在Android中使用FlatBuffers
TorriK95
7年前
<h2>总览</h2> <p>先来看一下 <strong>FlatBuffers</strong> 项目已经为我们提供了什么，而我们在将 <strong>FlatBuffers</strong> 用到我们的项目中时又需要做什么的整体流程。如下图：</p> <p style="text-align:center"><img src="https://simg.open-open.com/show/09e8b7cca7c810c32b0c8a3606d25be4.jpg"></p> <p>在使用 <strong>FlatBuffers</strong> 时，我们需要以特殊的格式定义我们的结构化数据，保存为 .fbs 文件。 <strong>FlatBuffers</strong> 项目为我们提供了编译器，可用于将 .fbs 文件编译为Java文件，C++文件等，以用于我们的项目。 <strong>FlatBuffers</strong> 编译器在我们的开发机，比如Ubuntu，Mac上运行。这些源代码文件是基于 <strong>FlatBuffers</strong> 提供的Java库生成的，同时我们也需要利用这个Java库的一些接口来序列化或解析数据。</p> <p>我们将 <strong>FlatBuffers</strong> 编译器生成的Java文件及 <strong>FlatBuffers</strong> 的Java库导入我们的项目，就可以用 <strong>FlatBuffers</strong> 来对我们的结构化数据执行序列化和反序列化了。尽管每次手动执行 <strong>FlatBuffers</strong> 编译器生成Java文件非常麻烦，但不像 <strong>Protocol Buffers</strong> 那样，当前还没有Google官方提供的gradle插件可用。不过，我们这边开发了一个简单的 <strong>FlatBuffers</strong> gradle插件，后面会简单介绍一下，欢迎大家使用。</p> <p>接下来我们更详细地看一下上面流程中的各个部分。</p> <h2>下载、编译FlatBuffers编译器</h2> <p>我们可以在如下位置：</p> <pre> <code class="language-java">https://github.com/google/flatbuffers/releases</code></pre> <p>获取官方发布的打包好的版本。针对Windows平台有编译好的可执行安装文件，对其它平台还是打包的源文件。我们也可以指向clone repo的代码，进行手动编译。这里我们从GitHub上clone代码并手动编译编译器：</p> <pre> <code class="language-java">$ git clone https://github.com/google/flatbuffers.git Cloning into 'flatbuffers'... remote: Counting objects: 7340, done. remote: Compressing objects: 100% (46/46), done. remote: Total 7340 (delta 16), reused 0 (delta 0), pack-reused 7290 Receiving objects: 100% (7340/7340), 3.64 MiB | 115.00 KiB/s, done. Resolving deltas: 100% (4692/4692), done. Checking connectivity... done.</code></pre> <p>下载代码之后，我们需要用cmake工具来为flatbuffers生成Makefile文件并编译：</p> <pre> <code class="language-java">$ cd flatbuffers/ $ cmake CMakeLists.txt -- The C compiler identification is AppleClang 7.3.0.7030031 -- The CXX compiler identification is AppleClang 7.3.0.7030031 -- Check for working C compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc -- Check for working C compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Detecting C compile features -- Detecting C compile features - done -- Check for working CXX compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ -- Check for working CXX compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ -- works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Detecting CXX compile features -- Detecting CXX compile features - done -- Configuring done -- Generating done -- Build files have been written to: /Users/netease/Projects/OpenSource/flatbuffers $ make && make install</code></pre> <p>安装之后执行如下命令以确认已经装好：</p> <pre> <code class="language-java">$ flatc --version flatc version 1.4.0 (Dec 7 2016)</code></pre> <p>flatc没有为我们提供 <strong>--help</strong> 选项，不过加了错误的参数时这个工具会为我们展示详细的用法：</p> <pre> <code class="language-java">$ flatc --help flatc: unknown commandline argument: --help usage: flatc [OPTION]... FILE... [-- FILE...] --binary -b Generate wire format binaries for any data definitions. --json -t Generate text output for any data definitions. --cpp -c Generate C++ headers for tables/structs. --go -g Generate Go files for tables/structs. --java -j Generate Java classes for tables/structs. --js -s Generate JavaScript code for tables/structs. --csharp -n Generate C# classes for tables/structs. --python -p Generate Python files for tables/structs. --php Generate PHP files for tables/structs. -o PATH Prefix PATH to all generated files. -I PATH Search for includes in the specified path. -M