Make compression options configurable. These include window-bits, level and strategy for ZlibCompression

Summary: Leveldb currently uses windowBits=-14 while using zlib compression.(It was earlier 15). This makes the setting configurable. Related changes here: https://reviews.facebook.net/D6105

Test Plan: make all check

Reviewers: dhruba, MarkCallaghan, sheki, heyongqiang

Differential Revision: https://reviews.facebook.net/D6393

db/c.cc

@@ -502,6 +502,13 @@ void leveldb_options_set_compression(leveldb_options_t* opt, int t) {
   opt->rep.compression = static_cast<CompressionType>(t);
 }
 
+void leveldb_options_set_compression_options(
+    leveldb_options_t* opt, int w_bits, int level, int strategy) {
+  opt->rep.compression_opts.window_bits = w_bits;
+  opt->rep.compression_opts.level = level;
+  opt->rep.compression_opts.strategy = strategy;
+}
+
 void leveldb_options_set_disable_data_sync(
     leveldb_options_t* opt, bool disable_data_sync) {
   opt->rep.disableDataSync = disable_data_sync;

db/c_test.c

@@ -187,6 +187,7 @@ int main(int argc, char** argv) {
   leveldb_options_set_block_size(options, 1024);
   leveldb_options_set_block_restart_interval(options, 8);
   leveldb_options_set_compression(options, leveldb_no_compression);
+  leveldb_options_set_compression_options(options, -14, -1, 0);
 
   roptions = leveldb_readoptions_create();
   leveldb_readoptions_set_verify_checksums(roptions, 1);

db/db_bench.cc

@@ -509,15 +509,18 @@ class Benchmark {
 
       switch (FLAGS_compression_type) {
         case kSnappyCompression:
-          result = port::Snappy_Compress(text, strlen(text), &compressed);
+          result = port::Snappy_Compress(Options().compression_opts, text,
+                                         strlen(text), &compressed);
           name = "Snappy";
           break;
         case kZlibCompression:
-          result = port::Zlib_Compress(text, strlen(text), &compressed);
+          result = port::Zlib_Compress(Options().compression_opts, text,
+                                       strlen(text), &compressed);
           name = "Zlib";
           break;
         case kBZip2Compression:
-          result = port::BZip2_Compress(text, strlen(text), &compressed);
+          result = port::BZip2_Compress(Options().compression_opts, text,
+                                        strlen(text), &compressed);
           name = "BZip2";
           break;
       }
@@ -855,7 +858,8 @@ class Benchmark {
     bool ok = true;
     std::string compressed;
     while (ok && bytes < 1024 * 1048576) {  // Compress 1G
-      ok = port::Snappy_Compress(input.data(), input.size(), &compressed);
+      ok = port::Snappy_Compress(Options().compression_opts, input.data(),
+                                 input.size(), &compressed);
       produced += compressed.size();
       bytes += input.size();
       thread->stats.FinishedSingleOp(NULL);
@@ -876,7 +880,8 @@ class Benchmark {
     RandomGenerator gen;
     Slice input = gen.Generate(Options().block_size);
     std::string compressed;
-    bool ok = port::Snappy_Compress(input.data(), input.size(), &compressed);
+    bool ok = port::Snappy_Compress(Options().compression_opts, input.data(),
+                                    input.size(), &compressed);
     int64_t bytes = 0;
     char* uncompressed = new char[input.size()];
     while (ok && bytes < 1024 * 1048576) {  // Compress 1G

db/db_test.cc

@@ -20,22 +20,22 @@
 
 namespace leveldb {
 
-static bool SnappyCompressionSupported() {	
+static bool SnappyCompressionSupported(const CompressionOptions& options) {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::Snappy_Compress(in.data(), in.size(), &out);	
+  return port::Snappy_Compress(options, in.data(), in.size(), &out);
 }
 
-static bool ZlibCompressionSupported() {	
+static bool ZlibCompressionSupported(const CompressionOptions& options) {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::Zlib_Compress(in.data(), in.size(), &out);	
+  return port::Zlib_Compress(options, in.data(), in.size(), &out);
 }
 
-static bool BZip2CompressionSupported() {	
+static bool BZip2CompressionSupported(const CompressionOptions& options) {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::BZip2_Compress(in.data(), in.size(), &out);	
+  return port::BZip2_Compress(options, in.data(), in.size(), &out);
 }
 
 static std::string RandomString(Random* rnd, int len) {
@@ -1105,22 +1105,25 @@ void MinLevelHelper(DBTest* self, Options& options) {
   ASSERT_EQ(self->NumTableFilesAtLevel(1), 1);
 }
 
