No public description

PiperOrigin-RevId: 715812031

Author: ArrayRecord Team

cpp/BUILD

@@ -115,6 +115,7 @@ cc_library(
         "@abseil-cpp//absl/status",
         "@abseil-cpp//absl/status:statusor",
         "@abseil-cpp//absl/strings",
+        "@abseil-cpp//absl/strings:cord",
         "@abseil-cpp//absl/synchronization",
         "@abseil-cpp//absl/types:span",
         "@protobuf//:protobuf_lite",
@@ -235,6 +236,8 @@ cc_test(
         ":test_utils",
         ":thread_pool",
         "@abseil-cpp//absl/strings",
+        "@abseil-cpp//absl/strings:cord",
+        "@abseil-cpp//absl/strings:cord_test_helpers",
         "@googletest//:gtest_main",
         "@riegeli//riegeli/base:initializer",
         "@riegeli//riegeli/bytes:string_reader",

cpp/array_record_writer.cc

@@ -31,6 +31,7 @@ limitations under the License.
 #include "absl/log/check.h"
 #include "absl/status/status.h"
 #include "absl/status/statusor.h"
+#include "absl/strings/cord.h"
 #include "absl/strings/match.h"
 #include "absl/strings/str_cat.h"
 #include "absl/strings/string_view.h"
@@ -39,8 +40,8 @@ limitations under the License.
 #include "cpp/common.h"
 #include "cpp/layout.pb.h"
 #include "cpp/sequenced_chunk_writer.h"
-#include "cpp/tri_state_ptr.h"
 #include "cpp/thread_pool.h"
+#include "cpp/tri_state_ptr.h"
 #include "google/protobuf/message_lite.h"
 #include "riegeli/base/object.h"
 #include "riegeli/base/options_parser.h"
@@ -418,6 +419,16 @@ bool ArrayRecordWriterBase::WriteRecord(absl::string_view record) {
   return WriteRecordImpl(std::move(record));
 }
 
+bool ArrayRecordWriterBase::WriteRecord(const absl::Cord& record) {
+  if (auto flat = record.TryFlat(); flat.has_value()) {
+    return WriteRecord(*flat);
+  }
+
+  std::string cord_string;
+  absl::AppendCordToString(record, &cord_string);
+  return WriteRecord(cord_string);
+}
+
 bool ArrayRecordWriterBase::WriteRecord(const void* data, size_t num_bytes) {
   auto view = absl::string_view(reinterpret_cast<const char*>(data), num_bytes);
   return WriteRecordImpl(std::move(view));

cpp/array_record_writer.h

@@ -66,12 +66,13 @@ limitations under the License.
 #include <utility>
 
 #include "absl/status/statusor.h"
+#include "absl/strings/cord.h"
 #include "absl/strings/string_view.h"
 #include "absl/types/span.h"
 #include "cpp/common.h"
 #include "cpp/sequenced_chunk_writer.h"
-#include "cpp/tri_state_ptr.h"
 #include "cpp/thread_pool.h"
+#include "cpp/tri_state_ptr.h"
 #include "riegeli/base/initializer.h"
 #include "riegeli/base/object.h"
 #include "riegeli/bytes/writer.h"
@@ -304,6 +305,7 @@ class ArrayRecordWriterBase : public riegeli::Object {
   // Write records of various types.
   bool WriteRecord(const google::protobuf::MessageLite& record);
   bool WriteRecord(absl::string_view record);
+  bool WriteRecord(const absl::Cord& record);
   bool WriteRecord(const void* data, size_t num_bytes);
   template <typename T>
   bool WriteRecord(absl::Span<const T> record) {

cpp/array_record_writer_test.cc

@@ -25,6 +25,8 @@ limitations under the License.
 #include <vector>
 
 #include "gtest/gtest.h"
+#include "absl/strings/cord.h"
+#include "absl/strings/cord_test_helpers.h"
 #include "absl/strings/string_view.h"
 #include "cpp/common.h"
 #include "cpp/layout.pb.h"
@@ -116,6 +118,39 @@ TEST_P(ArrayRecordWriterTest, MoveTest) {
   }
 }
 
+TEST_P(ArrayRecordWriterTest, CordTest) {
+  std::string encoded;
+  ARThreadPool* pool = nullptr;
+  if (std::get<3>(GetParam())) {
+    pool = ArrayRecordGlobalPool();
+  }
+  auto options = GetOptions();
+  options.set_group_size(2);
+  auto writer = ArrayRecordWriter(
+      riegeli::Maker<riegeli::StringWriter>(&encoded), options, pool);
+
+  absl::Cord flat_cord("test");
+  // Empty string should not crash the writer.
+  absl::Cord empty_cord("");
+  absl::Cord fragmented_cord = absl::MakeFragmentedCord({"aaa ", "", "c"});
+
+  EXPECT_TRUE(writer.WriteRecord(flat_cord));
+  EXPECT_TRUE(writer.WriteRecord(empty_cord));
+  EXPECT_TRUE(writer.WriteRecord(fragmented_cord));
+  ASSERT_TRUE(writer.Close());
+
+  // Empty string should not crash the reader.
+  std::vector<std::string> expected_strings{"test", "", "aaa c"};
+
+  auto reader =
+      riegeli::RecordReader(riegeli::Maker<riegeli::StringReader>(encoded));
+  for (const auto& expected : expected_strings) {
+    std::string result;
+    reader.ReadRecord(result);
+    EXPECT_EQ(result, expected);
+  }
+}
+
 TEST_P(ArrayRecordWriterTest, RandomDatasetTest) {
   std::mt19937 bitgen;
   constexpr uint32_t kGroupSize = 100;