Print make rules for generated files. --version Print the version number of flatc and exit. --strict-json Strict JSON: field names must be / will be quoted, no trailing commas in tables/vectors. --allow-non-utf8 Pass non-UTF-8 input through parser and emit nonstandard \x escapes in JSON. (Default is to raise parse error on non-UTF-8 input.) --defaults-json Output fields whose value is the default when writing JSON --unknown-json Allow fields in JSON that are not defined in the schema. These fields will be discared when generating binaries. --no-prefix Don't prefix enum values with the enum type in C++. --scoped-enums Use C++11 style scoped and strongly typed enums. also implies --no-prefix. --gen-includes (deprecated), this is the default behavior. If the original behavior is required (no include statements) use --no-includes. --no-includes Don't generate include statements for included schemas the generated file depends on (C++). --gen-mutable Generate accessors that can mutate buffers in-place. --gen-onefile Generate single output file for C#. --gen-name-strings Generate type name functions for C++. --escape-proto-ids Disable appending '_' in namespaces names. --gen-object-api Generate an additional object-based API. --cpp-ptr-type T Set object API pointer type (default std::unique_ptr) --raw-binary Allow binaries without file_indentifier to be read. This may crash flatc given a mismatched schema. --proto Input is a .proto, translate to .fbs. --schema Serialize schemas instead of JSON (use with -b) --conform FILE Specify a schema the following schemas should be an evolution of. Gives errors if not. --conform-includes Include path for the schema given with --conform PATH FILEs may be schemas, or JSON files (conforming to preceding schema) FILEs after the -- must be binary flatbuffer format files. Output files are named using the base file name of the input, and written to the current directory or the path given by -o. example: flatc -c -b schema1.fbs schema2.fbs data.json</code></pre> <h2>创建 .fbs 文件</h2> <p>flatc支持将为 <strong>Protocol Buffers</strong> 编写的 .proto 文件转换为 .fbs 文件，如：</p> <pre> <code class="language-java">$ ls addressbook.proto $ flatc --proto addressbook.proto $ ls -l total 16 -rw-r--r-- 1 netease staff 431 12 7 17:21 addressbook.fbs -rw-r--r--@ 1 netease staff 486 12 1 15:18 addressbook.proto</code></pre> <p>Protocol Buffers消息文件中的一些写法， <strong>FlatBuffers</strong> 编译器还不能很好的支持，如option java_package，option java_outer_classname，和嵌套类。这里我们基于 <strong>FlatBuffers</strong> 编译器转换的 .proto 文件来获得我们的 .fbs 文件：</p> <pre> <code class="language-java">// Generated from addressbook.proto namespace com.example.tutorial; enum PhoneType : int { MOBILE = 0, HOME = 1, WORK = 2, } namespace com.example.tutorial; table Person { name:string (required); id:int; email:string; phone:[com.example.tutorial._Person.PhoneNumber]; } namespace com.example.tutorial._Person; table PhoneNumber { number:string (required); type:int; } namespace com.example.tutorial; table AddressBook { person:[com.example.tutorial.Person]; } root_type AddressBook;</code></pre> <h2>编译 .fbs 文件</h2> <p>可以通过如下命令编译 .fbs 文件：</p> <pre> <code class="language-java">$ flatc --java -o out addressbook.fbs</code></pre> <p>--java用于指定编译的目标编程语言。-o 参数则用于指定输出文件的路径，如过没有提供则将当前目录用作输出目录。 <strong>FlatBuffers</strong> 编译器按照为不同的数据结构声明的namespace生成目录结构。