-TEST(DBTest, MinLevelToCompress) {	
+void MinLevelToCompress(CompressionType& type, Options& options, int wbits,
-  Options options = CurrentOptions();	
+                        int lev, int strategy) {
+  fprintf(stderr, "Test with compression options : window_bits = %d, level =  %d
+                   , strategy = %d}\n", wbits, lev, strategy);
   options.write_buffer_size = 100<<10; //100KB
   options.num_levels = 3;
   options.max_mem_compaction_level = 0;
   options.level0_file_num_compaction_trigger = 3;
   options.create_if_missing = true;
-  CompressionType type;
 
-  if (SnappyCompressionSupported()) {	
+  if (SnappyCompressionSupported(CompressionOptions(wbits, lev, strategy))) {
     type = kSnappyCompression;
     fprintf(stderr, "using snappy\n");
-  } else if (ZlibCompressionSupported()) {	
+  } else if (ZlibCompressionSupported(
+               CompressionOptions(wbits, lev, strategy))) {
     type = kZlibCompression;
     fprintf(stderr, "using zlib\n");
-  } else if (BZip2CompressionSupported()) {	
+  } else if (BZip2CompressionSupported(
+               CompressionOptions(wbits, lev, strategy))) {
     type = kBZip2Compression;
     fprintf(stderr, "using bzip2\n");
   } else {
@@ -1136,6 +1139,29 @@ TEST(DBTest, MinLevelToCompress) {
   for (int i = 1; i < options.num_levels; i++) {
     options.compression_per_level[i] = type;
   }
+}
+TEST(DBTest, MinLevelToCompress1) {
+  Options options = CurrentOptions();
+  CompressionType type;
+  MinLevelToCompress(type, options, -14, -1, 0);
+  Reopen(&options);
+  MinLevelHelper(this, options);
+
+  // do not compress L0 and L1
+  for (int i = 0; i < 2; i++) {
+    options.compression_per_level[i] = kNoCompression;
+  }
+  for (int i = 2; i < options.num_levels; i++) {
+    options.compression_per_level[i] = type;
+  }
+  DestroyAndReopen(&options);
+  MinLevelHelper(this, options);
+}
+
+TEST(DBTest, MinLevelToCompress2) {
+  Options options = CurrentOptions();
+  CompressionType type;
+  MinLevelToCompress(type, options, 15, -1, 0);
   Reopen(&options);
   MinLevelHelper(this, options);
 

include/leveldb/options.h

@@ -32,6 +32,19 @@ enum CompressionType {
   kBZip2Compression = 0x3
 };
 
+// Compression options for different compression algorithms like Zlib
+struct CompressionOptions {
+  int window_bits;
+  int level;
+  int strategy;
+  CompressionOptions():window_bits(-14),
+                       level(-1),
+                       strategy(0){}
+  CompressionOptions(int wbits, int lev, int strategy):window_bits(wbits),
+                                                       level(lev),
+                                                       strategy(strategy){}
+};
+
 // Options to control the behavior of a database (passed to DB::Open)
 struct Options {
   // -------------------
@@ -148,6 +161,9 @@ struct Options {
   // java/C api hard to construct.
   CompressionType* compression_per_level;
 
+  //different options for compression algorithms
+  CompressionOptions compression_opts;
+
   // If non-NULL, use the specified filter policy to reduce disk reads.
   // Many applications will benefit from passing the result of
   // NewBloomFilterPolicy() here.

port/port_posix.h

@@ -44,6 +44,7 @@
 #include <stdint.h>
 #include <string>
 #include <string.h>
+#include "leveldb/options.h"
 #include "port/atomic_pointer.h"
 
 #ifndef PLATFORM_IS_LITTLE_ENDIAN
@@ -131,8 +132,8 @@ typedef pthread_once_t OnceType;
 #define LEVELDB_ONCE_INIT PTHREAD_ONCE_INIT
 extern void InitOnce(OnceType* once, void (*initializer)());
 
-inline bool Snappy_Compress(const char* input, size_t length,
+inline bool Snappy_Compress(const CompressionOptions& opts, const char* input,
-                            ::std::string* output) {
+                            size_t length, ::std::string* output) {
 #ifdef SNAPPY
   output->resize(snappy::MaxCompressedLength(length));
   size_t outlen;
@@ -162,9 +163,8 @@ inline bool Snappy_Uncompress(const char* input, size_t length,
 #endif
 }
 
