google-protobuf 4.28.0.rc.1-arm64-darwin → 4.28.0.rc.2-arm64-darwin

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 428b6110d23efd5f0221381a83edbc9040c28b301ad4e78447a96b4aaee07b99
-  data.tar.gz: 4b7ed698163f9e96f9b41ab01ec02a0b2c0e878111896d5d157fe134032cfee7
+  metadata.gz: 6a43155e5a42dcc8b26c0c21b5df3c06dcd6112cfbd29807dd37e6ba155043d6
+  data.tar.gz: e89143a856d7dd4e0847dc4ab88a653d81f93f2073ddbd527ab4c40cfb24a9a6
 SHA512:
-  metadata.gz: b457f65a745a824a24cee33553a4e9fd07eada3a9a0b74828fec28f25796515b573796d6b93885be83a4003d5d9a25e1ec4c047ce468363218d0fbd9c3f5ec98
-  data.tar.gz: 1382c128d7a25635d959542f0bb900a1901eaa75086683f20dcb770fb3f828c38d16c56def1f62c75d30513f3d7ca30a31b84abf9a52073a1dfea201bb2a4499
+  metadata.gz: 5818e8489f11701254e8dadbe1e46b55ba55e7a8b19e2c60777fba90a21247b2f714d37d71cf426865819dc5dbebe7e248fb8d9de9cfb8195b5752c6792a4943
+  data.tar.gz: a9a0680957c1686891c8989e63be1bf33b404ab2cd041a7dd54d5417532b2f2fcc69ebad05ddac5407f0946e6497eeb8b2a6ef978feea6ec8ab5be0673b9b621

data/ext/google/protobuf_c/convert.c

@@ -104,6 +104,41 @@ unknownval:
   rb_raise(rb_eRangeError, "Unknown symbol value for enum field '%s'.", name);
 }
 
+VALUE Convert_CheckStringUtf8(VALUE str) {
+  VALUE utf8 = rb_enc_from_encoding(rb_utf8_encoding());
+
+  if (rb_obj_encoding(str) == utf8) {
+    // Note: Just because a string is marked as having UTF-8 encoding does
+    // not mean that it is *valid* UTF-8.  We have to check separately
+    // whether it is valid.
+    if (rb_enc_str_coderange(str) == ENC_CODERANGE_BROKEN) {
+      // TODO: For now
+      // we only warn for this case.  We will remove the warning and throw an
+      // exception below in the 30.x release
+
+      rb_warn(
+          "String is invalid UTF-8. This will be an error in a future "
+          "version.");
+      // VALUE exc = rb_const_get_at(
+      //     rb_cEncoding, rb_intern("InvalidByteSequenceError"));
+      // rb_raise(exc, "String is invalid UTF-8");
+    }
+  } else {
+    // Note: this will not duplicate underlying string data unless
+    // necessary.
+    //
+    // This will throw an exception if the conversion cannot be performed:
+    // - Encoding::UndefinedConversionError if certain characters cannot be
+    //   converted to UTF-8.
+    // - Encoding::InvalidByteSequenceError if certain characters were invalid
+    //   in the source encoding.
+    str = rb_str_encode(str, utf8, 0, Qnil);
+    PBRUBY_ASSERT(rb_enc_str_coderange(str) != ENC_CODERANGE_BROKEN);
+  }
+
+  return str;
+}
+
 upb_MessageValue Convert_RubyToUpb(VALUE value, const char* name,
                                    TypeInfo type_info, upb_Arena* arena) {
   upb_MessageValue ret;
@@ -137,8 +172,7 @@ upb_MessageValue Convert_RubyToUpb(VALUE value, const char* name,
       }
       break;
     }
-    case kUpb_CType_String: {
-      VALUE utf8 = rb_enc_from_encoding(rb_utf8_encoding());
+    case kUpb_CType_String:
       if (rb_obj_class(value) == rb_cSymbol) {
         value = rb_funcall(value, rb_intern("to_s"), 0);
       } else if (!rb_obj_is_kind_of(value, rb_cString)) {
@@ -147,19 +181,9 @@ upb_MessageValue Convert_RubyToUpb(VALUE value, const char* name,
                  rb_class2name(CLASS_OF(value)));
       }
 
-      if (rb_obj_encoding(value) != utf8) {
-        // Note: this will not duplicate underlying string data unless
-        // necessary.
-        value = rb_str_encode(value, utf8, 0, Qnil);
-
-        if (rb_enc_str_coderange(value) == ENC_CODERANGE_BROKEN) {
-          rb_raise(rb_eEncodingError, "String is invalid UTF-8");
-        }
-      }
-
+      value = Convert_CheckStringUtf8(value);
       ret.str_val = Convert_StringData(value, arena);
       break;
-    }
     case kUpb_CType_Bytes: {
       VALUE bytes = rb_enc_from_encoding(rb_ascii8bit_encoding());
       if (rb_obj_class(value) != rb_cString) {

data/lib/google/protobuf/ffi/internal/convert.rb

@@ -33,11 +33,18 @@ module Google
             return_value[:bool_val] = value
           when :string
             raise TypeError.new "Invalid argument for string field '#{name}' (given #{value.class})." unless value.is_a?(String) or value.is_a?(Symbol)
-            begin
+            value = value.to_s if value.is_a?(Symbol)
+            if value.encoding == Encoding::UTF_8
+              unless value.valid_encoding?
+                # TODO:
+                # For now we only warn for this case.  We will remove the
+                # warning and throw an exception below in the 30.x release
+                warn "String is invalid UTF-8. This will be an error in a future version."
+                # raise Encoding::InvalidByteSequenceError.new "String is invalid UTF-8"
+              end
+              string_value = value
+            else
               string_value = value.to_s.encode("UTF-8")
-            rescue Encoding::UndefinedConversionError
-              # TODO - why not include the field name here?
-              raise Encoding::UndefinedConversionError.new "String is invalid UTF-8"
             end
             return_value[:str_val][:size] = string_value.bytesize
             return_value[:str_val][:data] = Google::Protobuf::FFI.arena_malloc(arena, string_value.bytesize)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: google-protobuf
 version: !ruby/object:Gem::Version
-  version: 4.28.0.rc.1
+  version: 4.28.0.rc.2
 platform: arm64-darwin
 authors:
 - Protobuf Authors
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-07-11 00:00:00.000000000 Z
+date: 2024-08-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake-compiler-dock
@@ -190,7 +190,7 @@ homepage: https://developers.google.com/protocol-buffers
 licenses:
 - BSD-3-Clause
 metadata:
-  source_code_uri: https://github.com/protocolbuffers/protobuf/tree/v4.28.0-rc1/ruby
+  source_code_uri: https://github.com/protocolbuffers/protobuf/tree/v4.28.0-rc2/ruby
 post_install_message: 
 rdoc_options: []
 require_paths: