checks for extra (= unknown) fields
in record S-expressions.
of_sexp_error reason sexp
of_sexp_error exc sexp
unit_of_sexp sexp
converts S-expression sexp
to a value of type
unit
.
bool_of_sexp sexp
converts S-expression sexp
to a value of type
bool
.
string_of_sexp sexp
converts S-expression sexp
to a value of type
string
.
char_of_sexp sexp
converts S-expression sexp
to a value of type
char
.
int_of_sexp sexp
converts S-expression sexp
to a value of type
int
.
float_of_sexp sexp
converts S-expression sexp
to a value of type
float
.
int32_of_sexp sexp
converts S-expression sexp
to a value of type
int32
.
int64_of_sexp sexp
converts S-expression sexp
to a value of type
int64
.
nativeint_of_sexp sexp
converts S-expression sexp
to a value
of type nativeint
.
triple_of_sexp conv1 conv2 conv3 sexp
converts S-expression sexp
to a triple of type 'a * 'b * 'c
using conversion functions conv1
,
conv2
, and conv3
, which convert S-expressions to values of type
'a
, 'b
, and 'c
respectively.
hashtbl_of_sexp conv_key conv_value sexp
converts S-expression
sexp
to a value of type ('a, 'b) Hashtbl.t
using conversion
function conv_key
, which converts an S-expression to hashtable
key of type 'a
, and function conv_value
, which converts an
S-expression to hashtable value of type 'b
.
float32_vec_of_sexp sexp
converts S-expression sexp
to a
one-dimensional bigarray of 32-bit floats in Fortran-layout.
float64_vec_of_sexp sexp
converts S-expression sexp
to a
one-dimensional bigarray of 64-bit floats in Fortran-layout.
vec_of_sexp sexp
same as float64_vec_of_sexp.
float32_mat_of_sexp sexp
converts S-expression sexp
to a
two-dimensional bigarray of 32-bit floats in Fortran-layout.
float64_mat_of_sexp sexp
converts S-expression sexp
to a
two-dimensional bigarray of 64-bit floats in Fortran-layout.
mat_of_sexp sexp
same as Conv.float64_mat_of_sexp.
opaque_of_sexp sexp
fun_of_sexp sexp
of_string__of__of_sexp conv str
converts the S-expression str
represented as a string to an OCaml-value by using conversion function
conv
.
Exception converters
sexp_of_exn exc
converts exception exc
to an S-expression.
If no suitable converter is found, the standard converter in
Printexc
will be used to generate an atomic S-expression.
sexp_of_exn_opt exc
converts exception exc
to Some sexp
.
If no suitable converter is found, None
is returned instead.