对于上面的例子，会生成如下的这些文件：</p> <pre> <code class="language-java">$ find out p.p1 {margin: 0.0px 0.0px 0.0px 0.0px; font: 11.0px Menlo}span.s1 {font-variant-ligatures: no-common-ligatures} $ find out/ out/ out//com out//com/example out//com/example/tutorial out//com/example/tutorial/_Person out//com/example/tutorial/_Person/PhoneNumber.java out//com/example/tutorial/AddressBook.java out//com/example/tutorial/Person.java out//com/example/tutorial/PhoneType.java</code></pre> <h2>在Android项目中使用FlatBuffers</h2> <p>我们将前面由 .fbs 文件生成的Java文件拷贝到我们的项目中。我们前面提到的， <strong>FlatBuffers</strong> 的Java库比较薄，当前并没有发不到jcenter这样的maven仓库中，因而我们需要将这部分代码也拷贝到我们的额项目中。 <strong>FlatBuffers</strong> 的Java库在其repo仓库的 java 目录下。引入这些文件之后，我们的代码结构如下：</p> <p style="text-align:center"><img src="https://simg.open-open.com/show/e21068547bac05ca988f714e187613f4.jpg"></p> <p>添加访问 <strong>FlatBuffers</strong> 的类：</p> <pre> <code class="language-java">package com.netease.volleydemo; import com.example.tutorial.AddressBook; import com.example.tutorial.Person; import com.example.tutorial._Person.PhoneNumber; import com.google.flatbuffers.FlatBufferBuilder; import java.nio.ByteBuffer; /** * Created by hanpfei0306 on 16-12-5. */ public class AddressBookFlatBuffers { public static ByteBuffer encodeTest(String[] names) { FlatBufferBuilder builder = new FlatBufferBuilder(0); int[] personOffsets = new int[names.length]; for (int i = 0; i < names.length; ++ i) { int name = builder.createString(names[i]); int email = builder.createString("zhangsan@gmail.com"); int number1 = builder.createString("0157-23443276"); int type1 = 1; int phoneNumber1 = PhoneNumber.createPhoneNumber(builder, number1, type1); int number2 = builder.createString("136183667387"); int type2 = 0; int phoneNumber2 = PhoneNumber.createPhoneNumber(builder, number2, type2); int[] phoneNubers = new int[2]; phoneNubers[0] = phoneNumber1; phoneNubers[1] = phoneNumber2; int phoneNumbersPos = Person.createPhoneVector(builder, phoneNubers); int person = Person.createPerson(builder, name, 13958235, email, phoneNumbersPos); personOffsets[i] = person; } int persons = AddressBook.createPersonVector(builder, personOffsets); AddressBook.startAddressBook(builder); AddressBook.addPerson(builder, persons); int eab = AddressBook.endAddressBook(builder); builder.finish(eab); ByteBuffer buf = builder.dataBuffer(); return buf; } public static ByteBuffer encodeTest(String[] names, int times) { for (int i = 0; i < times - 1; ++ i) { encodeTest(names); } return encodeTest(names); } public static AddressBook decodeTest(ByteBuffer byteBuffer) { AddressBook addressBook = null; addressBook = AddressBook.getRootAsAddressBook(byteBuffer); return addressBook; } public static AddressBook decodeTest(ByteBuffer byteBuffer, int times) { AddressBook addressBook = null; for (int i = 0; i < times; ++ i) { addressBook = decodeTest(byteBuffer); } return addressBook; } }</code></pre> <h2>使用 flatbuf-gradle-plugin</h2> <p>我们有开发一个 <strong>FlatBuffers</strong> 的gradle插件，以方便开发 。这个插件的设计有参考Google的protobuf-gradle-plugin，功能与用法也与protobuf-gradle-plugin类似。在这个项目中，我们也有为 <strong>FlatBuffers</strong> 的Java库创建一个module。</p> <h2>编译并发布protobuf-gradle-plugin</h2> <p>从github上下载代码：</p> <pre> <code class="language-java">$ git clone https://github.com/hanpfei/flatbuffers.git</code></pre> <p>然后将代码导入Android Studio，将看到如下的代码结构：</p> <p style="text-align:center"><img src="https://simg.open-open.com/show/5cade82e96a09d223f859206c3127948.jpg"></p> <p>app 模块是一个demo程序，flatbuf-gradle-plugin 模块是 <strong>FlatBuffers</strong> 的gradle插件，而flatbuffers模块则是 <strong>FlatBuffers</strong> 的Java库。</p> <p>为了使用 flatbuf-gradle-plugin，可以将插件发布到本地文件系统。