1<html> 2<head> 3<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 4<title>Input Output</title> 5<link rel="stylesheet" href="../../multiprecision.css" type="text/css"> 6<meta name="generator" content="DocBook XSL Stylesheets V1.79.1"> 7<link rel="home" href="../../index.html" title="Chapter 1. Boost.Multiprecision"> 8<link rel="up" href="../tut.html" title="Tutorial"> 9<link rel="prev" href="limits/how_to_tell.html" title="How to Determine the Kind of a Number From std::numeric_limits"> 10<link rel="next" href="hash.html" title="Hash Function Support"> 11</head> 12<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"> 13<table cellpadding="2" width="100%"><tr> 14<td valign="top"><img alt="Boost C++ Libraries" width="277" height="86" src="../../../../../../boost.png"></td> 15<td align="center"><a href="../../../../../../index.html">Home</a></td> 16<td align="center"><a href="../../../../../../libs/libraries.htm">Libraries</a></td> 17<td align="center"><a href="http://www.boost.org/users/people.html">People</a></td> 18<td align="center"><a href="http://www.boost.org/users/faq.html">FAQ</a></td> 19<td align="center"><a href="../../../../../../more/index.htm">More</a></td> 20</tr></table> 21<hr> 22<div class="spirit-nav"> 23<a accesskey="p" href="limits/how_to_tell.html"><img src="../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../tut.html"><img src="../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../index.html"><img src="../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="hash.html"><img src="../../../../../../doc/src/images/next.png" alt="Next"></a> 24</div> 25<div class="section"> 26<div class="titlepage"><div><div><h3 class="title"> 27<a name="boost_multiprecision.tut.input_output"></a><a class="link" href="input_output.html" title="Input Output">Input Output</a> 28</h3></div></div></div> 29<h5> 30<a name="boost_multiprecision.tut.input_output.h0"></a> 31 <span class="phrase"><a name="boost_multiprecision.tut.input_output.loopback_testing"></a></span><a class="link" href="input_output.html#boost_multiprecision.tut.input_output.loopback_testing">Loopback 32 testing</a> 33 </h5> 34<p> 35 <span class="emphasis"><em>Loopback</em></span> or <span class="emphasis"><em>round-tripping</em></span> refers 36 to writing out a value as a decimal digit string using <code class="computeroutput"><span class="identifier">std</span><span class="special">::</span><span class="identifier">iostream</span></code>, 37 usually to a <code class="computeroutput"><span class="identifier">std</span><span class="special">::</span><span class="identifier">stringstream</span></code>, and then reading the string 38 back in to another value, and confirming that the two values are identical. 39 A trivial example using <code class="computeroutput"><span class="keyword">float</span></code> 40 is: 41 </p> 42<pre class="programlisting"><span class="keyword">float</span> <span class="identifier">write</span><span class="special">;</span> <span class="comment">// Value to round-trip.</span> 43<span class="identifier">std</span><span class="special">::</span><span class="identifier">stringstream</span> <span class="identifier">ss</span><span class="special">;</span> <span class="comment">// Read and write std::stringstream.</span> 44<span class="identifier">ss</span><span class="special">.</span><span class="identifier">precision</span><span class="special">(</span><span class="identifier">std</span><span class="special">::</span><span class="identifier">numeric_limits</span><span class="special"><</span><span class="identifier">T</span><span class="special">>::</span><span class="identifier">max_digits10</span><span class="special">);</span> <span class="comment">// Ensure all potentially significant bits are output.</span> 45<span class="identifier">ss</span><span class="special">.</span><span class="identifier">flags</span><span class="special">(</span><span class="identifier">std</span><span class="special">::</span><span class="identifier">ios_base</span><span class="special">::</span><span class="identifier">fmtflags</span><span class="special">(</span><span class="identifier">std</span><span class="special">::</span><span class="identifier">ios_base</span><span class="special">::</span><span class="identifier">scientific</span><span class="special">));</span> <span class="comment">// Use scientific format.</span> 46<span class="identifier">ss</span> <span class="special"><<</span> <span class="identifier">write</span><span class="special">;</span> <span class="comment">// Output to string.</span> 47<span class="keyword">float</span> <span class="identifier">read</span><span class="special">;</span> <span class="comment">// Expected.</span> 48<span class="identifier">ss</span> <span class="special">>></span> <span class="identifier">read</span><span class="special">;</span> <span class="comment">// Read decimal digits string from stringstream.</span> 49<span class="identifier">BOOST_CHECK_EQUAL</span><span class="special">(</span><span class="identifier">write</span><span class="special">,</span> <span class="identifier">read</span><span class="special">);</span> <span class="comment">// Should be the same.</span> 50</pre> 51<p> 52 and this can be run in a loop for all possible values of a 32-bit float. 