From 12743ab24f1d5389a09adabfb0513c7c4b164c10 Mon Sep 17 00:00:00 2001
From: Simon Cruanes <simon.cruanes.2007@m4x.org>
Date: Thu, 2 Feb 2017 21:27:46 +0100
Subject: [PATCH] add `count`

---
 src/Sequence.ml        | 22 ++++++++++++++++++++++
 src/Sequence.mli       |  6 ++++++
 src/sequenceLabels.mli |  6 ++++++
 3 files changed, 34 insertions(+)

diff --git a/src/Sequence.ml b/src/Sequence.ml
index 425bf8d..d36221d 100644
--- a/src/Sequence.ml
+++ b/src/Sequence.ml
@@ -411,6 +411,28 @@ let group_by (type k) ?(hash=Hashtbl.hash) ?(eq=(=)) seq =
     |> OUnit.assert_equal [[1];[2;2;2];[3;3;3];[4]]
 *)
 
+let count (type k) ?(hash=Hashtbl.hash) ?(eq=(=)) seq =
+  let module Tbl = Hashtbl.Make(struct
+      type t = k
+      let equal = eq
+      let hash = hash
+    end) in
+  (* compute group table *)
+  let tbl = Tbl.create 32 in
+  seq
+    (fun x ->
+       let n = try Tbl.find tbl x with Not_found -> 0 in
+       Tbl.replace tbl x (n+1)
+    );
+  fun yield ->
+    Tbl.iter (fun x n -> yield (x,n)) tbl
+
+(*$R
+  [1;2;3;3;2;2;3;4]
+    |> of_list |> count ?eq:None ?hash:None |> sort ?cmp:None |> to_list
+    |> OUnit.assert_equal [1,1;2,3;3,3;4,1]
+*)
+
 let uniq ?(eq=fun x y -> x = y) seq k =
   let has_prev = ref false
   and prev = ref (Obj.magic 0) in  (* avoid option type, costly *)
diff --git a/src/Sequence.mli b/src/Sequence.mli
index e7c5e3d..872289d 100644
--- a/src/Sequence.mli
+++ b/src/Sequence.mli
@@ -241,6 +241,12 @@ val group_by : ?hash:('a -> int) -> ?eq:('a -> 'a -> bool) ->
     The result sequence is traversable as many times as required.
     @since 0.6 *)
 
+val count : ?hash:('a -> int) -> ?eq:('a -> 'a -> bool) ->
+  'a t -> ('a * int) t
+(** Map each distinct element to its number of occurrences in the whole seq.
+    Similar to [group_by seq |> map (fun l->List.hd l, List.length l)]
+    @since NEXT_RELEASE *)
+
 val uniq : ?eq:('a -> 'a -> bool) -> 'a t -> 'a t
 (** Remove consecutive duplicate elements. Basically this is
     like [fun seq -> map List.hd (group seq)]. *)
diff --git a/src/sequenceLabels.mli b/src/sequenceLabels.mli
index 06bd60f..2829be4 100644
--- a/src/sequenceLabels.mli
+++ b/src/sequenceLabels.mli
@@ -212,6 +212,12 @@ val group_by : ?hash:('a -> int) -> ?eq:('a -> 'a -> bool) ->
     The result sequence is traversable as many times as required.
     @since 0.6 *)
 
+val count : ?hash:('a -> int) -> ?eq:('a -> 'a -> bool) ->
+  'a t -> ('a * int) t
+(** Map each distinct element to its number of occurrences in the whole seq.
+    Similar to [group_by seq |> map (fun l->List.hd l, List.length l)]
+    @since NEXT_RELEASE *)
+
 val uniq : ?eq:('a -> 'a -> bool) -> 'a t -> 'a t
 (** Remove consecutive duplicate elements. Basically this is
     like [fun seq -> map List.hd (group seq)]. *)