Add brotli content-encoding filter.

net/filter/brotli_filter_unittest.cc

+// Copyright 2015 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 "base/files/file_util.h"
+#include "base/memory/scoped_ptr.h"
+#include "base/path_service.h"
+#include "net/base/io_buffer.h"
+#include "net/filter/brotli_filter.h"
+#include "net/filter/mock_filter_context.h"
+#include "testing/gtest/include/gtest/gtest.h"
+#include "testing/platform_test.h"
+
+namespace {
+const int kDefaultBufferSize = 4096;
+const int kSmallBufferSize = 128;
+}  // namespace
+
+namespace net {
+
+// These tests use the path service, which uses autoreleased objects on the
+// Mac, so this needs to be a PlatformTest.
+class BrotliUnitTest : public PlatformTest {
+ protected:
+  void SetUp() override {
+    PlatformTest::SetUp();
+
+    // Get the path of data directory.
+    base::FilePath data_dir;
+    PathService::Get(base::DIR_SOURCE_ROOT, &data_dir);
+    data_dir = data_dir.AppendASCII("net");
+    data_dir = data_dir.AppendASCII("data");
+    data_dir = data_dir.AppendASCII("filter_unittests");
+
+    // Read data from the original file into buffer.
+    base::FilePath file_path;
+    file_path = data_dir.AppendASCII("google.txt");
+    ASSERT_TRUE(base::ReadFileToString(file_path, &source_buffer_));
+
+    // Read data from the encoded file into buffer.
+    base::FilePath encoded_file_path;
+    encoded_file_path = data_dir.AppendASCII("google.br");
+    ASSERT_TRUE(base::ReadFileToString(encoded_file_path, &encoded_buffer_));
+    ASSERT_GE(kDefaultBufferSize, static_cast<int>(encoded_buffer_.size()));
+  }
+
+  // Use filter to decode compressed data, and compare the decoded result with
+  // the orginal data.
+  // Parameters: |source| and |source_len| are original data and its size.
+  // |encoded_source| and |encoded_source_len| are compressed data and its size.
+  // |output_buffer_size| specifies the size of buffer to read out data from
+  // filter.
+  void DecodeAndCompareWithFilter(Filter* filter,
+                                  const char* source,
+                                  int source_len,

[Ryan Sleevi, 2015/12/17 01:04:40]

https://www.chromium.org/developers/coding-style#TOC-Types

"use size_t for object and allocation sizes, boject counts, array and pointer
offsets, vector indices, and so on"

[eustas, 2015/12/17 16:44:54]

I also prefer to use size_t where appropriate, but Filter class has "int" type
everywhere. If I turn source_len to size_t it would require adding a bunch of
static_casts

+                                  const char* encoded_source,
+                                  int encoded_source_len,
+                                  int output_buffer_size) {
+    // Make sure we have enough space to hold the decoding output.
+    ASSERT_GE(kDefaultBufferSize, source_len);
+    ASSERT_GE(kDefaultBufferSize, output_buffer_size);
+
+    char decode_buffer[kDefaultBufferSize];
+    char* decode_next = decode_buffer;
+    int decode_avail_size = kDefaultBufferSize;
+
+    const char* encode_next = encoded_source;
+    int encode_avail_size = encoded_source_len;
+
+    int code = Filter::FILTER_OK;
+    while (code != Filter::FILTER_DONE) {
+      int encode_data_len =
+          std::min(encode_avail_size, filter->stream_buffer_size());
+      memcpy(filter->stream_buffer()->data(), encode_next, encode_data_len);
+      filter->FlushStreamBuffer(encode_data_len);
+      encode_next += encode_data_len;
+      encode_avail_size -= encode_data_len;
+
+      while (true) {
+        int decode_data_len = std::min(decode_avail_size, output_buffer_size);
+
+        code = filter->ReadData(decode_next, &decode_data_len);
+        decode_next += decode_data_len;
+        decode_avail_size -= decode_data_len;
+
+        ASSERT_NE(Filter::FILTER_ERROR, code);
+
+        if (code == Filter::FILTER_NEED_MORE_DATA ||
+            code == Filter::FILTER_DONE) {
+          break;
+        }
+      }
+    }
+
+    // Compare the decoding result with source data
+    int decode_total_data_len = kDefaultBufferSize - decode_avail_size;
+    EXPECT_EQ(source_len, decode_total_data_len);
+    EXPECT_EQ(memcmp(source, decode_buffer, source_len), 0);
+  }
+
+  // Unsafe function to use filter to decode compressed data.
+  // Parameters: |source| and |source_len| are compressed data and its size.
+  // |dest| is the buffer for decoding results. Upon entry, |*dest_len| is the
+  // size of the output buffer. Upon exit, |*dest_len| is the number of chars
+  // written into the buffer.
+  int DecodeAllWithFilter(Filter* filter,
+                          const char* source,
+                          int source_len,
+                          char* dest,
+                          int* dest_len) {
+    memcpy(filter->stream_buffer()->data(), source, source_len);
+    filter->FlushStreamBuffer(source_len);
+    return filter->ReadData(dest, dest_len);
+  }
+
+  void InitFilter() {
+    std::vector<Filter::FilterType> filter_types;
+    filter_types.push_back(Filter::FILTER_TYPE_BROTLI);
+    filter_.reset(Filter::Factory(filter_types, filter_context_));
+    ASSERT_TRUE(filter_.get());
+    ASSERT_LE(kDefaultBufferSize, filter_->stream_buffer_size());
+  }
+
+  void InitFilterWithBufferSize(int buffer_size) {
+    std::vector<Filter::FilterType> filter_types;
+    filter_types.push_back(Filter::FILTER_TYPE_BROTLI);
+    filter_.reset(
+        Filter::FactoryForTests(filter_types, filter_context_, buffer_size));
+    ASSERT_TRUE(filter_.get());
+  }
+
+  const char* source_buffer() const { return source_buffer_.data(); }
+  int source_len() const { return static_cast<int>(source_buffer_.size()); }
+
+  const char* encoded_buffer() const { return encoded_buffer_.data(); }
+  int encoded_len() const { return static_cast<int>(encoded_buffer_.size()); }
+
+  scoped_ptr<Filter> filter_;
+
+ private:
+  MockFilterContext filter_context_;
+  std::string source_buffer_;
+  std::string encoded_buffer_;
+};
+
+// Basic scenario: decoding brotli data with big enough buffer.
+TEST_F(BrotliUnitTest, DecodeBrotli) {
+  InitFilter();
+  memcpy(filter_->stream_buffer()->data(), encoded_buffer(), encoded_len());
+  filter_->FlushStreamBuffer(encoded_len());
+
+  ASSERT_GE(kDefaultBufferSize, source_len());
+  char decode_buffer[kDefaultBufferSize];
+  int decode_size = kDefaultBufferSize;
+  filter_->ReadData(decode_buffer, &decode_size);
+
+  // Compare the decoding result with source data
+  EXPECT_EQ(source_len(), decode_size);
+  EXPECT_EQ(memcmp(source_buffer(), decode_buffer, source_len()), 0);
