filiptepper-leveldb-ruby 0.14

data/leveldb/table/filter_block.cc

@@ -0,0 +1,111 @@
+// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "table/filter_block.h"
+
+#include "leveldb/filter_policy.h"
+#include "util/coding.h"
+
+namespace leveldb {
+
+// See doc/table_format.txt for an explanation of the filter block format.
+
+// Generate new filter every 2KB of data
+static const size_t kFilterBaseLg = 11;
+static const size_t kFilterBase = 1 << kFilterBaseLg;
+
+FilterBlockBuilder::FilterBlockBuilder(const FilterPolicy* policy)
+    : policy_(policy) {
+}
+
+void FilterBlockBuilder::StartBlock(uint64_t block_offset) {
+  uint64_t filter_index = (block_offset / kFilterBase);
+  assert(filter_index >= filter_offsets_.size());
+  while (filter_index > filter_offsets_.size()) {
+    GenerateFilter();
+  }
+}
+
+void FilterBlockBuilder::AddKey(const Slice& key) {
+  Slice k = key;
+  start_.push_back(keys_.size());
+  keys_.append(k.data(), k.size());
+}
+
+Slice FilterBlockBuilder::Finish() {
+  if (!start_.empty()) {
+    GenerateFilter();
+  }
+
+  // Append array of per-filter offsets
+  const uint32_t array_offset = result_.size();
+  for (size_t i = 0; i < filter_offsets_.size(); i++) {
+    PutFixed32(&result_, filter_offsets_[i]);
+  }
+
+  PutFixed32(&result_, array_offset);
+  result_.push_back(kFilterBaseLg);  // Save encoding parameter in result
+  return Slice(result_);
+}
+
+void FilterBlockBuilder::GenerateFilter() {
+  const size_t num_keys = start_.size();
+  if (num_keys == 0) {
+    // Fast path if there are no keys for this filter
+    filter_offsets_.push_back(result_.size());
+    return;
+  }
+
+  // Make list of keys from flattened key structure
+  start_.push_back(keys_.size());  // Simplify length computation
+  tmp_keys_.resize(num_keys);
+  for (size_t i = 0; i < num_keys; i++) {
+    const char* base = keys_.data() + start_[i];
+    size_t length = start_[i+1] - start_[i];
+    tmp_keys_[i] = Slice(base, length);
+  }
+
+  // Generate filter for current set of keys and append to result_.
+  filter_offsets_.push_back(result_.size());
+  policy_->CreateFilter(&tmp_keys_[0], num_keys, &result_);
+
+  tmp_keys_.clear();
+  keys_.clear();
+  start_.clear();
+}
+
+FilterBlockReader::FilterBlockReader(const FilterPolicy* policy,
+                                     const Slice& contents)
+    : policy_(policy),
+      data_(NULL),
+      offset_(NULL),
+      num_(0),
+      base_lg_(0) {
+  size_t n = contents.size();
+  if (n < 5) return;  // 1 byte for base_lg_ and 4 for start of offset array
+  base_lg_ = contents[n-1];
+  uint32_t last_word = DecodeFixed32(contents.data() + n - 5);
+  if (last_word > n - 5) return;
+  data_ = contents.data();
+  offset_ = data_ + last_word;
+  num_ = (n - 5 - last_word) / 4;
+}
+
+bool FilterBlockReader::KeyMayMatch(uint64_t block_offset, const Slice& key) {
+  uint64_t index = block_offset >> base_lg_;
+  if (index < num_) {
+    uint32_t start = DecodeFixed32(offset_ + index*4);
+    uint32_t limit = DecodeFixed32(offset_ + index*4 + 4);
+    if (start <= limit && limit <= (offset_ - data_)) {
+      Slice filter = Slice(data_ + start, limit - start);
+      return policy_->KeyMayMatch(key, filter);
+    } else if (start == limit) {
+      // Empty filters do not match any keys
+      return false;
+    }
+  }
+  return true;  // Errors are treated as potential matches
+}
+
+}

