Remove dead _one_shot parameter from _make_iterencode

On main, _one_shot=True in iterencode() selected the C encoder path via
"if _one_shot and c_make_encoder is not None". On this branch the C path
is always taken when available, so _one_shot never controls anything.
_make_iterencode always returned _iterencode_once regardless of _one_shot,
so the parameter was dead there too. Remove it.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: eendebakpt

Lib/json/encoder.py

@@ -243,8 +243,6 @@ def iterencode(self, o, _one_shot=False):
         else:
             indent = ' ' * self.indent
         if c_make_encoder is not None:
-            # _one_shot is irrelevant on the C path: the C iterator has no
-            # mutually-recursive closure cycles to break.
             _iterencode = c_make_encoder(
                 markers, self.default, _encoder, indent,
                 self.key_separator, self.item_separator, self.sort_keys,
@@ -276,11 +274,11 @@ def floatstr(o, allow_nan=self.allow_nan,
             _iterencode = _make_iterencode(
                 markers, self.default, _encoder, indent, floatstr,
                 self.key_separator, self.item_separator, self.sort_keys,
-                self.skipkeys, _one_shot)
+                self.skipkeys)
             return _iterencode(o, 0)
 
 def _make_iterencode(markers, _default, _encoder, _indent, _floatstr,
-        _key_separator, _item_separator, _sort_keys, _skipkeys, _one_shot,
+        _key_separator, _item_separator, _sort_keys, _skipkeys,
     ):
 
     def _iterencode_list(lst, _current_indent_level):