snappy 0.1.0 → 0.2.0

data/vendor/snappy/NEWS

@@ -1,3 +1,11 @@
+Snappy v1.1.8, January 15th 2020:
+
+  * Small performance improvements.
+
+  * Removed snappy::string alias for std::string.
+
+  * Improved CMake configuration.
+
 Snappy v1.1.7, August 24th 2017:
 
   * Improved CMake build support for 64-bit Linux distributions.

data/vendor/snappy/README.md

@@ -51,7 +51,7 @@ In particular:
 
  - Snappy uses 64-bit operations in several places to process more data at
    once than would otherwise be possible.
- - Snappy assumes unaligned 32- and 64-bit loads and stores are cheap.
+ - Snappy assumes unaligned 32 and 64-bit loads and stores are cheap.
    On some platforms, these must be emulated with single-byte loads
    and stores, which is much slower.
  - Snappy assumes little-endian throughout, and needs to byte-swap data in
@@ -65,32 +65,37 @@ are of course most welcome; see "Contact", below.
 Building
 ========
 
-CMake is supported and autotools will soon be deprecated.
-You need CMake 3.4 or above to build:
-
-  mkdir build
-  cd build && cmake ../ && make
+You need the CMake version specified in [CMakeLists.txt](./CMakeLists.txt)
+or later to build:
 
+```bash
+mkdir build
+cd build && cmake ../ && make
+```
 
 Usage
 =====
 
 Note that Snappy, both the implementation and the main interface,
 is written in C++. However, several third-party bindings to other languages
-are available; see the home page at http://google.github.io/snappy/
-for more information. Also, if you want to use Snappy from C code, you can
-use the included C bindings in snappy-c.h.
+are available; see the [home page](docs/README.md) for more information.
+Also, if you want to use Snappy from C code, you can use the included C
+bindings in snappy-c.h.
 
 To use Snappy from your own C++ program, include the file "snappy.h" from
 your calling file, and link against the compiled library.
 
 There are many ways to call Snappy, but the simplest possible is
 
-  snappy::Compress(input.data(), input.size(), &output);
+```c++
+snappy::Compress(input.data(), input.size(), &output);
+```
 
 and similarly
 
-  snappy::Uncompress(input.data(), input.size(), &output);
+```c++
+snappy::Uncompress(input.data(), input.size(), &output);
+```
 
 where "input" and "output" are both instances of std::string.
 
@@ -112,12 +117,12 @@ tests to verify you have not broken anything. Note that if you have the
 Google Test library installed, unit test behavior (especially failures) will be
 significantly more user-friendly. You can find Google Test at
 
-  http://github.com/google/googletest
+  https://github.com/google/googletest
 
 You probably also want the gflags library for handling of command-line flags;
 you can find it at
 
-  http://gflags.github.io/gflags/
+  https://gflags.github.io/gflags/
 
 In addition to the unit tests, snappy contains microbenchmarks used to
 tune compression and decompression performance. These are automatically run
@@ -140,10 +145,4 @@ Contact
 =======
 
 Snappy is distributed through GitHub. For the latest version, a bug tracker,
-and other information, see
-
-  http://google.github.io/snappy/
-
-or the repository at
-
-  https://github.com/google/snappy
+and other information, see https://github.com/google/snappy.

data/vendor/snappy/cmake/SnappyConfig.cmake.in

@@ -0,0 +1,33 @@
+# Copyright 2019 Google Inc. All Rights Reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#     * Redistributions of source code must retain the above copyright
+# notice, this list of conditions and the following disclaimer.
+#     * Redistributions in binary form must reproduce the above
+# copyright notice, this list of conditions and the following disclaimer
+# in the documentation and/or other materials provided with the
+# distribution.
+#     * Neither the name of Google Inc. nor the names of its
+# contributors may be used to endorse or promote products derived from
+# this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+@PACKAGE_INIT@
+
+include("${CMAKE_CURRENT_LIST_DIR}/SnappyTargets.cmake")
+
+check_required_components(Snappy)

data/vendor/snappy/cmake/config.h.in

@@ -28,12 +28,6 @@
 /* Define to 1 if you have the `z' library (-lz). */
 #cmakedefine HAVE_LIBZ 1
 
-/* Define to 1 if you have the <stddef.h> header file. */
-#cmakedefine HAVE_STDDEF_H 1
-
-/* Define to 1 if you have the <stdint.h> header file. */
-#cmakedefine HAVE_STDINT_H 1
-
 /* Define to 1 if you have the <sys/endian.h> header file. */
 #cmakedefine HAVE_SYS_ENDIAN_H 1
 
@@ -55,6 +49,12 @@
 /* Define to 1 if you have the <windows.h> header file. */
 #cmakedefine HAVE_WINDOWS_H 1
 
+/* Define to 1 if you target processors with SSSE3+ and have <tmmintrin.h>. */
+#cmakedefine01 SNAPPY_HAVE_SSSE3
+
+/* Define to 1 if you target processors with BMI2+ and have <bmi2intrin.h>. */
+#cmakedefine01 SNAPPY_HAVE_BMI2
+
 /* Define to 1 if your processor stores words with the most significant byte
    first (like Motorola and SPARC, unlike Intel and VAX). */
 #cmakedefine SNAPPY_IS_BIG_ENDIAN 1

data/vendor/snappy/docs/README.md

@@ -0,0 +1,72 @@
+Snappy is a compression/decompression library. It does not aim for maximum
+compression, or compatibility with any other compression library; instead, it
+aims for very high speeds and reasonable compression. For instance, compared
+to the fastest mode of zlib, Snappy is an order of magnitude faster for most
+inputs, but the resulting compressed files are anywhere from 20% to 100%
+bigger. On a single core of a Core i7 processor in 64-bit mode, Snappy
+compresses at about 250 MB/sec or more and decompresses at about 500 MB/sec
+or more.
+
+Snappy is widely used inside Google, in everything from BigTable and MapReduce
+to our internal RPC systems. (Snappy has previously been referred to as "Zippy"
+in some presentations and the likes.)
+
+For more information, please see the [README](../README.md). Benchmarks against
+a few other compression libraries (zlib, LZO, LZF, FastLZ, and QuickLZ) are
+included in the source code distribution. The source code also contains a
+[formal format specification](../format_description.txt), as well
+as a specification for a [framing format](../framing_format.txt) useful for
+higher-level framing and encapsulation of Snappy data, e.g. for transporting
+Snappy-compressed data across HTTP in a streaming fashion. Note that the Snappy
+distribution currently has no code implementing the latter, but some of the
+ports do (see below).
+
+Snappy is written in C++, but C bindings are included, and several bindings to
+other languages are maintained by third parties:
+
+* C#: [Snappy for .NET](http://snappy4net.codeplex.com/) (P/Invoke wrapper),
+  [Snappy.NET](http://snappy.angeloflogic.com/) (P/Invoke wrapper),
+  [Snappy.Sharp](https://github.com/jeffesp/Snappy.Sharp) (native
+  reimplementation)
+* [C port](http://github.com/andikleen/snappy-c)
+* [C++ MSVC packaging](http://snappy.angeloflogic.com/) (plus Windows binaries,
+  NuGet packages and command-line tool)
+* Common Lisp: [Library bindings](http://flambard.github.com/thnappy/),
+  [native reimplementation](https://github.com/brown/snappy)
+* Erlang: [esnappy](https://github.com/thekvs/esnappy),
+  [snappy-erlang-nif](https://github.com/fdmanana/snappy-erlang-nif)
+* [Go](https://github.com/golang/snappy/)
+* [Haskell](http://hackage.haskell.org/package/snappy)
+* [Haxe](https://github.com/MaddinXx/hxsnappy) (C++/Neko)
+* [iOS packaging](https://github.com/ideawu/snappy-ios)
+* Java: [JNI wrapper](https://github.com/xerial/snappy-java) (including the
+  framing format), [native reimplementation](http://code.google.com/p/jsnappy/),
+  [other native reimplementation](https://github.com/dain/snappy) (including
+  the framing format)
+* [Lua](https://github.com/forhappy/lua-snappy)
+* [Node.js](https://github.com/kesla/node-snappy) (including the [framing
+  format](https://github.com/kesla/node-snappy-stream))
+* [Perl](http://search.cpan.org/dist/Compress-Snappy/)
+* [PHP](https://github.com/kjdev/php-ext-snappy)
+* [Python](http://pypi.python.org/pypi/python-snappy) (including a command-line
+  tool for the framing format)
+* [R](https://github.com/lulyon/R-snappy)
+* [Ruby](https://github.com/miyucy/snappy)
+* [Rust](https://github.com/BurntSushi/rust-snappy)
+* [Smalltalk](https://github.com/mumez/sqnappy) (including the framing format)
+
+Snappy is used or is available as an alternative in software such as
+
+* [MongoDB](https://www.mongodb.com/)
+* [Cassandra](http://cassandra.apache.org/)
+* [Couchbase](http://www.couchbase.com/)
+* [Hadoop](http://hadoop.apache.org/)
+* [LessFS](http://www.lessfs.com/wordpress/)
+* [LevelDB](https://github.com/google/leveldb) (which is in turn used by
+  [Google Chrome](http://chrome.google.com/))
+* [Lucene](http://lucene.apache.org/)
+* [VoltDB](http://voltdb.com/)
+
+If you know of more, do not hesitate to let us know. The easiest way to get in
+touch is via the
+[Snappy discussion mailing list](http://groups.google.com/group/snappy-compression).