这可以通过修改flatbuf-gradle-plugin/build.gradle来完成，修改 uploadArchives task 的 repository 指向本地文件系统，如：</p> <pre> <code class="language-java">uploadArchives { repositories { mavenDeployer { pom.groupId = 'com.netease.hearttouch' pom.artifactId = 'ht-flatbuf-gradle-plugin' pom.version = '0.0.1-SNAPSHOT' repository(url: 'file:///Users/netease/Projects/CorpProjects/ht-flatbuffers/app/plugin') } } }</code></pre> <p>执行uploadArchives task，编译并发布flatbuf-gradle-plugin到本地文件系统。</p> <h2>应用flatbuf-gradle-plugin</h2> <p>修改应用程序的 build.gradle 以应用 flatbuf-gradle-plugin 。</p> <ol> <li>为buildscript添加对 flatbuf-gradle-plugin 的依赖： <pre> <code class="language-java">buildscript { //目前先发布在本地，后面会通过maven进行引用 repositories { maven { url "file:///Users/netease/Projects/CorpProjects/ht-flatbuffers/app/plugin" } jcenter() mavenCentral() } dependencies { classpath 'com.netease.hearttouch:ht-flatbuf-gradle-plugin:0.0.1-SNAPSHOT' } }</code></pre> </li> <li>在 apply plugin: 'com.android.application' 后面应用flatbuf的plugin： <pre> <code class="language-java">apply plugin: 'com.android.application' apply plugin: 'com.netease.flatbuf'</code></pre> </li> <li> <p>添加flatbuf块，对flatbuf-gradle-plugin的执行做配置：</p> <pre> <code class="language-java">flatbuf { flatc { path = '/usr/local/bin/flatc' } generateFlatTasks { all().each { task -> task.builtins { remove java } task.builtins { java { } } } } }</code></pre> <p>flatc 块用于配置 <strong>FlatBuffers</strong> 编译器，这里我们指定用我们之前手动编译的编译器。</p> <p>task.builtins 的块必不可少，这个块用于指定我们要为那些编程语言生成代码，这里我们为Java生成代码。</p> </li> <li>指定 .fbs 文件的路径 <pre> <code class="language-java">sourceSets { main { flat { srcDir 'src/main/flat' } } }</code></pre> 我们将 <strong>FlatBuffers</strong> 的IDL文件放在src/main/flat目录下。</li> </ol> <p>这样我们就不用再那么麻烦每次手动执行protoc了。</p> <h2>FlatBuffers、Protobuf及JSON对比测试</h2> <p>FlatBuffers相对于Protobuf的表现又如何呢？这里我们用数据说话，对比一下FlatBuffers格式、JSON格式与Protobuf的表现。测试同样用fastjson作为JSON的编码解码工具。</p> <p>测试用的数据结构所有的数据结构，Protobuf相关的测试代码，及JSON的测试代码同 在Android中使用Protocol Buffers 一文所述，FlatBuffers的测试代码如下：</p> <pre> <code class="language-java">package hearttouch.netease.com.myapplication; import com.example.tutorial.AddressBook; import com.example.tutorial.Person; import com.example.tutorial._Person.PhoneNumber; import com.google.flatbuffers.FlatBufferBuilder; import java.nio.ByteBuffer; /** * Created by hanpfei0306 on 16-12-5. */ public class AddressBookFlatBuffers { public static ByteBuffer encodeTest(String[] names) { FlatBufferBuilder builder = new FlatBufferBuilder(0); int[] personOffsets = new int[names.length]; for (int i = 0; i < names.length; ++ i) { int name = builder.createString(names[i]); int email = builder.createString("zhangsan@gmail.com"); int number1 = builder.createString("0157-23443276"); int type1 = 1; int phoneNumber1 = PhoneNumber.createPhoneNumber(builder, number1, type1); int number2 = builder.createString("136183667387"); int type2 = 0; int phoneNumber2 = PhoneNumber.createPhoneNumber(builder, number2, type2); int[] phoneNubers = new int[2]; phoneNubers[0] = phoneNumber1; phoneNubers[1] = phoneNumber2; int phoneNumbersPos = Person.createPhoneVector(builder, phoneNubers); int person = Person.createPerson(builder, name, 13958235, email, phoneNumbersPos); personOffsets[i] = person; } int persons = AddressBook.createPersonVector(builder, personOffsets); AddressBook.startAddressBook(builder); AddressBook.addPerson(builder, persons); int eab = AddressBook.endAddressBook(builder); builder.finish(eab); ByteBuffer buf = builder.dataBuffer(); return buf; } public