simon/ocaml-containers
1
0
Fork 0
mirror of https://github.com/c-cube/ocaml-containers.git synced 2025-12-06 03:05:28 -05:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge pull request #467 from mobotsar/enrich_char

Browse source 
Additional functions for the `Char` module to check common character properties
This commit is contained in:
Simon Cruanes 2025-01-04 17:09:24 -05:00 • committed by GitHub
commit 31ad563044
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 108 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
src/core/CCChar.ml
View file

@ -23,3 +23,12 @@ module Infix = struct
end end
include Infix include Infix
let is_uppercase_ascii c = c > '\064' && c < '\091'
let is_lowercase_ascii c = c > '\096' && c < '\123'
let is_letter_ascii c =
(is_lowercase_ascii [@inlined]) c || (is_uppercase_ascii [@inlined]) c
let is_digit_ascii c = c > '\047' && c < '\058'
let is_whitespace_ascii c = c = '\032' || (c > '\008' && c < '\014')

26
src/core/CCChar.mli
View file

@ -40,6 +40,32 @@ val pp_buf : Buffer.t -> t -> unit
val pp : Format.formatter -> t -> unit val pp : Format.formatter -> t -> unit
(** Renamed from [print] since 2.0. *) (** Renamed from [print] since 2.0. *)
val is_uppercase_ascii : t -> bool
(** [is_uppercase_ascii c] is true exactly when [c] is an
uppercase ASCII character, i.e. ['\064'] < [c] < ['\091'].
@since 3.16 *)
val is_lowercase_ascii : t -> bool
(** [is_lowercase_ascii c] is true exactly when [c] is a
lowercase ASCII character, i.e. ['\096'] < [c] < ['\123'].
@since 3.16 *)
val is_letter_ascii : t -> bool
(** [is_letter_ascii c] is true exactly when [c] is an ASCII
letter, i.e. [is_uppercase_ascii c || is_lowercase_ascii c].
@since 3.16 *)
val is_digit_ascii : t -> bool
(** [is_digit_ascii c] is true exactly when [c] is an
ASCII digit, i.e. ['\047'] < [c] < ['\058'].
@since 3.16 *)
val is_whitespace_ascii : t -> bool
(** [is_whitespace_ascii c] is true exactly when [c] is an ASCII
whitespace character as defined by Unicode, i.e. either [c = ' ']
or ['\008'] < [c] < ['\014'].
@since 3.16 *)
(** {2 Infix Operators} (** {2 Infix Operators}
@since 3.3 *) @since 3.3 *)

73
tests/core/t_char.ml
View file

@ -8,3 +8,76 @@ eq None (of_int 257);;
q q
(Q.string_of_size (Q.Gen.return 1)) (Q.string_of_size (Q.Gen.return 1))
(fun s -> Stdlib.( = ) (to_string s.[0]) s) (fun s -> Stdlib.( = ) (to_string s.[0]) s)
;;
q (Q.int_range 65 90 |> Q.map Char.chr) CCChar.is_uppercase_ascii;;
q
(Q.int_range 0 64 |> Q.map Char.chr)
(fun c -> not @@ CCChar.is_uppercase_ascii c)
;;
q
(Q.int_range 91 127 |> Q.map Char.chr)
(fun c -> not @@ CCChar.is_uppercase_ascii c)
;;
q (Q.int_range 97 122 |> Q.map Char.chr) CCChar.is_lowercase_ascii;;
q
(Q.int_range 0 96 |> Q.map Char.chr)
(fun c -> not @@ CCChar.is_lowercase_ascii c)
;;
q
(Q.int_range 123 127 |> Q.map Char.chr)
(fun c -> not @@ CCChar.is_lowercase_ascii c)
;;
q (Q.int_range 48 57 |> Q.map Char.chr) CCChar.is_digit_ascii;;
q (Q.int_range 0 47 |> Q.map Char.chr) (fun c -> not @@ CCChar.is_digit_ascii c)
;;
q
(Q.int_range 58 127 |> Q.map Char.chr)
(fun c -> not @@ CCChar.is_digit_ascii c)
;;
eq true
(Stdlib.List.for_all CCChar.is_whitespace_ascii
[ '\n'; '\t'; ' '; '\010'; '\011'; '\012'; '\013' ])
;;
eq false
(Stdlib.List.exists CCChar.is_whitespace_ascii
[
'H';
'e';
'l';
'l';
'o';
'!';
'-';
'-';
'N';
'O';
't';
'h';
'i';
'n';
'a';
'\055';
'k';
'a';
'g';
'$';
'$';
'$';
'%';
'^';
'b';
'c';
'h';
'\008';
'h';
])