Simplified handling of UTF-8 symbols in YAML

We actually don't need to store the encoding, we can assume
everything is UTF-8. `Symbol#to_sym` will take care of coercing
to ASCII when applicable.

Any other encoding is not supported as the cache should be invoked
if Psych could generate any other encoding.

CHANGELOG.md

@@ -1,6 +1,6 @@
 # Unreleased
 
-* Improve the YAML compile cache to support `UTF-8` symbols. (#398)
+* Improve the YAML compile cache to support `UTF-8` symbols. (#398, #399)
   [The default `MessagePack` symbol serializer assumes all symbols are ASCII](https://github.com/msgpack/msgpack-ruby/pull/211),
   because of this, non-ASCII compatible symbol would be restored with `ASCII_8BIT` encoding (AKA `BINARY`).
   Bootsnap now properly cache them in `UTF-8`.

ext/bootsnap/bootsnap.c

@@ -2,7 +2,7 @@
  * Suggested reading order:
  * 1. Skim Init_bootsnap
  * 2. Skim bs_fetch
- * 3. The rest of everything
+ * 3. The rest of everyrything
  *
  * Init_bootsnap sets up the ruby objects and binds bs_fetch to
  * Bootsnap::CompileCache::Native.fetch.
@@ -75,7 +75,7 @@ struct bs_cache_key {
 STATIC_ASSERT(sizeof(struct bs_cache_key) == KEY_SIZE);
 
 /* Effectively a schema version. Bumping invalidates all previous caches */
-static const uint32_t current_version = 5;
+static const uint32_t current_version = 4;
 
 /* hash of e.g. "x86_64-darwin17", invalidating when ruby is recompiled on a
  * new OS ABI, etc. */

lib/bootsnap/compile_cache/yaml.rb

@@ -49,31 +49,8 @@ def supported_internal_encoding?
         module EncodingAwareSymbols
           extend self
 
-          if Symbol.method_defined?(:name)
-            def pack(symbol)
-              if symbol.encoding == Encoding::UTF_8
-                1.chr << symbol.name
-              else
-                0.chr << symbol.name
-              end
-            end
-          else
-            def pack(symbol)
-              if symbol.encoding == Encoding::UTF_8
-                1.chr << symbol.to_s
-              else
-                0.chr << symbol.to_s
-              end
-            end
-          end
-
           def unpack(payload)
-            payload.freeze
-            string = payload.byteslice(1..-1)
-            if payload.ord == 1 # Encoding::UTF_8
-              string.force_encoding(Encoding::UTF_8)
-            end
-            string.to_sym
+            payload.force_encoding(Encoding::UTF_8).to_sym
           end
         end
 
@@ -94,7 +71,7 @@ def init!
           factory.register_type(
             0x00,
             Symbol,
-            packer: EncodingAwareSymbols.method(:pack).to_proc,
+            packer: :to_msgpack_ext,
             unpacker: EncodingAwareSymbols.method(:unpack).to_proc,
           )
 

test/compile_cache_key_format_test.rb

@@ -21,7 +21,7 @@ class CompileCacheKeyFormatTest < Minitest::Test
 
   def test_key_version
     key = cache_key_for_file(FILE)
-    exp = [5].pack("L")
+    exp = [4].pack("L")
     assert_equal(exp, key[R[:version]])
   end