| |
| |
| |
| <!DOCTYPE html> |
| <html class="writer-html5" lang="en" > |
| <head> |
| <meta charset="utf-8"> |
| |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> |
| |
| <title>Streaming, Serialization, and IPC — Apache Arrow v1.0.1</title> |
| |
| |
| |
| <link rel="stylesheet" href="../_static/css/theme.css" type="text/css" /> |
| <link rel="stylesheet" href="../_static/pygments.css" type="text/css" /> |
| |
| |
| |
| |
| |
| |
| |
| <!--[if lt IE 9]> |
| <script src="../_static/js/html5shiv.min.js"></script> |
| <![endif]--> |
| |
| |
| <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script> |
| <script src="../_static/jquery.js"></script> |
| <script src="../_static/underscore.js"></script> |
| <script src="../_static/doctools.js"></script> |
| <script src="../_static/language_data.js"></script> |
| |
| <script type="text/javascript" src="../_static/js/theme.js"></script> |
| |
| |
| <link rel="canonical" href="https://arrow.apache.org/docs/python/ipc.html" /> |
| <link rel="index" title="Index" href="../genindex.html" /> |
| <link rel="search" title="Search" href="../search.html" /> |
| <link rel="next" title="Filesystem Interface" href="filesystems.html" /> |
| <link rel="prev" title="Data Types and In-Memory Data Model" href="data.html" /> |
|
|
|
|
| <!-- Matomo -->
|
| <script>
|
| var _paq = window._paq = window._paq || [];
|
| /* tracker methods like "setCustomDimension" should be called before "trackPageView" */
|
| _paq.push(["setDoNotTrack", true]);
|
| _paq.push(["disableCookies"]);
|
| _paq.push(['trackPageView']);
|
| _paq.push(['enableLinkTracking']);
|
| (function() {
|
| var u="https://analytics.apache.org/";
|
| _paq.push(['setTrackerUrl', u+'matomo.php']);
|
| _paq.push(['setSiteId', '20']);
|
| var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0];
|
| g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s);
|
| })();
|
| </script>
|
| <!-- End Matomo Code -->
|
|
|
| </head> |
| |
| <body class="wy-body-for-nav"> |
| |
| |
| <div class="wy-grid-for-nav"> |
| |
| <nav data-toggle="wy-nav-shift" class="wy-nav-side"> |
| <div class="wy-side-scroll"> |
| <div class="wy-side-nav-search" > |
| |
| |
| |
| <a href="../index.html" class="icon icon-home" alt="Documentation Home"> Apache Arrow |
| |
| |
| |
| </a> |
| |
| |
| |
| |
| <div class="version"> |
| 1.0.1 |
| </div> |
| |
| |
| |
| |
| <div role="search"> |
| <form id="rtd-search-form" class="wy-form" action="../search.html" method="get"> |
| <input type="text" name="q" placeholder="Search docs" /> |
| <input type="hidden" name="check_keywords" value="yes" /> |
| <input type="hidden" name="area" value="default" /> |
| </form> |
| </div> |
| |
| |
| </div> |
| |
| |
| <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation"> |
| |
| |
| |
| |
| |
| |
| <p class="caption"><span class="caption-text">Specifications and Protocols</span></p> |
| <ul> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Versioning.html">Format Versioning and Stability</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Columnar.html">Arrow Columnar Format</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Flight.html">Arrow Flight RPC</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Integration.html">Integration Testing</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/CDataInterface.html">The Arrow C data interface</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Other.html">Other Data Structures</a></li> |
| </ul> |
| <p class="caption"><span class="caption-text">Libraries</span></p> |
| <ul class="current"> |
| <li class="toctree-l1"><a class="reference internal" href="../status.html">Implementation Status</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/c_glib/">C/GLib</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../cpp/index.html">C++</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/csharp/README.md">C#</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://godoc.org/github.com/apache/arrow/go/arrow">Go</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../java/index.html">Java</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/js/">JavaScript</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/matlab/README.md">MATLAB</a></li> |
| <li class="toctree-l1 current"><a class="reference internal" href="index.html">Python</a><ul class="current"> |
| <li class="toctree-l2"><a class="reference internal" href="install.html">Installing PyArrow</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="memory.html">Memory and IO Interfaces</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="data.html">Data Types and In-Memory Data Model</a></li> |
| <li class="toctree-l2 current"><a class="current reference internal" href="#">Streaming, Serialization, and IPC</a><ul> |
| <li class="toctree-l3"><a class="reference internal" href="#writing-and-reading-streams">Writing and Reading Streams</a><ul> |
| <li class="toctree-l4"><a class="reference internal" href="#using-streams">Using streams</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#writing-and-reading-random-access-files">Writing and Reading Random Access Files</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#reading-from-stream-and-file-format-for-pandas">Reading from Stream and File Format for pandas</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l3"><a class="reference internal" href="#arbitrary-object-serialization">Arbitrary Object Serialization</a><ul> |
| <li class="toctree-l4"><a class="reference internal" href="#serializing-custom-data-types">Serializing Custom Data Types</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#component-based-serialization">Component-based Serialization</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l3"><a class="reference internal" href="#serializing-pandas-objects">Serializing pandas Objects</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l2"><a class="reference internal" href="filesystems.html">Filesystem Interface</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="filesystems_deprecated.html">Filesystem Interface (legacy)</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="plasma.html">The Plasma In-Memory Object Store</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="numpy.html">NumPy Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="pandas.html">Pandas Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="timestamps.html">Timestamps</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="csv.html">Reading CSV files</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="feather.html">Feather File Format</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="json.html">Reading JSON files</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="parquet.html">Reading and Writing the Apache Parquet Format</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="dataset.html">Tabular Datasets</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="cuda.html">CUDA Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="extending_types.html">Extending pyarrow</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="extending.html">Using pyarrow from C++ and Cython Code</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="api.html">API Reference</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="getting_involved.html">Getting Involved</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="benchmarks.html">Benchmarks</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/r/">R</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/ruby/README.md">Ruby</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://docs.rs/crate/arrow/">Rust</a></li> |
| </ul> |
| <p class="caption"><span class="caption-text">Development</span></p> |
| <ul> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/contributing.html">Contributing to Apache Arrow</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/cpp/index.html">C++ Development</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/python.html">Python Development</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/archery.html">Daily Development using Archery</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/crossbow.html">Packaging and Testing with Crossbow</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/docker.html">Running Docker Builds</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/benchmarks.html">Benchmarks</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/documentation.html">Building the Documentation</a></li> |
| </ul> |
| |
| |
| |
| </div> |
| |
| </div> |
| </nav> |
| |
| <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"> |
| |
| |
| <nav class="wy-nav-top" aria-label="top navigation"> |
| |
| <i data-toggle="wy-nav-top" class="fa fa-bars"></i> |
| <a href="../index.html">Apache Arrow</a> |
| |
| </nav> |
| |
| |
| <div class="wy-nav-content"> |
| |
| <div class="rst-content"> |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| <div role="navigation" aria-label="breadcrumbs navigation"> |
| |
| <ul class="wy-breadcrumbs"> |
| |
| <li><a href="../index.html" class="icon icon-home"></a> »</li> |
| |
| <li><a href="index.html">Python bindings</a> »</li> |
| |
| <li>Streaming, Serialization, and IPC</li> |
| |
| |
| <li class="wy-breadcrumbs-aside"> |
| |
| |
| <a href="../_sources/python/ipc.rst.txt" rel="nofollow"> View page source</a> |
| |
| |
| </li> |
| |
| </ul> |
| |
| |
| <hr/> |
| </div> |
| <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article"> |
| <div itemprop="articleBody"> |
| |
| <div class="section" id="streaming-serialization-and-ipc"> |
| <span id="ipc"></span><h1>Streaming, Serialization, and IPC<a class="headerlink" href="#streaming-serialization-and-ipc" title="Permalink to this headline">¶</a></h1> |
| <div class="section" id="writing-and-reading-streams"> |
| <h2>Writing and Reading Streams<a class="headerlink" href="#writing-and-reading-streams" title="Permalink to this headline">¶</a></h2> |
| <p>Arrow defines two types of binary formats for serializing record batches:</p> |
| <ul class="simple"> |
| <li><p><strong>Streaming format</strong>: for sending an arbitrary length sequence of record |
| batches. The format must be processed from start to end, and does not support |
| random access</p></li> |
| <li><p><strong>File or Random Access format</strong>: for serializing a fixed number of record |
| batches. Supports random access, and thus is very useful when used with |
| memory maps</p></li> |
| </ul> |
| <p>To follow this section, make sure to first read the section on <a class="reference internal" href="memory.html#io"><span class="std std-ref">Memory and |
| IO</span></a>.</p> |
| <div class="section" id="using-streams"> |
| <h3>Using streams<a class="headerlink" href="#using-streams" title="Permalink to this headline">¶</a></h3> |