+}
+
+// Tests we can call filter repeatedly to get all the data decoded.
+// To do that, we create a filter with a small buffer that can not hold all
+// the input data.
+TEST_F(BrotliUnitTest, DecodeWithSmallBuffer) {
+  InitFilterWithBufferSize(kSmallBufferSize);
+  EXPECT_EQ(kSmallBufferSize, filter_->stream_buffer_size());
+  DecodeAndCompareWithFilter(filter_.get(), source_buffer(), source_len(),
+                             encoded_buffer(), encoded_len(),
+                             kDefaultBufferSize);
+}
+
+// Tests we can still decode with just 1 byte buffer in the filter.
+// The purpose of this test: sometimes the filter will consume input without
+// generating output. Verify filter can handle it correctly.
+TEST_F(BrotliUnitTest, DecodeWithOneByteBuffer) {
+  InitFilterWithBufferSize(1);
+  EXPECT_EQ(1, filter_->stream_buffer_size());
+  DecodeAndCompareWithFilter(filter_.get(), source_buffer(), source_len(),
+                             encoded_buffer(), encoded_len(),
+                             kDefaultBufferSize);
+}
+
+// Tests we can decode when caller has small buffer to read out from filter.
+TEST_F(BrotliUnitTest, DecodeWithSmallOutputBuffer) {
+  InitFilter();
+  DecodeAndCompareWithFilter(filter_.get(), source_buffer(), source_len(),
+                             encoded_buffer(), encoded_len(), kSmallBufferSize);
+}
+
+// Tests we can decode when caller has small buffer and input is also broken
+// into small parts. This may uncover some corner cases that doesn't happen with
+// one-byte buffers.
+TEST_F(BrotliUnitTest, DecodeWithSmallInputAndOutputBuffer) {
+  InitFilterWithBufferSize(kSmallBufferSize);
+  DecodeAndCompareWithFilter(filter_.get(), source_buffer(), source_len(),
+                             encoded_buffer(), encoded_len(), kSmallBufferSize);
+}
+
+// Tests we can still decode with just 1 byte buffer in the filter and just 1
+// byte buffer in the caller.
+TEST_F(BrotliUnitTest, DecodeWithOneByteInputAndOutputBuffer) {
+  InitFilterWithBufferSize(1);
+  EXPECT_EQ(1, filter_->stream_buffer_size());
+  DecodeAndCompareWithFilter(filter_.get(), source_buffer(), source_len(),
+                             encoded_buffer(), encoded_len(), 1);
+}
+
+// Decoding deflate stream with corrupted data.
+TEST_F(BrotliUnitTest, DecodeCorruptedData) {
+  char corrupt_data[kDefaultBufferSize];
+  int corrupt_data_len = encoded_len();
+  memcpy(corrupt_data, encoded_buffer(), encoded_len());
+
+  int pos = corrupt_data_len / 2;
+  corrupt_data[pos] = !corrupt_data[pos];
+
+  // Decode the corrupted data with filter
+  InitFilter();
+  char corrupt_decode_buffer[kDefaultBufferSize];
+  int corrupt_decode_size = kDefaultBufferSize;
+
+  int code = DecodeAllWithFilter(filter_.get(), corrupt_data, corrupt_data_len,
+                                 corrupt_decode_buffer, &corrupt_decode_size);
+
+  // Expect failures
+  EXPECT_EQ(Filter::FILTER_ERROR, code);
+}
+
+// Decoding deflate stream with missing data.
+TEST_F(BrotliUnitTest, DecodeMissingData) {

[Ryan Sleevi, 2015/12/17 01:04:40]

Why don't any of these tests need to be conditionally enabled/disabled?

[eustas, 2015/12/17 16:44:54]

The whole BrotliUnitTest is excluded when brotli support is disabled.
FilterFactory returns nullptr in this case -> nothing to test.

+  char corrupt_data[kDefaultBufferSize];
+  int corrupt_data_len = encoded_len();
+  memcpy(corrupt_data, encoded_buffer(), encoded_len());
+
+  int pos = corrupt_data_len / 2;
+  int len = corrupt_data_len - pos - 1;
+  memmove(&corrupt_data[pos], &corrupt_data[pos + 1], len);
+  --corrupt_data_len;
+
+  // Decode the corrupted data with filter
+  InitFilter();
+  char corrupt_decode_buffer[kDefaultBufferSize];
+  int corrupt_decode_size = kDefaultBufferSize;
+
+  int code = DecodeAllWithFilter(filter_.get(), corrupt_data, corrupt_data_len,
+                                 corrupt_decode_buffer, &corrupt_decode_size);
+
+  // Expect failures
+  EXPECT_EQ(Filter::FILTER_ERROR, code);
+}
+
+}  // namespace net