CA-361221: utf8_recode: use Uutf.{Buffer.add_utf_8,String.fold_utf_8} instead of Uutf.{encode,decoder}

Uutf.encoder allocates a 64k buffer internally (in addition to our 1k buffer).
Use Uutf.Buffer to write utf8 chars directly into our buffer instead.
Also size the buffer based on initial string size (it can grow if needed).

This should reduce allocation rate.

Thanks to a hint from https://discuss.ocaml.org/t/decoding-many-unicode-strings-with-uutf/8910/2

Also avoid allocations if string is all utf8

This could be further optimized to stop on first invalid utf8 char and
do the recode on the rest.

Signed-off-by: Edwin Török <edvin.torok@citrix.com>

Author: edwintorok

lib/xenops_utils.ml

@@ -91,22 +91,21 @@ let ignore_bool (_: bool) = ()
 let ignore_int (_: int) = ()
 
 (* Recode an incoming string as valid UTF-8 *)
+let utf8_add_if_valid buf _ next =
+  let uchar = match next with `Malformed _ -> Uutf.u_rep | `Uchar u -> u in
+  Uutf.Buffer.add_utf_8 buf uchar ;
+  buf
+
+let is_valid prev _ = function `Malformed _ -> false | `Uchar _ -> prev
+
 let utf8_recode str =
-	let out_encoding = `UTF_8 in
-	let b = Buffer.create 1024 in
-	let dst = `Buffer b in
-	let src = `String str in
-	let rec loop d e =
-		match Uutf.decode d with
-		| `Uchar _ as u -> ignore (Uutf.encode e u); loop d e
-		| `End -> ignore (Uutf.encode e `End)
-		|`Malformed _ -> ignore (Uutf.encode e (`Uchar Uutf.u_rep)); loop d e
-		| `Await -> assert false
-	in
-	let d = Uutf.decoder src in
-	let e = Uutf.encoder out_encoding dst in
-	loop d e;
-	Buffer.contents b
+  (* optimistic assumption that the string contains only valid utf8,
+     avoids allocations *)
+  if Uutf.String.fold_utf_8 is_valid true str then
+    str
+  else
+    let b = Buffer.create (String.length str) in
+    Uutf.String.fold_utf_8 utf8_add_if_valid b str |> Buffer.contents
 
 module Mutex = struct
 	include Mutex