data/vendor/snappy/snappy-internal.h

@@ -36,19 +36,26 @@
 namespace snappy {
 namespace internal {
 
+// Working memory performs a single allocation to hold all scratch space
+// required for compression.
 class WorkingMemory {
  public:
-  WorkingMemory() : large_table_(NULL) { }
-  ~WorkingMemory() { delete[] large_table_; }
+  explicit WorkingMemory(size_t input_size);
+  ~WorkingMemory();
 
   // Allocates and clears a hash table using memory in "*this",
   // stores the number of buckets in "*table_size" and returns a pointer to
   // the base of the hash table.
-  uint16* GetHashTable(size_t input_size, int* table_size);
+  uint16* GetHashTable(size_t fragment_size, int* table_size) const;
+  char* GetScratchInput() const { return input_; }
+  char* GetScratchOutput() const { return output_; }
 
  private:
-  uint16 small_table_[1<<10];    // 2KB
-  uint16* large_table_;          // Allocated only when needed
+  char* mem_;      // the allocated memory, never nullptr
+  size_t size_;    // the size of the allocated memory, never 0
+  uint16* table_;  // the pointer to the hashtable
+  char* input_;    // the pointer to the input scratch buffer
+  char* output_;   // the pointer to the output scratch buffer
 
   // No copying
   WorkingMemory(const WorkingMemory&);

data/vendor/snappy/snappy-stubs-internal.cc

@@ -33,7 +33,7 @@
 
 namespace snappy {
 
-void Varint::Append32(string* s, uint32 value) {
+void Varint::Append32(std::string* s, uint32 value) {
   char buf[Varint::kMax32];
   const char* p = Varint::Encode32(buf, value);
   s->append(buf, p - buf);

data/vendor/snappy/snappy-stubs-internal.h

@@ -53,6 +53,18 @@
 #include <intrin.h>
 #endif  // defined(_MSC_VER)
 
+#ifndef __has_feature
+#define __has_feature(x) 0
+#endif
+
+#if __has_feature(memory_sanitizer)
+#include <sanitizer/msan_interface.h>
+#define SNAPPY_ANNOTATE_MEMORY_IS_INITIALIZED(address, size) \
+    __msan_unpoison((address), (size))
+#else
+#define SNAPPY_ANNOTATE_MEMORY_IS_INITIALIZED(address, size) /* empty */
+#endif  // __has_feature(memory_sanitizer)
+
 #include "snappy-stubs-public.h"
 
 #if defined(__x86_64__)
@@ -187,7 +199,7 @@ struct Unaligned32Struct {
     ((reinterpret_cast< ::snappy::base::internal::Unaligned32Struct *>(_p))->value = \
          (_val))
 
-// TODO(user): NEON supports unaligned 64-bit loads and stores.
+// TODO: NEON supports unaligned 64-bit loads and stores.
 // See if that would be more efficient on platforms supporting it,
 // at least for copies.
 
@@ -353,6 +365,9 @@ class LittleEndian {
 // Some bit-manipulation functions.
 class Bits {
  public:
+  // Return floor(log2(n)) for positive integer n.
+  static int Log2FloorNonZero(uint32 n);
+
   // Return floor(log2(n)) for positive integer n.  Returns -1 iff n == 0.
   static int Log2Floor(uint32 n);
 
@@ -373,50 +388,72 @@ class Bits {
 
 #ifdef HAVE_BUILTIN_CTZ
 
+inline int Bits::Log2FloorNonZero(uint32 n) {
+  assert(n != 0);
+  // (31 ^ x) is equivalent to (31 - x) for x in [0, 31]. An easy proof
+  // represents subtraction in base 2 and observes that there's no carry.
+  //
+  // GCC and Clang represent __builtin_clz on x86 as 31 ^ _bit_scan_reverse(x).
+  // Using "31 ^" here instead of "31 -" allows the optimizer to strip the
+  // function body down to _bit_scan_reverse(x).
+  return 31 ^ __builtin_clz(n);
+}
+
 inline int Bits::Log2Floor(uint32 n) {
-  return n == 0 ? -1 : 31 ^ __builtin_clz(n);
+  return (n == 0) ? -1 : Bits::Log2FloorNonZero(n);
 }
 
 inline int Bits::FindLSBSetNonZero(uint32 n) {
+  assert(n != 0);
   return __builtin_ctz(n);
 }
 
 #if defined(ARCH_K8) || defined(ARCH_PPC) || defined(ARCH_ARM)
 inline int Bits::FindLSBSetNonZero64(uint64 n) {
+  assert(n != 0);
   return __builtin_ctzll(n);
 }
 #endif  // defined(ARCH_K8) || defined(ARCH_PPC) || defined(ARCH_ARM)
 
 #elif defined(_MSC_VER)
 
+inline int Bits::Log2FloorNonZero(uint32 n) {
+  assert(n != 0);
+  unsigned long where;
+  _BitScanReverse(&where, n);
+  return static_cast<int>(where);
+}
+
 inline int Bits::Log2Floor(uint32 n) {
   unsigned long where;
-  if (_BitScanReverse(&where, n)) {
-    return where;
-  } else {
-    return -1;
-  }
+  if (_BitScanReverse(&where, n))
+    return static_cast<int>(where);
+  return -1;
 }
 
 inline int Bits::FindLSBSetNonZero(uint32 n) {
+  assert(n != 0);
   unsigned long where;
-  if (_BitScanForward(&where, n)) return static_cast<int>(where);
+  if (_BitScanForward(&where, n))
+    return static_cast<int>(where);
   return 32;
 }
 
 #if defined(ARCH_K8) || defined(ARCH_PPC) || defined(ARCH_ARM)
 inline int Bits::FindLSBSetNonZero64(uint64 n) {
+  assert(n != 0);
   unsigned long where;
-  if (_BitScanForward64(&where, n)) return static_cast<int>(where);
+  if (_BitScanForward64(&where, n))
+    return static_cast<int>(where);
   return 64;
 }
 #endif  // defined(ARCH_K8) || defined(ARCH_PPC) || defined(ARCH_ARM)
 
 #else  // Portable versions.
 
-inline int Bits::Log2Floor(uint32 n) {
-  if (n == 0)
-    return -1;
+inline int Bits::Log2FloorNonZero(uint32 n) {
+  assert(n != 0);
+
   int log = 0;
   uint32 value = n;
   for (int i = 4; i >= 0; --i) {
@@ -431,7 +468,13 @@ inline int Bits::Log2Floor(uint32 n) {
   return log;
 }
 
+inline int Bits::Log2Floor(uint32 n) {
+  return (n == 0) ? -1 : Bits::Log2FloorNonZero(n);
+}
+
 inline int Bits::FindLSBSetNonZero(uint32 n) {
+  assert(n != 0);
+
   int rc = 31;
   for (int i = 4, shift = 1 << 4; i >= 0; --i) {
     const uint32 x = n << shift;
@@ -447,6 +490,8 @@ inline int Bits::FindLSBSetNonZero(uint32 n) {
 #if defined(ARCH_K8) || defined(ARCH_PPC) || defined(ARCH_ARM)
 // FindLSBSetNonZero64() is defined in terms of FindLSBSetNonZero().
 inline int Bits::FindLSBSetNonZero64(uint64 n) {
+  assert(n != 0);
+
   const uint32 bottombits = static_cast<uint32>(n);
   if (bottombits == 0) {
     // Bottom bits are zero, so scan in top bits
@@ -479,7 +524,7 @@ class Varint {
   static char* Encode32(char* ptr, uint32 v);
 
   // EFFECTS    Appends the varint representation of "value" to "*s".
-  static void Append32(string* s, uint32 value);
+  static void Append32(std::string* s, uint32 value);
 };
 
 inline const char* Varint::Parse32WithLimit(const char* p,
@@ -536,7 +581,7 @@ inline char* Varint::Encode32(char* sptr, uint32 v) {
 // replace this function with one that resizes the string without
 // filling the new space with zeros (if applicable) --
 // it will be non-portable but faster.
-inline void STLStringResizeUninitialized(string* s, size_t new_size) {
+inline void STLStringResizeUninitialized(std::string* s, size_t new_size) {
   s->resize(new_size);
 }
 
@@ -552,7 +597,7 @@ inline void STLStringResizeUninitialized(string* s, size_t new_size) {
 // (http://www.open-std.org/JTC1/SC22/WG21/docs/lwg-defects.html#530)
 // proposes this as the method. It will officially be part of the standard
 // for C++0x. This should already work on all current implementations.
-inline char* string_as_array(string* str) {
+inline char* string_as_array(std::string* str) {
   return str->empty() ? NULL : &*str->begin();
 }
 

data/vendor/snappy/snappy-stubs-public.h.in

@@ -1,5 +1,4 @@
 // Copyright 2011 Google Inc. All Rights Reserved.
-// Author: sesse@google.com (Steinar H. Gunderson)
 //
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
@@ -36,56 +35,37 @@
 #ifndef THIRD_PARTY_SNAPPY_OPENSOURCE_SNAPPY_STUBS_PUBLIC_H_
 #define THIRD_PARTY_SNAPPY_OPENSOURCE_SNAPPY_STUBS_PUBLIC_H_
 
-#if ${HAVE_STDINT_H_01}  // HAVE_STDINT_H
-#include <stdint.h>
-#endif  // HAVE_STDDEF_H
-
-#if ${HAVE_STDDEF_H_01}  // HAVE_STDDEF_H
-#include <stddef.h>
-#endif  // HAVE_STDDEF_H
+#include <cstddef>
+#include <cstdint>
+#include <string>
 
 #if ${HAVE_SYS_UIO_H_01}  // HAVE_SYS_UIO_H
 #include <sys/uio.h>
 #endif  // HAVE_SYS_UIO_H
 
-#define SNAPPY_MAJOR ${SNAPPY_MAJOR}
-#define SNAPPY_MINOR ${SNAPPY_MINOR}
-#define SNAPPY_PATCHLEVEL ${SNAPPY_PATCHLEVEL}
+#define SNAPPY_MAJOR ${PROJECT_VERSION_MAJOR}
+#define SNAPPY_MINOR ${PROJECT_VERSION_MINOR}
+#define SNAPPY_PATCHLEVEL ${PROJECT_VERSION_PATCH}
 #define SNAPPY_VERSION \
     ((SNAPPY_MAJOR << 16) | (SNAPPY_MINOR << 8) | SNAPPY_PATCHLEVEL)
 
-#include <string>
-
 namespace snappy {
 
-#if ${HAVE_STDINT_H_01}  // HAVE_STDINT_H
-typedef int8_t int8;
-typedef uint8_t uint8;
-typedef int16_t int16;
-typedef uint16_t uint16;
-typedef int32_t int32;
-typedef uint32_t uint32;
-typedef int64_t int64;
-typedef uint64_t uint64;
-#else
-typedef signed char int8;
-typedef unsigned char uint8;
-typedef short int16;
-typedef unsigned short uint16;
-typedef int int32;
-typedef unsigned int uint32;
-typedef long long int64;
-typedef unsigned long long uint64;
-#endif  // HAVE_STDINT_H
-
-typedef std::string string;
+using int8 = std::int8_t;
+using uint8 = std::uint8_t;
+using int16 = std::int16_t;
+using uint16 = std::uint16_t;
+using int32 = std::int32_t;
+using uint32 = std::uint32_t;
+using int64 = std::int64_t;
+using uint64 = std::uint64_t;
 
 #if !${HAVE_SYS_UIO_H_01}  // !HAVE_SYS_UIO_H
 // Windows does not have an iovec type, yet the concept is universally useful.
 // It is simple to define it ourselves, so we put it inside our own namespace.
 struct iovec {
-	void* iov_base;
-	size_t iov_len;
+  void* iov_base;
+  size_t iov_len;
 };
 #endif  // !HAVE_SYS_UIO_H