53 For other floating-point types <code class="computeroutput"><span class="identifier">T</span></code>, 54 including <a href="https://en.cppreference.com/w/cpp/language/types" target="_top">fundamental 55 (built-in)</a> <code class="computeroutput"><span class="keyword">double</span></code>, it 56 takes far too long to test all values, so a reasonable test strategy is to 57 use a large number of random values. 58 </p> 59<pre class="programlisting"><span class="identifier">T</span> <span class="identifier">write</span><span class="special">;</span> 60<span class="identifier">std</span><span class="special">::</span><span class="identifier">stringstream</span> <span class="identifier">ss</span><span class="special">;</span> 61<span class="identifier">ss</span><span class="special">.</span><span class="identifier">precision</span><span class="special">(</span><span class="identifier">std</span><span class="special">::</span><span class="identifier">numeric_limits</span><span class="special"><</span><span class="identifier">T</span><span class="special">>::</span><span class="identifier">max_digits10</span><span class="special">);</span> <span class="comment">// Ensure all potentially significant bits are output.</span> 62<span class="identifier">ss</span><span class="special">.</span><span class="identifier">flags</span><span class="special">(</span><span class="identifier">f</span><span class="special">);</span> <span class="comment">// Changed from default iostream format flags if desired.</span> 63<span class="identifier">ss</span> <span class="special"><<</span> <span class="identifier">write</span><span class="special">;</span> <span class="comment">// Output to stringstream.</span> 64 65<span class="identifier">T</span> <span class="identifier">read</span><span class="special">;</span> 66<span class="identifier">ss</span> <span class="special">>></span> <span class="identifier">read</span><span class="special">;</span> <span class="comment">// Get read using operator>> from stringstream.</span> 67<span class="identifier">BOOST_CHECK_EQUAL</span><span class="special">(</span><span class="identifier">read</span><span class="special">,</span> <span class="identifier">write</span><span class="special">);</span> 68 69<span class="identifier">read</span> <span class="special">=</span> <span class="keyword">static_cast</span><span class="special"><</span><span class="identifier">T</span><span class="special">>(</span><span class="identifier">ss</span><span class="special">.</span><span class="identifier">str</span><span class="special">());</span> <span class="comment">// Get read by converting from decimal digits string representation of write.</span> 70<span class="identifier">BOOST_CHECK_EQUAL</span><span class="special">(</span><span class="identifier">read</span><span class="special">,</span> <span class="identifier">write</span><span class="special">);</span> 71 72<span class="identifier">read</span> <span class="special">=</span> <span class="keyword">static_cast</span><span class="special"><</span><span class="identifier">T</span><span class="special">>(</span><span class="identifier">write</span><span class="special">.</span><span class="identifier">str</span><span class="special">(</span><span class="number">0</span><span class="special">,</span> <span class="identifier">f</span><span class="special">));</span> <span class="comment">// Get read using format specified when written.</span> 73<span class="identifier">BOOST_CHECK_EQUAL</span><span class="special">(</span><span class="identifier">read</span><span class="special">,</span> <span class="identifier">write</span><span class="special">);</span> 74</pre> 75<p> 76 The test at <a href="http://www.boost.org/doc/libs/release/libs/multiprecision/doc/html/../../test/test_cpp_bin_float_io.cpp" target="_top">test_cpp_bin_float_io.cpp</a> 77 allows any floating-point type to be <span class="emphasis"><em>round_tripped</em></span> using 78 a wide range of fairly random values. It also includes tests compared a collection 79 of <a href="http://www.boost.org/doc/libs/release/libs/multiprecision/doc/html/../../test/string_data.ipp" target="_top">stringdata</a> test cases 80 in a file. 81 </p> 82<h5> 83<a name="boost_multiprecision.tut.input_output.h1"></a> 84 <span class="phrase"><a name="boost_multiprecision.tut.input_output.comparing_with_output_using_fund"></a></span><a class="link" href="input_output.html#boost_multiprecision.tut.input_output.comparing_with_output_using_fund">Comparing 85 with output using fundamental 86 (built-in) types</a> 87 </h5> 88<p> 89 One can make some comparisons with the output of 90 </p> 91<pre class="programlisting"><span class="special"><</span><span class="identifier">number</span><span class="special"><</span><span class="identifier">cpp_bin_float</span><span class="special"><</span><span class="number">53</span><span class="special">,</span> <span class="identifier">digit_count_2</span><span class="special">></span> <span class="special">></span> 92</pre> 93<p> 94 which has the same number of significant bits (53) as 64-bit double precision 95 floating-point. 