google-protobuf 4.28.0.rc.1-x64-mingw-ucrt → 4.28.0.rc.2-x64-mingw-ucrt

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 966eae41c608975fd8b13468491f4074c34e78d536470d7ab1518edf5c2f12d7
-  data.tar.gz: 936f14dd49b32c9d4d82ede6f6d16992494a91387c25ce00b484f65e0cf0621f
+  metadata.gz: 1b12a1d072dfe2812a3f37698dafb0d9aaa3eef4e69c099d5e96f3b387b5dc6e
+  data.tar.gz: 4dc0c5c947e047cb031a50203afed3b39ab4bfe787bf3a122b19e76ac9aa27ac
 SHA512:
-  metadata.gz: 7afc3ae0b87c39def2ba218dc11c682db9789c56fcf6f7196bd9caf3b5bef123f280f7c58b6845bf3149792c6ede7330813132e5fb864cc1a8e326ee3db3e13f
-  data.tar.gz: 0526b18d5ad8d696586a627358b64cf03da6225003f79d928351415cf48c8cbebdf897c6183fadeedb0039db2dbb50cfcb51c3de13c0a8de60cd55f2e3485098
+  metadata.gz: 340017c3e664516359924c79a920b5db8d4b3fe26660e37c77fb484aec6b15546a0d6b31449d747ca5fc47470a098ecde9d83f616babb712b506a68edeba67f7
+  data.tar.gz: 86bce0e29e740d6bbf15705551f1069af2c6e5413285fbd0da5796d081f4e698761240d386d42c1def628d139229a6412010e4af8f30dc816da0cd3e45b6172a

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: x64-mingw-ucrt
 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
@@ -189,7 +189,7 @@ 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: