ProtobufSingle input format

Allows to read Single protobuf message at once (w/o length-delemeters).

src/Core/Settings.h

@@ -398,7 +398,12 @@ class IColumn;
     M(Bool, force_optimize_skip_unused_shards_no_nested, false, "Obsolete setting, does nothing. Will be removed after 2020-12-01. Use force_optimize_skip_unused_shards_nesting instead.", 0) \
     M(Bool, experimental_use_processors, true, "Obsolete setting, does nothing. Will be removed after 2020-11-29.", 0) \
     M(Bool, optimize_trivial_insert_select, true, "Optimize trivial 'INSERT INTO table SELECT ... FROM TABLES' query", 0) \
-    M(Bool, allow_experimental_database_atomic, true, "Obsolete setting, does nothing. Will be removed after 2021-02-12", 0)
+    M(Bool, allow_experimental_database_atomic, true, "Obsolete setting, does nothing. Will be removed after 2021-02-12", 0) \
+    M(Bool, allow_non_metadata_alters, true, "Allow to execute alters which affects not only tables metadata, but also data on disk", 0) \
+    M(Bool, enable_global_with_statement, false, "Propagate WITH statements to UNION queries and all subqueries", 0)
+
+// End of COMMON_SETTINGS
+// !!! please add settings related to formats into the FORMAT_FACTORY_SETTINGS below !!!
 
 #define FORMAT_FACTORY_SETTINGS(M) \
     M(Char, format_csv_delimiter, ',', "The character to be considered as a delimiter in CSV data. If setting with a string, a string has to have a length of 1.", 0) \
@@ -462,9 +467,10 @@ class IColumn;
     \
     M(Bool, output_format_enable_streaming, false, "Enable streaming in output formats that support it.", 0) \
     M(Bool, output_format_write_statistics, true, "Write statistics about read rows, bytes, time elapsed in suitable output formats.", 0) \
-    M(Bool, allow_non_metadata_alters, true, "Allow to execute alters which affects not only tables metadata, but also data on disk", 0) \
-    M(Bool, output_format_pretty_row_numbers, false, "Add row numbers before each row for pretty output format", 0) \
-    M(Bool, enable_global_with_statement, false, "Propagate WITH statements to UNION queries and all subqueries", 0) \
+    M(Bool, output_format_pretty_row_numbers, false, "Add row numbers before each row for pretty output format", 0)
+
+// End of FORMAT_FACTORY_SETTINGS
+// !!! please add settings non-related to formats into the COMMON_SETTINGS above !!!
 
 #define LIST_OF_SETTINGS(M)    \
     COMMON_SETTINGS(M)         \

src/Formats/ProtobufReader.cpp

@@ -38,36 +38,47 @@ namespace
     // Those inequations helps checking conditions in ProtobufReader::SimpleReader.
     constexpr Int64 END_OF_VARINT = -1;
     constexpr Int64 END_OF_GROUP = -2;
+    constexpr Int64 END_OF_FILE = -3;
 
     Int64 decodeZigZag(UInt64 n) { return static_cast<Int64>((n >> 1) ^ (~(n & 1) + 1)); }
 
-    [[noreturn]] void throwUnknownFormat()
-    {
-        throw Exception("Protobuf messages are corrupted or don't match the provided schema. Please note that Protobuf stream is length-delimited: every message is prefixed by its length in varint.", ErrorCodes::UNKNOWN_PROTOBUF_FORMAT);
-    }
 }
 
 
 // SimpleReader is an utility class to deserialize protobufs.
 // Knows nothing about protobuf schemas, just provides useful functions to deserialize data.
-ProtobufReader::SimpleReader::SimpleReader(ReadBuffer & in_)
+ProtobufReader::SimpleReader::SimpleReader(ReadBuffer & in_, const bool single_message_mode_)
     : in(in_)
     , cursor(0)
     , current_message_level(0)
     , current_message_end(0)
     , field_end(0)
     , last_string_pos(-1)
+    , single_message_mode(single_message_mode_)
 {
 }
 
+[[noreturn]] void ProtobufReader::SimpleReader::throwUnknownFormat()
+{
+    throw Exception(std::string("Protobuf messages are corrupted or don't match the provided schema.") + (single_message_mode ? "" : " Please note that Protobuf stream is length-delimited: every message is prefixed by its length in varint."), ErrorCodes::UNKNOWN_PROTOBUF_FORMAT);
+}
+
 bool ProtobufReader::SimpleReader::startMessage()
 {
     // Start reading a root message.
     assert(!current_message_level);
     if (unlikely(in.eof()))
         return false;
+
+    if (single_message_mode)
+    {
+        current_message_end = END_OF_FILE;
+    }
+    else
+    {
         size_t size_of_message = readVarint();
         current_message_end = cursor + size_of_message;
+    }
     ++current_message_level;
     field_end = cursor;
     return true;
@@ -150,7 +161,15 @@ bool ProtobufReader::SimpleReader::readFieldNumber(UInt32 & field_number)
             throwUnknownFormat();
     }
 
-    if ((cursor >= current_message_end) && (current_message_end != END_OF_GROUP))
+    if (current_message_end == END_OF_FILE)
+    {
+        if (unlikely(in.eof()))
+        {
+            current_message_end = cursor;
+            return false;
+        }
+    }
+    else if ((cursor >= current_message_end) && (current_message_end != END_OF_GROUP))
         return false;
 
     UInt64 varint = readVarint();
@@ -1077,8 +1096,8 @@ std::unique_ptr<ProtobufReader::IConverter> ProtobufReader::createConverter<goog
 
 
 ProtobufReader::ProtobufReader(
-    ReadBuffer & in_, const google::protobuf::Descriptor * message_type, const std::vector<String> & column_names)
-    : simple_reader(in_)
+    ReadBuffer & in_, const google::protobuf::Descriptor * message_type, const std::vector<String> & column_names, const bool single_message_mode_)
+    : simple_reader(in_, single_message_mode_)
 {
     root_message = ProtobufColumnMatcher::matchColumns<ColumnMatcherTraits>(column_names, message_type);
     setTraitsDataAfterMatchingColumns(root_message.get());

src/Formats/ProtobufReader.h

@@ -37,7 +37,7 @@ using AggregateFunctionPtr = std::shared_ptr<IAggregateFunction>;
 class ProtobufReader : private boost::noncopyable
 {
 public:
-    ProtobufReader(ReadBuffer & in_, const google::protobuf::Descriptor * message_type, const std::vector<String> & column_names);
+    ProtobufReader(ReadBuffer & in_, const google::protobuf::Descriptor * message_type, const std::vector<String> & column_names, const bool single_message_mode_);
     ~ProtobufReader();
 
     /// Should be called when we start reading a new message.
@@ -93,7 +93,7 @@ private:
     class SimpleReader
     {
     public:
-        SimpleReader(ReadBuffer & in_);
+        SimpleReader(ReadBuffer & in_, const bool single_message_mode_);
         bool startMessage();
         void endMessage(bool ignore_errors);
         void startNestedMessage();
@@ -126,6 +126,7 @@ private:
         UInt64 continueReadingVarint(UInt64 first_byte);
         void ignoreVarint();
         void ignoreGroup();
+        [[noreturn]] void throwUnknownFormat();
 
         ReadBuffer & in;
         Int64 cursor;
@@ -134,6 +135,7 @@ private:
         std::vector<Int64> parent_message_ends;
         Int64 field_end;
         Int64 last_string_pos;
+        const bool single_message_mode;
     };
 
     class IConverter

src/Processors/Formats/Impl/ProtobufRowInputFormat.cpp

@@ -11,10 +11,10 @@
 namespace DB
 {
 
-ProtobufRowInputFormat::ProtobufRowInputFormat(ReadBuffer & in_, const Block & header_, Params params_, const FormatSchemaInfo & info_)
+ProtobufRowInputFormat::ProtobufRowInputFormat(ReadBuffer & in_, const Block & header_, Params params_, const FormatSchemaInfo & info_, const bool single_message_mode_)
     : IRowInputFormat(header_, in_, params_)
     , data_types(header_.getDataTypes())
-    , reader(in, ProtobufSchemas::instance().getMessageTypeForFormatSchema(info_), header_.getNames())
+    , reader(in, ProtobufSchemas::instance().getMessageTypeForFormatSchema(info_), header_.getNames(), single_message_mode_)
 {
 }
 