data/leveldb/table/filter_block.h

@@ -0,0 +1,68 @@
+// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+//
+// A filter block is stored near the end of a Table file.  It contains
+// filters (e.g., bloom filters) for all data blocks in the table combined
+// into a single filter block.
+
+#ifndef STORAGE_LEVELDB_TABLE_FILTER_BLOCK_H_
+#define STORAGE_LEVELDB_TABLE_FILTER_BLOCK_H_
+
+#include <stddef.h>
+#include <stdint.h>
+#include <string>
+#include <vector>
+#include "leveldb/slice.h"
+#include "util/hash.h"
+
+namespace leveldb {
+
+class FilterPolicy;
+
+// A FilterBlockBuilder is used to construct all of the filters for a
+// particular Table.  It generates a single string which is stored as
+// a special block in the Table.
+//
+// The sequence of calls to FilterBlockBuilder must match the regexp:
+//      (StartBlock AddKey*)* Finish
+class FilterBlockBuilder {
+ public:
+  explicit FilterBlockBuilder(const FilterPolicy*);
+
+  void StartBlock(uint64_t block_offset);
+  void AddKey(const Slice& key);
+  Slice Finish();
+
+ private:
+  void GenerateFilter();
+
+  const FilterPolicy* policy_;
+  std::string keys_;              // Flattened key contents
+  std::vector<size_t> start_;     // Starting index in keys_ of each key
+  std::string result_;            // Filter data computed so far
+  std::vector<Slice> tmp_keys_;   // policy_->CreateFilter() argument
+  std::vector<uint32_t> filter_offsets_;
+
+  // No copying allowed
+  FilterBlockBuilder(const FilterBlockBuilder&);
+  void operator=(const FilterBlockBuilder&);
+};
+
+class FilterBlockReader {
+ public:
+ // REQUIRES: "contents" and *policy must stay live while *this is live.
+  FilterBlockReader(const FilterPolicy* policy, const Slice& contents);
+  bool KeyMayMatch(uint64_t block_offset, const Slice& key);
+
+ private:
+  const FilterPolicy* policy_;
+  const char* data_;    // Pointer to filter data (at block-start)
+  const char* offset_;  // Pointer to beginning of offset array (at block-end)
+  size_t num_;          // Number of entries in offset array
+  size_t base_lg_;      // Encoding parameter (see kFilterBaseLg in .cc file)
+};
+
+}
+
+#endif  // STORAGE_LEVELDB_TABLE_FILTER_BLOCK_H_