| <p>First, let’s create a small record batch:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [1]: </span><span class="kn">import</span> <span class="nn">pyarrow</span> <span class="kn">as</span> <span class="nn">pa</span> |
| |
| <span class="gp">In [2]: </span><span class="n">data</span> <span class="o">=</span> <span class="p">[</span> |
| <span class="gp"> ...: </span> <span class="n">pa</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">]),</span> |
| <span class="gp"> ...: </span> <span class="n">pa</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'baz'</span><span class="p">,</span> <span class="bp">None</span><span class="p">]),</span> |
| <span class="gp"> ...: </span> <span class="n">pa</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="bp">True</span><span class="p">,</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">False</span><span class="p">,</span> <span class="bp">True</span><span class="p">])</span> |
| <span class="gp"> ...: </span><span class="p">]</span> |
| <span class="gp"> ...: </span> |
| |
| <span class="gp">In [3]: </span><span class="n">batch</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">record_batch</span><span class="p">(</span><span class="n">data</span><span class="p">,</span> <span class="n">names</span><span class="o">=</span><span class="p">[</span><span class="s1">'f0'</span><span class="p">,</span> <span class="s1">'f1'</span><span class="p">,</span> <span class="s1">'f2'</span><span class="p">])</span> |
| |
| <span class="gp">In [4]: </span><span class="n">batch</span><span class="o">.</span><span class="n">num_rows</span> |
| <span class="gh">Out[4]: </span><span class="go">4</span> |
| |
| <span class="gp">In [5]: </span><span class="n">batch</span><span class="o">.</span><span class="n">num_columns</span> |
| <span class="gh">Out[5]: </span><span class="go">3</span> |
| </pre></div> |
| </div> |
| <p>Now, we can begin writing a stream containing some number of these batches. For |
| this we use <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchStreamWriter</span></code>, which can write to a |
| writeable <code class="docutils literal notranslate"><span class="pre">NativeFile</span></code> object or a writeable Python object. For convenience, |
| this one can be created with <a class="reference internal" href="generated/pyarrow.ipc.new_stream.html#pyarrow.ipc.new_stream" title="pyarrow.ipc.new_stream"><code class="xref py py-func docutils literal notranslate"><span class="pre">new_stream()</span></code></a>:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [6]: </span><span class="n">sink</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">BufferOutputStream</span><span class="p">()</span> |
| |
| <span class="gp">In [7]: </span><span class="n">writer</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">new_stream</span><span class="p">(</span><span class="n">sink</span><span class="p">,</span> <span class="n">batch</span><span class="o">.</span><span class="n">schema</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>Here we used an in-memory Arrow buffer stream, but this could have been a |
| socket or some other IO sink.</p> |
| <p>When creating the <code class="docutils literal notranslate"><span class="pre">StreamWriter</span></code>, we pass the schema, since the schema |
| (column names and types) must be the same for all of the batches sent in this |
| particular stream. Now we can do:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [8]: </span><span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">):</span> |
| <span class="gp"> ...: </span> <span class="n">writer</span><span class="o">.</span><span class="n">write_batch</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span> |
| <span class="gp"> ...: </span> |
| |
| <span class="gp">In [9]: </span><span class="n">writer</span><span class="o">.</span><span class="n">close</span><span class="p">()</span> |
| |
| <span class="gp">In [10]: </span><span class="n">buf</span> <span class="o">=</span> <span class="n">sink</span><span class="o">.</span><span class="n">getvalue</span><span class="p">()</span> |
| |
| <span class="gp">In [11]: </span><span class="n">buf</span><span class="o">.</span><span class="n">size</span> |
| <span class="gh">Out[11]: </span><span class="go">1984</span> |
| </pre></div> |
| </div> |
| <p>Now <code class="docutils literal notranslate"><span class="pre">buf</span></code> contains the complete stream as an in-memory byte buffer. We can |
| read such a stream with <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchStreamReader</span></code> or the |
| convenience function <code class="docutils literal notranslate"><span class="pre">pyarrow.ipc.open_stream</span></code>:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [12]: </span><span class="n">reader</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">open_stream</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| |
| <span class="gp">In [13]: </span><span class="n">reader</span><span class="o">.</span><span class="n">schema</span> |
| <span class="gh">Out[13]: </span><span class="go"></span> |
| <span class="go">f0: int64</span> |
| <span class="go">f1: string</span> |
| <span class="go">f2: bool</span> |
| |
| <span class="gp">In [14]: </span><span class="n">batches</span> <span class="o">=</span> <span class="p">[</span><span class="n">b</span> <span class="k">for</span> <span class="n">b</span> <span class="ow">in</span> <span class="n">reader</span><span class="p">]</span> |
| |
| <span class="gp">In [15]: </span><span class="nb">len</span><span class="p">(</span><span class="n">batches</span><span class="p">)</span> |
| <span class="gh">Out[15]: </span><span class="go">5</span> |
| </pre></div> |
| </div> |
| <p>We can check the returned batches are the same as the original input:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [16]: </span><span class="n">batches</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">equals</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span> |
| <span class="gh">Out[16]: </span><span class="go">True</span> |
| </pre></div> |
| </div> |
| <p>An important point is that if the input source supports zero-copy reads |
| (e.g. like a memory map, or <code class="docutils literal notranslate"><span class="pre">pyarrow.BufferReader</span></code>), then the returned |
| batches are also zero-copy and do not allocate any new memory on read.</p> |
| </div> |
| <div class="section" id="writing-and-reading-random-access-files"> |
| <h3>Writing and Reading Random Access Files<a class="headerlink" href="#writing-and-reading-random-access-files" title="Permalink to this headline">¶</a></h3> |
| <p>The <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchFileWriter</span></code> has the same API as |
| <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchStreamWriter</span></code>. You can create one with |
| <a class="reference internal" href="generated/pyarrow.ipc.new_file.html#pyarrow.ipc.new_file" title="pyarrow.ipc.new_file"><code class="xref py py-func docutils literal notranslate"><span class="pre">new_file()</span></code></a>:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [17]: </span><span class="n">sink</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">BufferOutputStream</span><span class="p">()</span> |
| |
| <span class="gp">In [18]: </span><span class="n">writer</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">new_file</span><span class="p">(</span><span class="n">sink</span><span class="p">,</span> <span class="n">batch</span><span class="o">.</span><span class="n">schema</span><span class="p">)</span> |
| |
| <span class="gp">In [19]: </span><span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">10</span><span class="p">):</span> |
| <span class="gp"> ....: </span> <span class="n">writer</span><span class="o">.</span><span class="n">write_batch</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span> |
| <span class="gp"> ....: </span> |
| |
| <span class="gp">In [20]: </span><span class="n">writer</span><span class="o">.</span><span class="n">close</span><span class="p">()</span> |
| |
| <span class="gp">In [21]: </span><span class="n">buf</span> <span class="o">=</span> <span class="n">sink</span><span class="o">.</span><span class="n">getvalue</span><span class="p">()</span> |
| |
| <span class="gp">In [22]: </span><span class="n">buf</span><span class="o">.</span><span class="n">size</span> |
| <span class="gh">Out[22]: </span><span class="go">4226</span> |
| </pre></div> |
| </div> |
| <p>The difference between <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchFileReader</span></code> and |
| <code class="xref py py-class docutils literal notranslate"><span class="pre">RecordBatchStreamReader</span></code> is that the input source must have a |
| <code class="docutils literal notranslate"><span class="pre">seek</span></code> method for random access. The stream reader only requires read |
| operations. We can also use the <a class="reference internal" href="generated/pyarrow.ipc.open_file.html#pyarrow.ipc.open_file" title="pyarrow.ipc.open_file"><code class="xref py py-func docutils literal notranslate"><span class="pre">open_file()</span></code></a> method to open a file:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [23]: </span><span class="n">reader</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">open_file</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>Because we have access to the entire payload, we know the number of record |
| batches in the file, and can read any at random:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [24]: </span><span class="n">reader</span><span class="o">.</span><span class="n">num_record_batches</span> |
| <span class="gh">Out[24]: </span><span class="go">10</span> |
| |
| <span class="gp">In [25]: </span><span class="n">b</span> <span class="o">=</span> <span class="n">reader</span><span class="o">.</span><span class="n">get_batch</span><span class="p">(</span><span class="mi">3</span><span class="p">)</span> |
| |
| <span class="gp">In [26]: </span><span class="n">b</span><span class="o">.</span><span class="n">equals</span><span class="p">(</span><span class="n">batch</span><span class="p">)</span> |
| <span class="gh">Out[26]: </span><span class="go">True</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="reading-from-stream-and-file-format-for-pandas"> |
| <h3>Reading from Stream and File Format for pandas<a class="headerlink" href="#reading-from-stream-and-file-format-for-pandas" title="Permalink to this headline">¶</a></h3> |
| <p>The stream and file reader classes have a special <code class="docutils literal notranslate"><span class="pre">read_pandas</span></code> method to |
| simplify reading multiple record batches and converting them to a single |
| DataFrame output:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [27]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">open_file</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span><span class="o">.</span><span class="n">read_pandas</span><span class="p">()</span> |
| |
| <span class="gp">In [28]: </span><span class="n">df</span><span class="p">[:</span><span class="mi">5</span><span class="p">]</span> |
| <span class="gh">Out[28]: </span><span class="go"></span> |
| <span class="go"> f0 f1 f2</span> |
| <span class="go">0 1 foo True</span> |
| <span class="go">1 2 bar None</span> |
| <span class="go">2 3 baz False</span> |
| <span class="go">3 4 None True</span> |
| <span class="go">4 1 foo True</span> |
| </pre></div> |
| </div> |
| </div> |
| </div> |
| <div class="section" id="arbitrary-object-serialization"> |
| <h2>Arbitrary Object Serialization<a class="headerlink" href="#arbitrary-object-serialization" title="Permalink to this headline">¶</a></h2> |
| <p>In <code class="docutils literal notranslate"><span class="pre">pyarrow</span></code> we are able to serialize and deserialize many kinds of Python |
| objects. While not a complete replacement for the <code class="docutils literal notranslate"><span class="pre">pickle</span></code> module, these |
| functions can be significantly faster, particular when dealing with collections |
| of NumPy arrays.</p> |
| <div class="admonition warning"> |
| <p class="admonition-title">Warning</p> |
| <p>While the functions in this section utilize the Arrow stream protocol |
| internally, they do not produce data that is compatible with the above |
| <code class="docutils literal notranslate"><span class="pre">ipc.open_file</span></code> and <code class="docutils literal notranslate"><span class="pre">ipc.open_stream</span></code> functions.</p> |
| </div> |
| <p>As an example, consider a dictionary containing NumPy arrays:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [29]: </span><span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span> |
| |
| <span class="gp">In [30]: </span><span class="n">data</span> <span class="o">=</span> <span class="p">{</span> |
| <span class="gp"> ....: </span> <span class="n">i</span><span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">500</span><span class="p">,</span> <span class="mi">500</span><span class="p">)</span> |
| <span class="gp"> ....: </span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">100</span><span class="p">)</span> |
| <span class="gp"> ....: </span><span class="p">}</span> |
| <span class="gp"> ....: </span> |
| </pre></div> |
| </div> |
| <p>We use the <code class="docutils literal notranslate"><span class="pre">pyarrow.serialize</span></code> function to convert this data to a byte |
| buffer:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [31]: </span><span class="n">buf</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">serialize</span><span class="p">(</span><span class="n">data</span><span class="p">)</span><span class="o">.</span><span class="n">to_buffer</span><span class="p">()</span> |
| |
| <span class="gp">In [32]: </span><span class="nb">type</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| <span class="gh">Out[32]: </span><span class="go">pyarrow.lib.Buffer</span> |
| |
| <span class="gp">In [33]: </span><span class="n">buf</span><span class="o">.</span><span class="n">size</span> |
| <span class="gh">Out[33]: </span><span class="go">200028928</span> |
| </pre></div> |
| </div> |
| <p><code class="docutils literal notranslate"><span class="pre">pyarrow.serialize</span></code> creates an intermediate object which can be converted to |
| a buffer (the <code class="docutils literal notranslate"><span class="pre">to_buffer</span></code> method) or written directly to an output stream.</p> |
| <p><code class="docutils literal notranslate"><span class="pre">pyarrow.deserialize</span></code> converts a buffer-like object back to the original |
| Python object:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [34]: </span><span class="n">restored_data</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">deserialize</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| |
| <span class="gp">In [35]: </span><span class="n">restored_data</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span> |
| <span class="gh">Out[35]: </span><span class="go"></span> |
| <span class="go">array([[ 0.31908496, -0.38497429, -0.00908503, ..., 0.59388712,</span> |
| <span class="go"> -0.4781306 , 0.05585997],</span> |
| <span class="go"> [ 0.17211542, 0.36950389, 1.65803031, ..., 0.13289332,</span> |
| <span class="go"> 0.81641384, -1.10876572],</span> |
| <span class="go"> [-0.28750533, 0.02887317, -1.16949495, ..., 0.68167732,</span> |
| <span class="go"> -0.94371386, 0.15949882],</span> |
| <span class="go"> ...,</span> |
| <span class="go"> [ 1.40029811, 0.4470198 , 1.55334769, ..., 1.96692071,</span> |
| <span class="go"> -0.2367957 , -1.20272686],</span> |
| <span class="go"> [-2.01700399, -1.10887375, 1.50930246, ..., 1.75533091,</span> |
| <span class="go"> 0.86458875, 0.0209724 ],</span> |
| <span class="go"> [-1.01071012, 0.7951678 , 1.43853346, ..., -0.84843836,</span> |
| <span class="go"> -1.50469422, -0.06000656]])</span> |
| </pre></div> |
| </div> |
| <p>When dealing with NumPy arrays, <code class="docutils literal notranslate"><span class="pre">pyarrow.deserialize</span></code> can be significantly |
| faster than <code class="docutils literal notranslate"><span class="pre">pickle</span></code> because the resulting arrays are zero-copy references |
| into the input buffer. The larger the arrays, the larger the performance |
| savings.</p> |
| <p>Consider this example, we have for <code class="docutils literal notranslate"><span class="pre">pyarrow.deserialize</span></code></p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [36]: </span><span class="o">%</span><span class="k">timeit</span> restored_data = pa.deserialize(buf) |
| <span class="go">8.81 ms +- 10.1 us per loop (mean +- std. dev. of 7 runs, 100 loops each)</span> |
| </pre></div> |
| </div> |
| <p>And for pickle:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [37]: </span><span class="kn">import</span> <span class="nn">pickle</span> |
| |
| <span class="gp">In [38]: </span><span class="n">pickled</span> <span class="o">=</span> <span class="n">pickle</span><span class="o">.</span><span class="n">dumps</span><span class="p">(</span><span class="n">data</span><span class="p">)</span> |
| |
| <span class="gp">In [39]: </span><span class="o">%</span><span class="k">timeit</span> unpickled_data = pickle.loads(pickled) |
| <span class="go">74.7 ms +- 67.9 us per loop (mean +- std. dev. of 7 runs, 10 loops each)</span> |
| </pre></div> |
| </div> |
| <p>We aspire to make these functions a high-speed alternative to pickle for |
| transient serialization in Python big data applications.</p> |
| <div class="section" id="serializing-custom-data-types"> |
| <h3>Serializing Custom Data Types<a class="headerlink" href="#serializing-custom-data-types" title="Permalink to this headline">¶</a></h3> |
| <p>If an unrecognized data type is encountered when serializing an object, |
| <code class="docutils literal notranslate"><span class="pre">pyarrow</span></code> will fall back on using <code class="docutils literal notranslate"><span class="pre">pickle</span></code> for converting that type to a |
| byte string. There may be a more efficient way, though.</p> |
| <p>Consider a class with two members, one of which is a NumPy array:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">MyData</span><span class="p">:</span> |
| <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">name</span><span class="p">,</span> <span class="n">data</span><span class="p">):</span> |
| <span class="bp">self</span><span class="o">.</span><span class="n">name</span> <span class="o">=</span> <span class="n">name</span> |
| <span class="bp">self</span><span class="o">.</span><span class="n">data</span> <span class="o">=</span> <span class="n">data</span> |
| </pre></div> |
| </div> |
| <p>We write functions to convert this to and from a dictionary with simpler types:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">_serialize_MyData</span><span class="p">(</span><span class="n">val</span><span class="p">):</span> |
| <span class="k">return</span> <span class="p">{</span><span class="s1">'name'</span><span class="p">:</span> <span class="n">val</span><span class="o">.</span><span class="n">name</span><span class="p">,</span> <span class="s1">'data'</span><span class="p">:</span> <span class="n">val</span><span class="o">.</span><span class="n">data</span><span class="p">}</span> |
| |
| <span class="k">def</span> <span class="nf">_deserialize_MyData</span><span class="p">(</span><span class="n">data</span><span class="p">):</span> |
| <span class="k">return</span> <span class="n">MyData</span><span class="p">(</span><span class="n">data</span><span class="p">[</span><span class="s1">'name'</span><span class="p">],</span> <span class="n">data</span><span class="p">[</span><span class="s1">'data'</span><span class="p">]</span> |
| </pre></div> |
| </div> |
| <p>then, we must register these functions in a <code class="docutils literal notranslate"><span class="pre">SerializationContext</span></code> so that |
| <code class="docutils literal notranslate"><span class="pre">MyData</span></code> can be recognized:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">context</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">SerializationContext</span><span class="p">()</span> |
| <span class="n">context</span><span class="o">.</span><span class="n">register_type</span><span class="p">(</span><span class="n">MyData</span><span class="p">,</span> <span class="s1">'MyData'</span><span class="p">,</span> |
| <span class="n">custom_serializer</span><span class="o">=</span><span class="n">_serialize_MyData</span><span class="p">,</span> |
| <span class="n">custom_deserializer</span><span class="o">=</span><span class="n">_deserialize_MyData</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>Lastly, we use this context as an additional argument to <code class="docutils literal notranslate"><span class="pre">pyarrow.serialize</span></code>:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">buf</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">serialize</span><span class="p">(</span><span class="n">val</span><span class="p">,</span> <span class="n">context</span><span class="o">=</span><span class="n">context</span><span class="p">)</span><span class="o">.</span><span class="n">to_buffer</span><span class="p">()</span> |
| <span class="n">restored_val</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">deserialize</span><span class="p">(</span><span class="n">buf</span><span class="p">,</span> <span class="n">context</span><span class="o">=</span><span class="n">context</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>The <code class="docutils literal notranslate"><span class="pre">SerializationContext</span></code> also has convenience methods <code class="docutils literal notranslate"><span class="pre">serialize</span></code> and |
| <code class="docutils literal notranslate"><span class="pre">deserialize</span></code>, so these are equivalent statements:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">buf</span> <span class="o">=</span> <span class="n">context</span><span class="o">.</span><span class="n">serialize</span><span class="p">(</span><span class="n">val</span><span class="p">)</span><span class="o">.</span><span class="n">to_buffer</span><span class="p">()</span> |
| <span class="n">restored_val</span> <span class="o">=</span> <span class="n">context</span><span class="o">.</span><span class="n">deserialize</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="component-based-serialization"> |
| <h3>Component-based Serialization<a class="headerlink" href="#component-based-serialization" title="Permalink to this headline">¶</a></h3> |
| <p>For serializing Python objects containing some number of NumPy arrays, Arrow |
| buffers, or other data types, it may be desirable to transport their serialized |
| representation without having to produce an intermediate copy using the |
| <code class="docutils literal notranslate"><span class="pre">to_buffer</span></code> method. To motivate this, suppose we have a list of NumPy arrays:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [40]: </span><span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span> |
| |
| <span class="gp">In [41]: </span><span class="n">data</span> <span class="o">=</span> <span class="p">[</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">10</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">5</span><span class="p">)]</span> |
| </pre></div> |
| </div> |
| <p>The call <code class="docutils literal notranslate"><span class="pre">pa.serialize(data)</span></code> does not copy the memory inside each of these |
| NumPy arrays. This serialized representation can be then decomposed into a |
| dictionary containing a sequence of <code class="docutils literal notranslate"><span class="pre">pyarrow.Buffer</span></code> objects containing |
| metadata for each array and references to the memory inside the arrays. To do |
| this, use the <code class="docutils literal notranslate"><span class="pre">to_components</span></code> method:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [42]: </span><span class="n">serialized</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">serialize</span><span class="p">(</span><span class="n">data</span><span class="p">)</span> |
| |
| <span class="gp">In [43]: </span><span class="n">components</span> <span class="o">=</span> <span class="n">serialized</span><span class="o">.</span><span class="n">to_components</span><span class="p">()</span> |
| </pre></div> |
| </div> |
| <p>The particular details of the output of <code class="docutils literal notranslate"><span class="pre">to_components</span></code> are not too |
| important. The objects in the <code class="docutils literal notranslate"><span class="pre">'data'</span></code> field are <code class="docutils literal notranslate"><span class="pre">pyarrow.Buffer</span></code> objects, |
| which are zero-copy convertible to Python <code class="docutils literal notranslate"><span class="pre">memoryview</span></code> objects:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [44]: </span><span class="n">memoryview</span><span class="p">(</span><span class="n">components</span><span class="p">[</span><span class="s1">'data'</span><span class="p">][</span><span class="mi">0</span><span class="p">])</span> |
| <span class="gh">Out[44]: </span><span class="go"><memory at 0x7f2f699cda08></span> |
| </pre></div> |
| </div> |
| <p>A memoryview can be converted back to a Arrow <code class="docutils literal notranslate"><span class="pre">Buffer</span></code> with |
| <code class="docutils literal notranslate"><span class="pre">pyarrow.py_buffer</span></code>:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [45]: </span><span class="n">mv</span> <span class="o">=</span> <span class="n">memoryview</span><span class="p">(</span><span class="n">components</span><span class="p">[</span><span class="s1">'data'</span><span class="p">][</span><span class="mi">0</span><span class="p">])</span> |
| |
| <span class="gp">In [46]: </span><span class="n">buf</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">py_buffer</span><span class="p">(</span><span class="n">mv</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>An object can be reconstructed from its component-based representation using |
| <code class="docutils literal notranslate"><span class="pre">deserialize_components</span></code>:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [47]: </span><span class="n">restored_data</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">deserialize_components</span><span class="p">(</span><span class="n">components</span><span class="p">)</span> |
| |
| <span class="gp">In [48]: </span><span class="n">restored_data</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span> |
| <span class="gh">Out[48]: </span><span class="go"></span> |
| <span class="go">array([[ 0.08272937, -1.62435083, -0.05553385, -0.17009072, -1.42342936,</span> |
| <span class="go"> -2.14976376, 2.08674725, -1.13165575, 0.35196661, 0.30062211],</span> |
| <span class="go"> [-1.64855238, 2.15327101, 0.24985174, 0.73395161, 0.33972162,</span> |
| <span class="go"> -0.14478569, 1.10654142, 0.67399034, 0.16480965, -1.40790188],</span> |
| <span class="go"> [ 1.52062309, 0.85876101, -0.78834157, 1.50678204, 1.21548628,</span> |
| <span class="go"> -0.768725 , 1.29781535, -2.04542496, 0.08160128, 0.92438068],</span> |
| <span class="go"> [-0.27505366, -0.96845967, -0.73970018, -2.03152045, -0.25347227,</span> |
| <span class="go"> -0.09009048, -1.33786189, 0.47357543, -0.66451895, -0.7304969 ],</span> |
| <span class="go"> [-1.53565458, 0.17464885, 1.1882689 , 1.39913389, -2.55505597,</span> |
| <span class="go"> -0.99447036, -1.3056554 , 0.63831016, 0.44225673, 0.1876823 ],</span> |
| <span class="go"> [ 2.03372544, -0.87916257, -0.65075426, 0.2109289 , -0.82366568,</span> |
| <span class="go"> -0.4311721 , 0.01720903, 2.31279004, 0.44759669, 0.50426095],</span> |
| <span class="go"> [ 0.74985502, -0.87754096, -1.84156315, 0.96375487, 0.0179234 ,</span> |
| <span class="go"> -0.20506921, 0.48606781, 0.80249654, -0.82538998, 0.15040379],</span> |
| <span class="go"> [ 0.29645408, 0.33675047, -0.36390788, -2.28558262, -0.76621032,</span> |
| <span class="go"> -1.40421301, -1.55362434, 0.1556816 , -0.26782507, -0.65797254],</span> |
| <span class="go"> [-0.4841557 , 0.04571714, 0.3263099 , -0.07581112, -0.16210661,</span> |
| <span class="go"> -0.35836661, 0.88125115, 0.33318043, 0.72369647, 0.59428462],</span> |
| <span class="go"> [-0.03647698, 0.71456644, 0.09034961, -1.17933729, -1.03682108,</span> |
| <span class="go"> -0.65117059, 1.80962159, -0.6355785 , 0.44556374, -0.19257742]])</span> |
| </pre></div> |
| </div> |
| <p><code class="docutils literal notranslate"><span class="pre">deserialize_components</span></code> is also available as a method on |
| <code class="docutils literal notranslate"><span class="pre">SerializationContext</span></code> objects.</p> |
| </div> |
| </div> |
| <div class="section" id="serializing-pandas-objects"> |
| <h2>Serializing pandas Objects<a class="headerlink" href="#serializing-pandas-objects" title="Permalink to this headline">¶</a></h2> |
| <p>The default serialization context has optimized handling of pandas |
| objects like <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> and <code class="docutils literal notranslate"><span class="pre">Series</span></code>. Combined with component-based |
| serialization above, this enables zero-copy transport of pandas DataFrame |
| objects not containing any Python objects:</p> |
| <div class="highlight-ipython notranslate"><div class="highlight"><pre><span></span><span class="gp">In [49]: </span><span class="kn">import</span> <span class="nn">pandas</span> <span class="kn">as</span> <span class="nn">pd</span> |
| |
| <span class="gp">In [50]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'a'</span><span class="p">:</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">]})</span> |
| |
| <span class="gp">In [51]: </span><span class="n">context</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">default_serialization_context</span><span class="p">()</span> |
| |
| <span class="gp">In [52]: </span><span class="n">serialized_df</span> <span class="o">=</span> <span class="n">context</span><span class="o">.</span><span class="n">serialize</span><span class="p">(</span><span class="n">df</span><span class="p">)</span> |
| |
| <span class="gp">In [53]: </span><span class="n">df_components</span> <span class="o">=</span> <span class="n">serialized_df</span><span class="o">.</span><span class="n">to_components</span><span class="p">()</span> |
| |
| <span class="gp">In [54]: </span><span class="n">original_df</span> <span class="o">=</span> <span class="n">context</span><span class="o">.</span><span class="n">deserialize_components</span><span class="p">(</span><span class="n">df_components</span><span class="p">)</span> |
| |
| <span class="gp">In [55]: </span><span class="n">original_df</span> |
| <span class="gh">Out[55]: </span><span class="go"></span> |
| <span class="go"> a</span> |
| <span class="go">0 1</span> |
| <span class="go">1 2</span> |
| <span class="go">2 3</span> |
| <span class="go">3 4</span> |
| <span class="go">4 5</span> |
| </pre></div> |
| </div> |
| </div> |
| </div> |
| |
| |
| </div> |
| |
| </div> |
| <footer> |
| |
| <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation"> |
| |
| <a href="filesystems.html" class="btn btn-neutral float-right" title="Filesystem Interface" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right"></span></a> |
| |
| |
| <a href="data.html" class="btn btn-neutral float-left" title="Data Types and In-Memory Data Model" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left"></span> Previous</a> |
| |
| </div> |
| |
| |
| <hr/> |
| |
| <div role="contentinfo"> |
| <p> |
| |
| © Copyright 2016-2019 Apache Software Foundation |
| |
| </p> |
| </div> |
| |
| |
| |
| Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a |
| |
| <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> |
| |
| provided by <a href="https://readthedocs.org">Read the Docs</a>. |
| |
| </footer> |
| |
| </div> |
| </div> |
| |
| </section> |
| |
| </div> |
| |
| |
| <script type="text/javascript"> |
| jQuery(function () { |
| SphinxRtdTheme.Navigation.enable(true); |
| }); |
| </script> |
| |
| |
| |
| |
| |
| |
| |
| <script type="text/javascript" src="/docs/_static/versionwarning.js"></script></body> |
| </html> |