96 </p> 97<p> 98 However, although most outputs are identical, there are differences on some 99 platforms caused by the implementation-dependent behaviours allowed by the 100 C99 specification <a href="http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf" target="_top">C99 101 ISO/IEC 9899:TC2</a>, incorporated by C++. 102 </p> 103<div class="blockquote"><blockquote class="blockquote"><p> 104 <span class="emphasis"><em>"For e, E, f, F, g, and G conversions, if the number of 105 significant decimal digits is at most DECIMAL_DIG, then the result should 106 be correctly rounded. If the number of significant decimal digits is more 107 than DECIMAL_DIG but the source value is exactly representable with DECIMAL_DIG 108 digits, then the result should be an exact representation with trailing 109 zeros. Otherwise, the source value is bounded by two adjacent decimal strings 110 L < U, both having DECIMAL_DIG significant digits; the value of the 111 resultant decimal string D should satisfy L<= D <= U, with the extra 112 stipulation that the error should have a correct sign for the current rounding 113 direction."</em></span> 114 </p></blockquote></div> 115<p> 116 So not only is correct rounding for the full number of digits not required, 117 but even if the <span class="bold"><strong>optional</strong></span> recommended practice 118 is followed, then the value of these last few digits is unspecified as long 119 as the value is within certain bounds. 120 </p> 121<div class="note"><table border="0" summary="Note"> 122<tr> 123<td rowspan="2" align="center" valign="top" width="25"><img alt="[Note]" src="../../../../../../doc/src/images/note.png"></td> 124<th align="left">Note</th> 125</tr> 126<tr><td align="left" valign="top"><p> 127 Do not expect the output from different platforms to be <span class="bold"><strong>identical</strong></span>, 128 but <code class="computeroutput"><span class="identifier">cpp_dec_float</span></code>, <code class="computeroutput"><span class="identifier">cpp_bin_float</span></code> (and other backends) outputs 129 should be correctly rounded to the number of digits requested by the set 130 precision and format. 131 </p></td></tr> 132</table></div> 133<h5> 134<a name="boost_multiprecision.tut.input_output.h2"></a> 135 <span class="phrase"><a name="boost_multiprecision.tut.input_output.macro_boost_mp_min_exponent_digi"></a></span><a class="link" href="input_output.html#boost_multiprecision.tut.input_output.macro_boost_mp_min_exponent_digi">Macro 136 BOOST_MP_MIN_EXPONENT_DIGITS</a> 137 </h5> 138<p> 139 <a href="http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf" target="_top">C99 140 Standard</a> for format specifiers, 7.19.6 Formatted input/output functions 141 requires: 142 </p> 143<p> 144 "The exponent always contains at least two digits, and only as many 145 more digits as necessary to represent the exponent." 146 </p> 147<p> 148 So to conform to the C99 standard (incorporated by C++) 149 </p> 150<pre class="programlisting"><span class="preprocessor">#define</span> <span class="identifier">BOOST_MP_MIN_EXPONENT_DIGITS</span> <span class="number">2</span> 151</pre> 152<p> 153 Confusingly, Microsoft (and MinGW) do not conform to this standard and provide 154 <span class="bold"><strong>at least three digits</strong></span>, for example <code class="computeroutput"><span class="number">1e+001</span></code>. So if you want the output to match 155 that from <a href="https://en.cppreference.com/w/cpp/language/types" target="_top">fundamental 156 (built-in)</a> floating-point types on compilers that use Microsofts 157 runtime then use: 158 </p> 159<pre class="programlisting"><span class="preprocessor">#define</span> <span class="identifier">BOOST_MP_MIN_EXPONENT_DIGITS</span> <span class="number">3</span> 160</pre> 161<p> 162 Also useful to get the minimum exponent field width is 163 </p> 164<pre class="programlisting"><span class="preprocessor">#define</span> <span class="identifier">BOOST_MP_MIN_EXPONENT_DIGITS</span> <span class="number">1</span> 165</pre> 166<p> 167 producing a compact output like <code class="computeroutput"><span class="number">2e+4</span></code>, 168 useful when conserving space is important. 169 </p> 170<p> 171 Larger values are also supported, for example, value 4 for <code class="computeroutput"><span class="number">2e+0004</span></code> which may be useful to ensure that 172 columns line up. 173 </p> 174</div> 175<table xmlns:rev="http://www.cs.rpi.edu/~gregod/boost/tools/doc/revision" width="100%"><tr> 176<td align="left"></td> 177<td align="right"><div class="copyright-footer">Copyright © 2002-2020 John 178 Maddock and Christopher Kormanyos<p> 179 Distributed under the Boost Software License, Version 1.0. (See accompanying 180 file LICENSE_1_0.txt or copy at <a href="http://www.boost.org/LICENSE_1_0.txt" target="_top">http://www.boost.org/LICENSE_1_0.txt</a>) 181 </p> 182</div></td> 183</tr></table> 184<hr> 185<div class="spirit-nav"> 186<a accesskey="p" href="limits/how_to_tell.html"><img src="../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../tut.html"><img src="../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../index.html"><img src="../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="hash.html"><img src="../../../../../../doc/src/images/next.png" alt="Next"></a> 187</div> 188</body> 189</html> 190