data/leveldb/table/filter_block_test.cc

@@ -0,0 +1,128 @@
+// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "table/filter_block.h"
+
+#include "leveldb/filter_policy.h"
+#include "util/coding.h"
+#include "util/hash.h"
+#include "util/logging.h"
+#include "util/testharness.h"
+#include "util/testutil.h"
+
+namespace leveldb {
+
+// For testing: emit an array with one hash value per key
+class TestHashFilter : public FilterPolicy {
+ public:
+  virtual const char* Name() const {
+    return "TestHashFilter";
+  }
+
+  virtual void CreateFilter(const Slice* keys, int n, std::string* dst) const {
+    for (int i = 0; i < n; i++) {
+      uint32_t h = Hash(keys[i].data(), keys[i].size(), 1);
+      PutFixed32(dst, h);
+    }
+  }
+
+  virtual bool KeyMayMatch(const Slice& key, const Slice& filter) const {
+    uint32_t h = Hash(key.data(), key.size(), 1);
+    for (int i = 0; i + 4 <= filter.size(); i += 4) {
+      if (h == DecodeFixed32(filter.data() + i)) {
+        return true;
+      }
+    }
+    return false;
+  }
+};
+
+class FilterBlockTest {
+ public:
+  TestHashFilter policy_;
+};
+
+TEST(FilterBlockTest, EmptyBuilder) {
+  FilterBlockBuilder builder(&policy_);
+  Slice block = builder.Finish();
+  ASSERT_EQ("\\x00\\x00\\x00\\x00\\x0b", EscapeString(block));
+  FilterBlockReader reader(&policy_, block);
+  ASSERT_TRUE(reader.KeyMayMatch(0, "foo"));
+  ASSERT_TRUE(reader.KeyMayMatch(100000, "foo"));
+}
+
+TEST(FilterBlockTest, SingleChunk) {
+  FilterBlockBuilder builder(&policy_);
+  builder.StartBlock(100);
+  builder.AddKey("foo");
+  builder.AddKey("bar");
+  builder.AddKey("box");
+  builder.StartBlock(200);
+  builder.AddKey("box");
+  builder.StartBlock(300);
+  builder.AddKey("hello");
+  Slice block = builder.Finish();
+  FilterBlockReader reader(&policy_, block);
+  ASSERT_TRUE(reader.KeyMayMatch(100, "foo"));
+  ASSERT_TRUE(reader.KeyMayMatch(100, "bar"));
+  ASSERT_TRUE(reader.KeyMayMatch(100, "box"));
+  ASSERT_TRUE(reader.KeyMayMatch(100, "hello"));
+  ASSERT_TRUE(reader.KeyMayMatch(100, "foo"));
+  ASSERT_TRUE(! reader.KeyMayMatch(100, "missing"));
+  ASSERT_TRUE(! reader.KeyMayMatch(100, "other"));
+}
+
+TEST(FilterBlockTest, MultiChunk) {
+  FilterBlockBuilder builder(&policy_);
+
+  // First filter
+  builder.StartBlock(0);
+  builder.AddKey("foo");
+  builder.StartBlock(2000);
+  builder.AddKey("bar");
+
+  // Second filter
+  builder.StartBlock(3100);
+  builder.AddKey("box");
+
+  // Third filter is empty
+
+  // Last filter
+  builder.StartBlock(9000);
+  builder.AddKey("box");
+  builder.AddKey("hello");
+
+  Slice block = builder.Finish();
+  FilterBlockReader reader(&policy_, block);
+
+  // Check first filter
+  ASSERT_TRUE(reader.KeyMayMatch(0, "foo"));
+  ASSERT_TRUE(reader.KeyMayMatch(2000, "bar"));
+  ASSERT_TRUE(! reader.KeyMayMatch(0, "box"));
+  ASSERT_TRUE(! reader.KeyMayMatch(0, "hello"));
+
+  // Check second filter
+  ASSERT_TRUE(reader.KeyMayMatch(3100, "box"));
+  ASSERT_TRUE(! reader.KeyMayMatch(3100, "foo"));
+  ASSERT_TRUE(! reader.KeyMayMatch(3100, "bar"));
+  ASSERT_TRUE(! reader.KeyMayMatch(3100, "hello"));
+
+  // Check third filter (empty)
+  ASSERT_TRUE(! reader.KeyMayMatch(4100, "foo"));
+  ASSERT_TRUE(! reader.KeyMayMatch(4100, "bar"));
+  ASSERT_TRUE(! reader.KeyMayMatch(4100, "box"));
+  ASSERT_TRUE(! reader.KeyMayMatch(4100, "hello"));
+
+  // Check last filter
+  ASSERT_TRUE(reader.KeyMayMatch(9000, "box"));
+  ASSERT_TRUE(reader.KeyMayMatch(9000, "hello"));
+  ASSERT_TRUE(! reader.KeyMayMatch(9000, "foo"));
+  ASSERT_TRUE(! reader.KeyMayMatch(9000, "bar"));
+}
+
+}  // namespace leveldb
+
+int main(int argc, char** argv) {
+  return leveldb::test::RunAllTests();
+}