-inline bool Zlib_Compress(const char* input, size_t length,
+inline bool Zlib_Compress(const CompressionOptions& opts, const char* input,
-    ::std::string* output, int windowBits = -14, int level = -1,
+                          size_t length, ::std::string* output) {
-     int strategy = 0) {
 #ifdef ZLIB
   // The memLevel parameter specifies how much memory should be allocated for
   // the internal compression state.
@@ -174,8 +174,8 @@ inline bool Zlib_Compress(const char* input, size_t length,
   static const int memLevel = 8;
   z_stream _stream;
   memset(&_stream, 0, sizeof(z_stream));
-  int st = deflateInit2(&_stream, level, Z_DEFLATED, windowBits,
+  int st = deflateInit2(&_stream, opts.level, Z_DEFLATED, opts.window_bits,
-                        memLevel, strategy);
+                        memLevel, opts.strategy);
   if (st != Z_OK) {
     return false;
   }
@@ -284,8 +284,8 @@ inline char* Zlib_Uncompress(const char* input_data, size_t input_length,
   return NULL;
 }
 
-inline bool BZip2_Compress(const char* input, size_t length,
+inline bool BZip2_Compress(const CompressionOptions& opts, const char* input,
-    ::std::string* output) {
+                           size_t length, ::std::string* output) {
 #ifdef BZIP2
   bz_stream _stream;
   memset(&_stream, 0, sizeof(bz_stream));

table/table_builder.cc

@@ -175,7 +175,8 @@ void TableBuilder::WriteBlock(BlockBuilder* block, BlockHandle* handle) {
 
     case kSnappyCompression: {
       std::string* compressed = &r->compressed_output;
-      if (port::Snappy_Compress(raw.data(), raw.size(), compressed) &&
+      if (port::Snappy_Compress(r->options.compression_opts, raw.data(),
+                                raw.size(), compressed) &&
           GoodCompressionRatio(compressed->size(), raw.size())) {
         block_contents = *compressed;
       } else {
@@ -187,7 +188,8 @@ void TableBuilder::WriteBlock(BlockBuilder* block, BlockHandle* handle) {
       break;
     }
     case kZlibCompression:
-      if (port::Zlib_Compress(raw.data(), raw.size(), compressed) &&
+      if (port::Zlib_Compress(r->options.compression_opts, raw.data(),
+                              raw.size(), compressed) &&
           GoodCompressionRatio(compressed->size(), raw.size())) {
         block_contents = *compressed;
       } else {
@@ -198,7 +200,8 @@ void TableBuilder::WriteBlock(BlockBuilder* block, BlockHandle* handle) {
       }
       break;
     case kBZip2Compression:
-      if (port::BZip2_Compress(raw.data(), raw.size(), compressed) &&
+      if (port::BZip2_Compress(r->options.compression_opts, raw.data(),
+                               raw.size(), compressed) &&
           GoodCompressionRatio(compressed->size(), raw.size())) {
         block_contents = *compressed;
       } else {

table/table_test.cc

@@ -247,6 +247,7 @@ class TableConstructor: public Constructor {
     source_ = new StringSource(sink.contents());
     Options table_options;
     table_options.comparator = options.comparator;
+    table_options.compression_opts = options.compression_opts;
     return Table::Open(table_options, source_, sink.contents().size(), &table_);
   }
 
@@ -399,19 +400,22 @@ class DBConstructor: public Constructor {
 static bool SnappyCompressionSupported() {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::Snappy_Compress(in.data(), in.size(), &out);
+  return port::Snappy_Compress(Options().compression_opts, in.data(), in.size(),
+                               &out);
 }
 
 static bool ZlibCompressionSupported() {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::Zlib_Compress(in.data(), in.size(), &out);
+  return port::Zlib_Compress(Options().compression_opts, in.data(), in.size(),
+                             &out);
 }
 
 static bool BZip2CompressionSupported() {
   std::string out;
   Slice in = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
-  return port::BZip2_Compress(in.data(), in.size(), &out);
+  return port::BZip2_Compress(Options().compression_opts, in.data(), in.size(),
+                              &out);
 }
 
 enum TestType {

util/options.cc

@@ -85,6 +85,12 @@ Options::Dump(
     Log(log,"     Options.max_log_file_size: %d", max_log_file_size);
     Log(log," Options.db_stats_log_interval: %d",
         db_stats_log_interval);
+    Log(log,"           Options.compression_opts.window_bits: %d",
+        compression_opts.window_bits);
+    Log(log,"                 Options.compression_opts.level: %d",
+        compression_opts.level);
+    Log(log,"              Options.compression_opts.strategy: %d",
+        compression_opts.strategy);
     Log(log,"     Options.level0_file_num_compaction_trigger: %d",
         level0_file_num_compaction_trigger);
     Log(log,"         Options.level0_slowdown_writes_trigger: %d",