only allow certain characters after interpolated vars

NEWS.md

@@ -167,7 +167,12 @@ Language changes
     backslashes and the end of the literal while 2n+1 backslashes followed by a quote encodes n
     backslashes followed by a quote character ([#22926]).
 
-  * The syntax `(x...)` for constructing a tuple is deprecated; use `(x...,)` instead (#24452).
+  * The syntax `(x...)` for constructing a tuple is deprecated; use `(x...,)` instead ([#24452]).
+
+  * Non-parenthesized interpolated variables in strings, e.g. `"$x"`, must be followed
+    by a character that will never be an allowed identifier character (currently
+    operators, space/control characters, or common punctuation characters) ([#25231]).
+
 
 Breaking changes
 ----------------

src/flisp/julia_extensions.c

@@ -162,6 +162,21 @@ int jl_op_suffix_char(uint32_t wc)
     return HT_NOTFOUND != wcharhash_get_r(&jl_opsuffs, (void*)((uintptr_t)wc), NULL);
 }
 
+// chars that we will never allow to be part of a valid non-operator identifier
+static int never_id_char(uint32_t wc)
+{
+     utf8proc_category_t cat = utf8proc_category((utf8proc_int32_t) wc);
+     return (
+          // spaces and control characters:
+          (cat >= UTF8PROC_CATEGORY_ZS && cat <= UTF8PROC_CATEGORY_CS) ||
+
+          // ASCII and Latin1 non-connector punctuation
+          (wc < 0xff &&
+           cat >= UTF8PROC_CATEGORY_PD && cat <= UTF8PROC_CATEGORY_PO) ||
+
+          wc == '`');
+}
+
 value_t fl_julia_identifier_char(fl_context_t *fl_ctx, value_t *args, uint32_t nargs)
 {
     argcount(fl_ctx, "identifier-char?", nargs, 1);
@@ -180,6 +195,16 @@ value_t fl_julia_identifier_start_char(fl_context_t *fl_ctx, value_t *args, uint
     return jl_id_start_char(wc) ? fl_ctx->T : fl_ctx->F;
 }
 
+value_t fl_julia_never_identifier_char(fl_context_t *fl_ctx, value_t *args, uint32_t nargs)
+{
+    argcount(fl_ctx, "never-identifier-char?", nargs, 1);
+    if (!iscprim(args[0]) || ((cprim_t*)ptr(args[0]))->type != fl_ctx->wchartype)
+        type_error(fl_ctx, "never-identifier-char?", "wchar", args[0]);
+    uint32_t wc = *(uint32_t*)cp_data((cprim_t*)ptr(args[0]));
+    return never_id_char(wc) ? fl_ctx->T : fl_ctx->F;
+}
+
+
 value_t fl_julia_op_suffix_char(fl_context_t *fl_ctx, value_t *args, uint32_t nargs)
 {
     argcount(fl_ctx, "op-suffix-char?", nargs, 1);
@@ -307,6 +332,7 @@ static const builtinspec_t julia_flisp_func_info[] = {
     { "accum-julia-symbol", fl_accum_julia_symbol },
     { "identifier-char?", fl_julia_identifier_char },
     { "identifier-start-char?", fl_julia_identifier_start_char },
+    { "never-identifier-char?", fl_julia_never_identifier_char },
     { "op-suffix-char?", fl_julia_op_suffix_char },
     { "strip-op-suffix", fl_julia_strip_op_suffix },
     { "underscore-symbol?", fl_julia_underscore_symbolp },

src/julia-parser.scm

@@ -2072,11 +2072,18 @@
 (define (string-replace s a b)
   (string.join (string-split s a) b))
 
+(define (ends-interpolated-atom? c)
+  (or (eof-object? c) (opchar? c) (never-identifier-char? c)))
+
 (define (parse-interpolate s)
   (let* ((p (ts:port s))
          (c (peek-char p)))
     (cond ((identifier-start-char? c)
-           (parse-atom s))
+           (let* ((atom (parse-atom s))
+                  (c (peek-char p)))
+             (if (ends-interpolated-atom? c)
+                 atom
+                 (error (string "interpolated variable $" atom " ends with invalid character \"" c "\"; use \"$(" atom ")\" instead.")))))
           ((eqv? c #\()
            (read-char p)
            (let ((ex (parse-eq* s))

test/parse.jl

@@ -258,3 +258,7 @@ end
 # added ⟂ to operator precedence (#24404)
 @test Meta.parse("a ⟂ b ⟂ c") == Expr(:comparison, :a, :⟂, :b, :⟂, :c)
 @test Meta.parse("a ⟂ b ∥ c") == Expr(:comparison, :a, :⟂, :b, :∥, :c)
+
+# only allow certain characters after interpolated vars (#25231)
+@test Meta.parse("\"\$x෴  \"",raise=false) == Expr(:error, "interpolated variable \$x ends with invalid character \"෴\"; use \"\$(x)\" instead.")
+@test Base.incomplete_tag(Meta.parse("\"\$foo", raise=false)) == :string