static ByteBuffer encodeTest(String[] names, int times) { for (int i = 0; i < times - 1; ++ i) { encodeTest(names); } return encodeTest(names); } public static AddressBook decodeTest(ByteBuffer byteBuffer) { AddressBook addressBook = null; addressBook = AddressBook.getRootAsAddressBook(byteBuffer); return addressBook; } public static AddressBook decodeTest(ByteBuffer byteBuffer, int times) { AddressBook addressBook = null; for (int i = 0; i < times; ++ i) { addressBook = decodeTest(byteBuffer); } return addressBook; } }</code></pre> <p>通过如下的这段代码来执行测试：</p> <pre> <code class="language-java">private class ProtoTestTask extends AsyncTask<Void, Void, Void> { private static final int BUFFER_LEN = 8192; private void compress(InputStream is, OutputStream os) throws Exception { GZIPOutputStream gos = new GZIPOutputStream(os); int count; byte data[] = new byte[BUFFER_LEN]; while ((count = is.read(data, 0, BUFFER_LEN)) != -1) { gos.write(data, 0, count); } gos.finish(); gos.close(); } private int getCompressedDataLength(byte[] data) { ByteArrayInputStream bais =new ByteArrayInputStream(data); ByteArrayOutputStream baos = new ByteArrayOutputStream(); try { compress(bais, baos); } catch (Exception e) { } return baos.toByteArray().length; } private void dumpDataLengthInfo(byte[] protobufData, String jsonData, ByteBuffer flatbufData) { int compressedProtobufLength = getCompressedDataLength(protobufData); int compressedJSONLength = getCompressedDataLength(jsonData.getBytes()); int compressedFlatbufLength = getCompressedDataLength(flatbufData.array()); Log.i(TAG, String.format("%-120s", "Data length")); Log.i(TAG, String.format("%-20s%-20s%-20s%-20s%-20s%-20s", "Protobuf", "Protobuf (GZIP)", "JSON", "JSON (GZIP)", "Flatbuf", "Flatbuf (GZIP)")); Log.i(TAG, String.format("%-20s%-20s%-20s%-20s%-20s%-20s", String.valueOf(protobufData.length), compressedProtobufLength, String.valueOf(jsonData.getBytes().length), compressedJSONLength, String.valueOf(flatbufData.array().length), compressedFlatbufLength)); } private void doEncodeTest(String[] names, int times) { long startTime = System.nanoTime(); byte[] protobufData = AddressBookProtobuf.encodeTest(names, times); long protobufTime = System.nanoTime(); protobufTime = protobufTime - startTime; startTime = System.nanoTime(); String jsonData = AddressBookJson.encodeTest(names, times); long jsonTime = System.nanoTime(); jsonTime = jsonTime - startTime; startTime = System.nanoTime(); ByteBuffer flatbufData = AddressBookFlatBuffers.encodeTest(names, times); long flatbufTime = System.nanoTime(); flatbufTime = flatbufTime - startTime; dumpDataLengthInfo(protobufData, jsonData, flatbufData); Log.i(TAG, String.format("%-20s%-20s%-20s%-20s", "Encode Times", String.valueOf(times), "Names Length", String.valueOf(names.length))); Log.i(TAG, String.format("%-20s%-20s%-20s%-20s%-20s%-20s", "ProtobufTime", String.valueOf(protobufTime), "JsonTime", String.valueOf(jsonTime), "FlatbufTime", String.valueOf(flatbufTime))); } private void doEncodeTest10(int times) { doEncodeTest(TestUtils.sTestNames10, times); } private void doEncodeTest50(int times) { doEncodeTest(TestUtils.sTestNames50, times); } private void doEncodeTest100(int times) { doEncodeTest(TestUtils.sTestNames100, times); } private void doEncodeTest(int times) { doEncodeTest10(times); doEncodeTest50(times); doEncodeTest100(times); } private void doDecodeTest(String[] names, int times) { byte[] protobufBytes = AddressBookProtobuf.encodeTest(names); ByteArrayInputStream