google-protobuf 4.28.0.rc.1-x86-linux → 4.28.0.rc.2-x86-linux

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bda9338435ea6c549627a7863f50aba8e07fedbb34e2f3afceb443a9ad272b90
-  data.tar.gz: c4a044023592c3f28dc543f4916aa4e6f80c9cc42c1851da2dec9eaca00de27a
+  metadata.gz: 14d2fcfa5ca6e62e1d5e65be5cf84af2e507abbaf46a76bd7f7250eb00476e97
+  data.tar.gz: 44f37b0d25166e3553c1e32d3f6847375e9e6430a30984fd75eadac49406ebfc
 SHA512:
-  metadata.gz: 700a7d87c55a937efbc945b16f65f4b036823a7f88b7e0e425ace636882197bc7dbf410454c3c87615982bc47e35a1283482a570523970caebd44204e221d4a8
-  data.tar.gz: a659dcfc8e13f7029126dc74708c46716e42547b0e17926d64778480cabddab7574760d0cb56d8112497e52fb2a1d6b2576ba4c8b57fd1c371c02eeeff34c4d6
+  metadata.gz: 928b6d8251afef7ff1846ca52997f8606e28a4c2930dec259b6fd8543c72b08afdb2d16980a2821642936983bd011a69015b4ce21a4119b657b1211d483084e3
+  data.tar.gz: 9affa82d9e7995349b02d9f5954da9f24ebc6b0cb3a38e168ded81d754bfd093d4f7b8adcb8a087de405a8fae546720f1f801bf333728d75c82e8234cedf1ff7

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: x86-linux
 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: