Introduced ChunkedByteBuffer class which will be required by next speech recognition CLs (2.x).

BUG=116954
TEST=content_unittest (ChunkedByteBufferTest.BasicTest)

Review URL: http://codereview.chromium.org/10123008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@133885 0039d316-1c4b-4281-b951-d872f2087c98

3 files changed, 283 insertions, 0 deletions

diff --git a/content/browser/speech/chunked_byte_buffer.cc b/content/browser/speech/chunked_byte_buffer.cc
new file mode 100644
index 0000000..123755a
--- /dev/null
+++ b/content/browser/speech/chunked_byte_buffer.cc
@@ -0,0 +1,131 @@
+// Copyright (c) 2012 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include "content/browser/speech/chunked_byte_buffer.h"
+
+#include <algorithm>
+
+#include "base/basictypes.h"
+#include "base/lazy_instance.h"
+#include "base/logging.h"
+
+namespace {
+
+static const size_t kHeaderLength = sizeof(uint32);
+
+COMPILE_ASSERT(sizeof(size_t) >= kHeaderLength,
+               ChunkedByteBufferNotSupportedOnThisArchitecture);
+
+uint32 ReadBigEndian32(const uint8* buffer) {
+  return (static_cast<uint32>(buffer[3])) |
+         (static_cast<uint32>(buffer[2]) <<  8) |
+         (static_cast<uint32>(buffer[1]) << 16) |
+         (static_cast<uint32>(buffer[0]) << 24);
+}
+
+}  // namespace
+
+namespace speech {
+
+ChunkedByteBuffer::ChunkedByteBuffer()
+    : partial_chunk_(new Chunk()),
+      total_bytes_stored_(0) {
+}
+
+ChunkedByteBuffer::~ChunkedByteBuffer() {
+  Clear();
+}
+
+void ChunkedByteBuffer::Append(const uint8* start, size_t length) {
+  DCHECK(length > 0);
+  size_t remaining_bytes = length;
+  const uint8* next_data = start;
+
+  while (remaining_bytes > 0) {
+    DCHECK(partial_chunk_ != NULL);
+    size_t insert_length = 0;
+    bool header_completed = false;
+    bool content_completed = false;
+    std::vector<uint8>* insert_target;
+
+    if (partial_chunk_->header.size() < kHeaderLength) {
+      const size_t bytes_to_complete_header =
+          kHeaderLength - partial_chunk_->header.size();
+      insert_length = std::min(bytes_to_complete_header, remaining_bytes);
+      insert_target = &partial_chunk_->header;
+      header_completed = (remaining_bytes >= bytes_to_complete_header);
+    } else {
+      DCHECK_LT(partial_chunk_->content->size(),
+                partial_chunk_->ExpectedContentLength());
+      const size_t bytes_to_complete_chunk =
+          partial_chunk_->ExpectedContentLength() -
+          partial_chunk_->content->size();
+      insert_length = std::min(bytes_to_complete_chunk, remaining_bytes);
+      insert_target = partial_chunk_->content.get();
+      content_completed = (remaining_bytes >= bytes_to_complete_chunk);
+    }
+
+    DCHECK_GT(insert_length, 0U);
+    DCHECK_LE(insert_length, remaining_bytes);
+    DCHECK_LE(next_data + insert_length, start + length);
+    insert_target->insert(insert_target->end(),
+                          next_data,
+                          next_data + insert_length);
+    next_data += insert_length;
+    remaining_bytes -= insert_length;
+
+    if (header_completed) {
+      DCHECK_EQ(partial_chunk_->header.size(), kHeaderLength);
+      DCHECK_NE(partial_chunk_->ExpectedContentLength(), 0U);
+      partial_chunk_->content->reserve(partial_chunk_->ExpectedContentLength());
+    } else if (content_completed) {
+      DCHECK_EQ(partial_chunk_->content->size(),
+                partial_chunk_->ExpectedContentLength());
+      chunks_.push_back(partial_chunk_.release());
+      partial_chunk_.reset(new Chunk());
+    }
+  }
+  DCHECK_EQ(next_data, start + length);
+  total_bytes_stored_ += length;
+}
+
+void ChunkedByteBuffer::Append(const std::string& string) {
+  Append(reinterpret_cast<const uint8*>(string.data()), string.size());
+}
+
+bool ChunkedByteBuffer::HasChunks() const {
+  return !chunks_.empty();
+}
+
+scoped_ptr< std::vector<uint8> > ChunkedByteBuffer::PopChunk() {
+  if (chunks_.empty())
+    return scoped_ptr< std::vector<uint8> >();
+  scoped_ptr<Chunk> chunk(*chunks_.begin());
+  chunks_.weak_erase(chunks_.begin());
+  DCHECK_EQ(chunk->header.size(), kHeaderLength);
+  DCHECK_EQ(chunk->content->size(), chunk->ExpectedContentLength());
+  total_bytes_stored_ -= chunk->content->size();
+  total_bytes_stored_ -= kHeaderLength;
+  return chunk->content.Pass();
+}
+
+void ChunkedByteBuffer::Clear() {
+  chunks_.reset();
+  partial_chunk_.reset(new Chunk());
+  total_bytes_stored_ = 0;
+}
+
+ChunkedByteBuffer::Chunk::Chunk()
+    : content(new std::vector<uint8>()) {
+}
+
+ChunkedByteBuffer::Chunk::~Chunk() {
+}
+
+size_t ChunkedByteBuffer::Chunk::ExpectedContentLength() const {
+  DCHECK_EQ(header.size(), kHeaderLength);
+  return static_cast<size_t>(ReadBigEndian32(&header[0]));
+}
+
+}  // namespace speech
diff --git a/content/browser/speech/chunked_byte_buffer.h b/content/browser/speech/chunked_byte_buffer.h
new file mode 100644
index 0000000..71ccc06
--- /dev/null
+++ b/content/browser/speech/chunked_byte_buffer.h
@@ -0,0 +1,76 @@
+// Copyright (c) 2012 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef CONTENT_BROWSER_SPEECH_CHUNKED_BYTE_BUFFER_H_
+#define CONTENT_BROWSER_SPEECH_CHUNKED_BYTE_BUFFER_H_
+#pragma once
+
+#include <string>
+#include <vector>
+
+#include "base/basictypes.h"
+#include "base/memory/scoped_ptr.h"
+#include "base/memory/scoped_vector.h"
+#include "content/common/content_export.h"
+
+namespace speech {
+
+// Models a chunk-oriented byte buffer. The term chunk is herein defined as an
+// arbitrary sequence of bytes that is preceeded by N header bytes, indicating
+// its size. Data may be appended to the buffer with no particular respect of
+// chunks boundaries. However, chunks can be extracted (FIFO) only when their
+// content (according to their header) is fully available in the buffer.
+// The current implementation support only 4 byte Big Endian headers.
+// Empty chunks (i.e. the sequence 00 00 00 00) are NOT allowed.
+//
+// E.g. 00 00 00 04 xx xx xx xx 00 00 00 02 yy yy 00 00 00 04 zz zz zz zz
+//      [----- CHUNK 1 -------] [--- CHUNK 2 ---] [------ CHUNK 3 ------]
+class CONTENT_EXPORT ChunkedByteBuffer {
+ public:
+  ChunkedByteBuffer();
+  ~ChunkedByteBuffer();
+
+  // Appends |length| bytes starting from |start| to the buffer.
+  void Append(const uint8* start, size_t length);
+
+  // Appends bytes contained in the |string| to the buffer.
+  void Append(const std::string& string);
+
+  // Checks whether one or more complete chunks are available in the buffer.
+  bool HasChunks() const;
+
+  // If enough data is available, reads and removes the first complete chunk
+  // from the buffer. Returns a NULL pointer if no complete chunk is available.
+  scoped_ptr< std::vector<uint8> > PopChunk();
+
+  // Clears all the content of the buffer.
+  void Clear();
+
+  // Returns the number of raw bytes (including headers) present.
+  size_t GetTotalLength() const { return total_bytes_stored_; }
+
+ private:
+  struct Chunk {
+    Chunk();
+    ~Chunk();
+
+    std::vector<uint8> header;
+    scoped_ptr< std::vector<uint8> > content;
+    size_t ExpectedContentLength() const;
+
+   private:
+    DISALLOW_COPY_AND_ASSIGN(Chunk);
+  };
+
+  ScopedVector<Chunk> chunks_;
+  scoped_ptr<Chunk> partial_chunk_;
+  size_t total_bytes_stored_;
+
+  DISALLOW_COPY_AND_ASSIGN(ChunkedByteBuffer);
+};
+
+
+}  // namespace speech
+
+#endif  // CONTENT_BROWSER_SPEECH_CHUNKED_BYTE_BUFFER_H_
diff --git a/content/browser/speech/chunked_byte_buffer_unittest.cc b/content/browser/speech/chunked_byte_buffer_unittest.cc
new file mode 100644
index 0000000..1633101
--- /dev/null
+++ b/content/browser/speech/chunked_byte_buffer_unittest.cc
@@ -0,0 +1,76 @@
+// Copyright (c) 2012 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include <string>
+#include <vector>
+
+#include "content/browser/speech/chunked_byte_buffer.h"
+#include "testing/gtest/include/gtest/gtest.h"
+
+namespace speech {
+
+typedef std::vector<uint8> ByteVector;
+
+TEST(ChunkedByteBufferTest, BasicTest) {
+  ChunkedByteBuffer buffer;
+
+  const uint8 kChunks[] = {
+      0x00, 0x00, 0x00, 0x04, 0x01, 0x02, 0x03, 0x04,  // Chunk 1: 4 bytes
+      0x00, 0x00, 0x00, 0x02, 0x05, 0x06,  // Chunk 2: 2 bytes
+      0x00, 0x00, 0x00, 0x01, 0x07  // Chunk 3: 1 bytes
+  };
+
+  EXPECT_EQ(0U, buffer.GetTotalLength());
+  EXPECT_FALSE(buffer.HasChunks());
+
+  // Append partially chunk 1.
+  buffer.Append(kChunks, 2);
+  EXPECT_EQ(2U, buffer.GetTotalLength());
+  EXPECT_FALSE(buffer.HasChunks());
+
+  // Complete chunk 1.
+  buffer.Append(kChunks + 2, 6);
+  EXPECT_EQ(8U, buffer.GetTotalLength());
+  EXPECT_TRUE(buffer.HasChunks());
+
+  // Append fully chunk 2.
+  buffer.Append(kChunks + 8, 6);
+  EXPECT_EQ(14U, buffer.GetTotalLength());
+  EXPECT_TRUE(buffer.HasChunks());
+
+  // Remove and check chunk 1.
+  scoped_ptr<ByteVector> chunk;
+  chunk = buffer.PopChunk();
+  EXPECT_TRUE(chunk != NULL);
+  EXPECT_EQ(4U, chunk->size());
+  EXPECT_EQ(0, std::char_traits<uint8>::compare(kChunks + 4,
+                                                &(*chunk)[0],
+                                                chunk->size()));
+  EXPECT_EQ(6U, buffer.GetTotalLength());
+  EXPECT_TRUE(buffer.HasChunks());
+
+  // Read and check chunk 2.
+  chunk = buffer.PopChunk();
+  EXPECT_TRUE(chunk != NULL);
+  EXPECT_EQ(2U, chunk->size());
+  EXPECT_EQ(0, std::char_traits<uint8>::compare(kChunks + 12,
+                                                &(*chunk)[0],
+                                                chunk->size()));
+  EXPECT_EQ(0U, buffer.GetTotalLength());
+  EXPECT_FALSE(buffer.HasChunks());
+
+  // Append fully chunk 3.
+  buffer.Append(kChunks + 14, 5);
+  EXPECT_EQ(5U, buffer.GetTotalLength());
+
+  // Remove and check chunk 3.
+  chunk = buffer.PopChunk();
+  EXPECT_TRUE(chunk != NULL);
+  EXPECT_EQ(1U, chunk->size());
+  EXPECT_EQ((*chunk)[0], kChunks[18]);
+  EXPECT_EQ(0U, buffer.GetTotalLength());
+  EXPECT_FALSE(buffer.HasChunks());
+}
+
+}  // namespace speech