@@ -67,7 +67,9 @@ void ProtobufRowInputFormat::syncAfterError()
 
 void registerInputFormatProcessorProtobuf(FormatFactory & factory)
 {
-    factory.registerInputFormatProcessor("Protobuf", [](
+    for (bool single_message_mode : {false, true})
+    {
+        factory.registerInputFormatProcessor(single_message_mode ? "ProtobufSingle" : "Protobuf", [single_message_mode](
             ReadBuffer & buf,
             const Block & sample,
             IRowInputFormat::Params params,
@@ -75,9 +77,11 @@ void registerInputFormatProcessorProtobuf(FormatFactory & factory)
         {
             return std::make_shared<ProtobufRowInputFormat>(buf, sample, std::move(params),
                 FormatSchemaInfo(settings.schema.format_schema, "Protobuf", true,
-                             settings.schema.is_server, settings.schema.format_schema_path));
+                                settings.schema.is_server, settings.schema.format_schema_path),
+                single_message_mode);
         });
     }
+}
 
 }
 

src/Processors/Formats/Impl/ProtobufRowInputFormat.h

@@ -16,17 +16,20 @@ class FormatSchemaInfo;
 
 
 /** Stream designed to deserialize data from the google protobuf format.
-  * Each row is read as a separated message.
-  * These messages are delimited according to documentation
+  * One Protobuf message is parsed as one row of data.
+  *
+  * Input buffer may contain single protobuf message (single_message_mode_ = true),
+  * or any number of messages (single_message_mode_ = false). In the second case
+  * parser assumes messages are length-delimited according to documentation
   * https://github.com/protocolbuffers/protobuf/blob/master/src/google/protobuf/util/delimited_message_util.h
-  * Serializing in the protobuf format requires the 'format_schema' setting to be set, e.g.
+  * Parsing of the protobuf format requires the 'format_schema' setting to be set, e.g.
   * INSERT INTO table FORMAT Protobuf SETTINGS format_schema = 'schema:Message'
   * where schema is the name of "schema.proto" file specifying protobuf schema.
   */
 class ProtobufRowInputFormat : public IRowInputFormat
 {
 public:
-    ProtobufRowInputFormat(ReadBuffer & in_, const Block & header_, Params params_, const FormatSchemaInfo & info_);
+    ProtobufRowInputFormat(ReadBuffer & in_, const Block & header_, Params params_, const FormatSchemaInfo & info_, const bool single_message_mode_);
     ~ProtobufRowInputFormat() override;
 
     String getName() const override { return "ProtobufRowInputFormat"; }

tests/integration/test_storage_kafka/test.py

@@ -103,6 +103,17 @@ def kafka_produce_protobuf_messages(topic, start_index, num_messages):
     producer.flush()
     print(("Produced {} messages for topic {}".format(num_messages, topic)))
 
+def kafka_produce_protobuf_messages_no_delimeters(topic, start_index, num_messages):
+    data = ''
+    producer = KafkaProducer(bootstrap_servers="localhost:9092")
+    for i in range(start_index, start_index + num_messages):
+        msg = kafka_pb2.KeyValuePair()
+        msg.key = i
+        msg.value = str(i)
+        serialized_msg = msg.SerializeToString()
+        producer.send(topic=topic, value=serialized_msg)
+    producer.flush()
+    print("Produced {} messages for topic {}".format(num_messages, topic))
 
 def avro_confluent_message(schema_registry_client, value):
     # type: (CachedSchemaRegistryClient, dict) -> str
@@ -971,6 +982,30 @@ def test_kafka_protobuf(kafka_cluster):
     kafka_check_result(result, True)
 
 
+@pytest.mark.timeout(180)
+def test_kafka_protobuf_no_delimiter(kafka_cluster):
+    instance.query('''
+        CREATE TABLE test.kafka (key UInt64, value String)
+            ENGINE = Kafka
+            SETTINGS kafka_broker_list = 'kafka1:19092',
+                     kafka_topic_list = 'pb_no_delimiter',
+                     kafka_group_name = 'pb_no_delimiter',
+                     kafka_format = 'ProtobufSingle',
+                     kafka_schema = 'kafka.proto:KeyValuePair';
+        ''')
+
+    kafka_produce_protobuf_messages_no_delimeters('pb_no_delimiter', 0, 20)
+    kafka_produce_protobuf_messages_no_delimeters('pb_no_delimiter', 20, 1)
+    kafka_produce_protobuf_messages_no_delimeters('pb_no_delimiter', 21, 29)
+
+    result = ''
+    while True:
+        result += instance.query('SELECT * FROM test.kafka', ignore_error=True)
+        if kafka_check_result(result):
+            break
+
+    kafka_check_result(result, True)
+
 @pytest.mark.timeout(180)
 def test_kafka_materialized_view(kafka_cluster):
     instance.query('''