Preserve unicode format strings

Changes.md

@@ -127,6 +127,8 @@ Unreleased
   ([#995](https://github.com/melange-re/melange/pull/995))
     - PPXes will no longer see the internal `*j` delimiter in unicode strings
       and can hook on either `j` or `js`
+- Preserve unicode in format strings
+  ([#1001](https://github.com/melange-re/melange/pull/1001))
 
 2.2.0 2023-12-05
 ---------------

test/blackbox-tests/unicode-format-strings.t

@@ -0,0 +1,56 @@
+
+  $ . ./setup.sh
+  $ cat > dune-project <<EOF
+  > (lang dune 3.8)
+  > (using melange 0.1)
+  > EOF
+  $ cat > dune <<EOF
+  > (melange.emit
+  >  (target output)
+  >  (alias mel)
+  >  (preprocess (pps melange.ppx)))
+  > EOF
+
+  $ cat > x.ml <<EOF
+  > let () = Js.log (Format.sprintf {js|你 %s|js} "xx")
+  > EOF
+
+  $ dune build @mel
+  $ cat _build/default/output/x.js
+  // Generated by Melange
+  'use strict';
+
+  var Curry = require("melange.js/curry.js");
+  var Stdlib__Format = require("melange/format.js");
+
+  console.log(Curry._1(Stdlib__Format.sprintf(/* Format */{
+                _0: {
+                  TAG: /* String_literal */11,
+                  _0: "你 ",
+                  _1: {
+                    TAG: /* String */2,
+                    _0: /* No_padding */0,
+                    _1: /* End_of_format */0
+                  }
+                },
+                _1: "你 %s"
+              }), "xx"));
+
+  /*  Not a pure module */
+
+  $ node _build/default/output/x.js
+  你 xx
+
+Show that `{j| ... |j}` with interpolation doesn't work because string
+concatenation isn't a literal
+
+  $ cat > x.ml <<EOF
+  > let x = "xx"
+  > let () = Js.log (Format.sprintf {j|你 \$(x) %s|j} "xx")
+  > EOF
+  $ dune build @mel
+  File "x.ml", line 1:
+  Error: This expression has type string but an expression was expected of type
+           ('a -> 'b, unit, string) format =
+             ('a -> 'b, unit, string, string, string, string) format6
+  [1]