bais = new ByteArrayInputStream(protobufBytes); long startTime = System.nanoTime(); AddressBookProtobuf.decodeTest(bais, times); long protobufTime = System.nanoTime(); protobufTime = protobufTime - startTime; String jsonStr = AddressBookJson.encodeTest(names); startTime = System.nanoTime(); AddressBookJson.decodeTest(jsonStr, times); long jsonTime = System.nanoTime(); jsonTime = jsonTime - startTime; ByteBuffer flatbufData = AddressBookFlatBuffers.encodeTest(names); startTime = System.nanoTime(); AddressBookFlatBuffers.decodeTest(flatbufData, times); long flatbufTime = System.nanoTime(); flatbufTime = flatbufTime - startTime; Log.i(TAG, String.format("%-20s%-20s%-20s%-20s", "Decode Times", String.valueOf(times), "Names Length", String.valueOf(names.length))); Log.i(TAG, String.format("%-20s%-20s%-20s%-20s%-20s%-20s", "ProtobufTime", String.valueOf(protobufTime), "JsonTime", String.valueOf(jsonTime), "FlatbufTime", String.valueOf(flatbufTime))); } private void doDecodeTest10(int times) { doDecodeTest(TestUtils.sTestNames10, times); } private void doDecodeTest50(int times) { doDecodeTest(TestUtils.sTestNames50, times); } private void doDecodeTest100(int times) { doDecodeTest(TestUtils.sTestNames100, times); } private void doDecodeTest(int times) { doDecodeTest10(times); doDecodeTest50(times); doDecodeTest100(times); } @Override protected Void doInBackground(Void... params) { TestUtils.initTest(); doEncodeTest(5000); doDecodeTest(5000); return null; } @Override protected void onPostExecute(Void aVoid) { super.onPostExecute(aVoid); } }</code></pre> <p>这里我们执行3组编码测试及3组解码测试。对于编码测试，第一组的单个数据中包含10个Person，第二组的包含50个，第三组的包含100个，然后对每个数据分别执行5000次的编码操作。</p> <p>对于解码测试，三组中单个数据同样包含10个Person、50个及100个，然后对每个数据分别执行5000次的解码码操作。</p> <p>在Galaxy Nexus的Android 4.4.4 CM平台上执行上述测试，最终得到如下结果：</p> <p>编码后数据长度对比 (Bytes)</p> <table> <thead> <tr> <th>Person个数</th> <th>Protobuf</th> <th>Protobuf（GZIP）</th> <th>JSON</th> <th>JSON（GZIP）</th> <th>Flatbuf</th> <th>Flatbuf（GZIP)</th> </tr> </thead> <tbody> <tr> <td>10</td> <td>860</td> <td>290</td> <td>1703</td> <td>343</td> <td>2048</td> <td>521</td> </tr> <tr> <td>50</td> <td>4300</td> <td>978</td> <td>8463</td> <td>1043</td> <td>8192</td> <td>1822</td> </tr> <tr> <td>100</td> <td>8600</td> <td>1825</td> <td>16913</td> <td>1902</td> <td>16384</td> <td>3417</td> </tr> </tbody> </table> <p>相同的数据，经过编码，在压缩前FlatBuffers的数据长度和JSON的数据长度接近，而Protobuf的数据长度则只有前两者的大概一半。而在用GZIP压缩后，Protobuf的数据长度与JSON的接近，而FlatBuffers的数据长度则接近两者的两倍。</p> <p>编码性能对比 (S)</p> <table> <thead> <tr> <th>Person个数</th> <th>Protobuf</th> <th>JSON</th> <th>FlatBuffers</th> </tr> </thead> <tbody> <tr> <td>10</td> <td>6.000</td> <td>8.952</td> <td>12.464</td> </tr> <tr> <td>50</td> <td>26.847</td> <td>45.782</td> <td>56.752</td> </tr> <tr> <td>100</td> <td>50.602</td> <td>73.688</td> <td>108.426</td> </tr> </tbody> </table> <p>编码性能Protobuf相对于JSON有较大幅度的提高，而FlatBuffers则有较大幅度的降低。</p> <p>解码性能对比 (S)</p> <table> <thead> <tr> <th>Person个数</th> <th>Protobuf</th> <th>JSON</th> <th>FlatBuffers</th> </tr> </thead> <tbody> <tr> <td>10</td> <td>0.255</td> <td>10.766</td> <td>0.014</td> </tr> <tr> <td>50</td> <td>0.245</td> <td>51.134</td> <td>0.014</td> </tr> <tr> <td>100</td> <td>0.323</td> <td>101.070</td> <td>0.006</td> </tr> </tbody> </table> <p>解码性能方面，Protobuf相对于JSON，有着惊人的提升。Protobuf的解码时间几乎不随着数据长度的增长而有太大的增长，而JSON则随着数据长度的增加，解码所需要的时间也越来越长。而FlatBuffers则由于无需解码，在性能方面相对于前两者更有着非常大的提升。</p> <p> </p> <p> </p> <p>来自：http://www.jianshu.com/p/3bac6bc80db7</p> <p> </p>