Fix GC compatibility: Don't stash encodings in global constants

This value should either be pinned, or looked up when needed at runtime.
Without pinning, the GC may move the encoding object, and that could
cause a crash.

In this case it is easier to find the value at runtime, and there is no
performance penalty (as Ruby caches encoding indexes).  We can shorten
the code, be compaction friendly, and incur no performance penalty.

ext/json/generator/generator.c

@@ -1,11 +1,6 @@
 #include "../fbuffer/fbuffer.h"
 #include "generator.h"
 
-#ifdef HAVE_RUBY_ENCODING_H
-static VALUE CEncoding_UTF_8;
-static ID i_encoding, i_encode;
-#endif
-
 static VALUE mJSON, mExt, mGenerator, cState, mGeneratorMethods, mObject,
              mHash, mArray,
 #ifdef RUBY_INTEGER_UNIFICATION
@@ -948,7 +943,7 @@ static void generate_json_string(FBuffer *buffer, VALUE Vstate, JSON_Generator_S
     fbuffer_append_char(buffer, '"');
 #ifdef HAVE_RUBY_ENCODING_H
     if (!enc_utf8_compatible_p(rb_enc_get(obj))) {
-        obj = rb_str_encode(obj, CEncoding_UTF_8, 0, Qnil);
+        obj = rb_str_export_to_enc(obj, rb_utf8_encoding());
     }
 #endif
     if (state->ascii_only) {
@@ -1610,9 +1605,4 @@ void Init_generator(void)
     i_match = rb_intern("match");
     i_keys = rb_intern("keys");
     i_dup = rb_intern("dup");
-#ifdef HAVE_RUBY_ENCODING_H
-    CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8"));
-    i_encoding = rb_intern("encoding");
-    i_encode = rb_intern("encode");
-#endif
 }