data/leveldb/table/format.cc

@@ -0,0 +1,145 @@
+// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "table/format.h"
+
+#include "leveldb/env.h"
+#include "port/port.h"
+#include "table/block.h"
+#include "util/coding.h"
+#include "util/crc32c.h"
+
+namespace leveldb {
+
+void BlockHandle::EncodeTo(std::string* dst) const {
+  // Sanity check that all fields have been set
+  assert(offset_ != ~static_cast<uint64_t>(0));
+  assert(size_ != ~static_cast<uint64_t>(0));
+  PutVarint64(dst, offset_);
+  PutVarint64(dst, size_);
+}
+
+Status BlockHandle::DecodeFrom(Slice* input) {
+  if (GetVarint64(input, &offset_) &&
+      GetVarint64(input, &size_)) {
+    return Status::OK();
+  } else {
+    return Status::Corruption("bad block handle");
+  }
+}
+
+void Footer::EncodeTo(std::string* dst) const {
+#ifndef NDEBUG
+  const size_t original_size = dst->size();
+#endif
+  metaindex_handle_.EncodeTo(dst);
+  index_handle_.EncodeTo(dst);
+  dst->resize(2 * BlockHandle::kMaxEncodedLength);  // Padding
+  PutFixed32(dst, static_cast<uint32_t>(kTableMagicNumber & 0xffffffffu));
+  PutFixed32(dst, static_cast<uint32_t>(kTableMagicNumber >> 32));
+  assert(dst->size() == original_size + kEncodedLength);
+}
+
+Status Footer::DecodeFrom(Slice* input) {
+  const char* magic_ptr = input->data() + kEncodedLength - 8;
+  const uint32_t magic_lo = DecodeFixed32(magic_ptr);
+  const uint32_t magic_hi = DecodeFixed32(magic_ptr + 4);
+  const uint64_t magic = ((static_cast<uint64_t>(magic_hi) << 32) |
+                          (static_cast<uint64_t>(magic_lo)));
+  if (magic != kTableMagicNumber) {
+    return Status::InvalidArgument("not an sstable (bad magic number)");
+  }
+
+  Status result = metaindex_handle_.DecodeFrom(input);
+  if (result.ok()) {
+    result = index_handle_.DecodeFrom(input);
+  }
+  if (result.ok()) {
+    // We skip over any leftover data (just padding for now) in "input"
+    const char* end = magic_ptr + 8;
+    *input = Slice(end, input->data() + input->size() - end);
+  }
+  return result;
+}
+
+Status ReadBlock(RandomAccessFile* file,
+                 const ReadOptions& options,
+                 const BlockHandle& handle,
+                 BlockContents* result) {
+  result->data = Slice();
+  result->cachable = false;
+  result->heap_allocated = false;
+
+  // Read the block contents as well as the type/crc footer.
+  // See table_builder.cc for the code that built this structure.
+  size_t n = static_cast<size_t>(handle.size());
+  char* buf = new char[n + kBlockTrailerSize];
+  Slice contents;
+  Status s = file->Read(handle.offset(), n + kBlockTrailerSize, &contents, buf);
+  if (!s.ok()) {
+    delete[] buf;
+    return s;
+  }
+  if (contents.size() != n + kBlockTrailerSize) {
+    delete[] buf;
+    return Status::Corruption("truncated block read");
+  }
+
+  // Check the crc of the type and the block contents
+  const char* data = contents.data();    // Pointer to where Read put the data
+  if (options.verify_checksums) {
+    const uint32_t crc = crc32c::Unmask(DecodeFixed32(data + n + 1));
+    const uint32_t actual = crc32c::Value(data, n + 1);
+    if (actual != crc) {
+      delete[] buf;
+      s = Status::Corruption("block checksum mismatch");
+      return s;
+    }
+  }
+
+  switch (data[n]) {
+    case kNoCompression:
+      if (data != buf) {
+        // File implementation gave us pointer to some other data.
+        // Use it directly under the assumption that it will be live
+        // while the file is open.
+        delete[] buf;
+        result->data = Slice(data, n);
+        result->heap_allocated = false;
+        result->cachable = false;  // Do not double-cache
+      } else {
+        result->data = Slice(buf, n);
+        result->heap_allocated = true;
+        result->cachable = true;
+      }
+
+      // Ok
+      break;
+    case kSnappyCompression: {
+      size_t ulength = 0;
+      if (!port::Snappy_GetUncompressedLength(data, n, &ulength)) {
+        delete[] buf;
+        return Status::Corruption("corrupted compressed block contents");
+      }
+      char* ubuf = new char[ulength];
+      if (!port::Snappy_Uncompress(data, n, ubuf)) {
+        delete[] buf;
+        delete[] ubuf;
+        return Status::Corruption("corrupted compressed block contents");
+      }
+      delete[] buf;
+      result->data = Slice(ubuf, ulength);
+      result->heap_allocated = true;
+      result->cachable = true;
+      break;
+    }
+    default:
+      delete[] buf;
+      return Status::Corruption("bad block type");
+  }
+
+  return Status::OK();
+}
+
+}  // namespace leveldb