--- /dev/null
+ <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+ <html xmlns="http://www.w3.org/1999/xhtml">
+<!--
+This file was generated by Devel::NYTProf version 4.06
+-->
+<head>
+ <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
+ <meta http-equiv="Content-Language" content="en-us" />
+ <title>Profile of JSON/XS.pm</title>
+<link rel="stylesheet" type="text/css" href="style.css" />
+ <script type="text/javascript" src="js/jquery-min.js"></script>
+
+ <script type="text/javascript" src="js/jquery-tablesorter-min.js"></script>
+ <link rel="stylesheet" type="text/css" href="js/style-tablesorter.css" />
+ <script type="text/javascript">
+ // when a column is first clicked on to sort it, use descending order
+ // XXX doesn't seem to work (and not just because the tablesorter formatSortingOrder() is broken)
+ $.tablesorter.defaults.sortInitialOrder = "desc";
+ // add parser through the tablesorter addParser method
+ $.tablesorter.addParser({
+ id: 'fmt_time', // name of this parser
+ is: function(s) {
+ return false; // return false so this parser is not auto detected
+ },
+ format: function(orig) { // format data for normalization
+ // console.log(orig);
+ val = orig.replace(/ns/,'');
+ if (val != orig) { return val / (1000*1000*1000); }
+ val = orig.replace(/µs/,''); /* XXX use µ ? */
+ if (val != orig) { return val / (1000*1000); }
+ var val = orig.replace(/ms/,'');
+ if (val != orig) { return val / (1000); }
+ var val = orig.replace(/s/,'');
+ if (val != orig) { return val; }
+ if (orig == '0') { return orig; }
+ console.log('no match for fmt_time of '.concat(orig));
+ return orig;
+ },
+ type: 'numeric' // set type, either numeric or text
+ });
+ </script>
+</head>
+
+<body >
+<div class="header" style="position: relative; overflow-x: hidden; overflow-y: hidden; z-index: 0; ">
+<div class="header_back">
+ <a href="index.html">← Index</a>
+ </div>
+<div class="headerForeground" style="float: left">
+ <span class="siteTitle">NYTProf Performance Profile</span>
+ <span class="siteSubtitle">  <span>« <span class="mode_btn mode_btn_selected">block view</span> • <span class="mode_btn"><a href="JSON-XS-pm-679-line.html">line view</a></span> • <span class="mode_btn"><a href="JSON-XS-pm-679-sub.html">sub view</a></span> »</span><br />
+ For script/nytprof.pl
+ </span>
+</div>
+<div class="headerForeground" style="float: right; text-align: right">
+ <span class="siteTitle"> </span>
+ <span class="siteSubtitle">Run on Thu May 31 16:29:39 2012<br />Reported on Thu May 31 16:36:10 2012</span>
+</div>
+<div style="position: absolute; left: 0px; top: 0%; width: 100%; height: 101%; z-index: -1; background-color: rgb(17, 136, 255); "></div>
+<div style="position: absolute; left: 0px; top: 2%; width: 100%; height: 99%; z-index: -1; background-color: rgb(16, 134, 253); "></div>
+<div style="position: absolute; left: 0px; top: 4%; width: 100%; height: 97%; z-index: -1; background-color: rgb(16, 133, 252); "></div>
+<div style="position: absolute; left: 0px; top: 6%; width: 100%; height: 95%; z-index: -1; background-color: rgb(15, 131, 250); "></div>
+<div style="position: absolute; left: 0px; top: 8%; width: 100%; height: 93%; z-index: -1; background-color: rgb(15, 130, 249); "></div>
+<div style="position: absolute; left: 0px; top: 10%; width: 100%; height: 91%; z-index: -1; background-color: rgb(15, 129, 248); "></div>
+<div style="position: absolute; left: 0px; top: 12%; width: 100%; height: 89%; z-index: -1; background-color: rgb(14, 127, 246); "></div>
+<div style="position: absolute; left: 0px; top: 14%; width: 100%; height: 87%; z-index: -1; background-color: rgb(14, 126, 245); "></div>
+<div style="position: absolute; left: 0px; top: 16%; width: 100%; height: 85%; z-index: -1; background-color: rgb(14, 125, 244); "></div>
+<div style="position: absolute; left: 0px; top: 18%; width: 100%; height: 83%; z-index: -1; background-color: rgb(13, 123, 242); "></div>
+<div style="position: absolute; left: 0px; top: 20%; width: 100%; height: 81%; z-index: -1; background-color: rgb(13, 122, 241); "></div>
+<div style="position: absolute; left: 0px; top: 22%; width: 100%; height: 79%; z-index: -1; background-color: rgb(13, 121, 240); "></div>
+<div style="position: absolute; left: 0px; top: 24%; width: 100%; height: 77%; z-index: -1; background-color: rgb(12, 119, 238); "></div>
+<div style="position: absolute; left: 0px; top: 26%; width: 100%; height: 75%; z-index: -1; background-color: rgb(12, 118, 237); "></div>
+<div style="position: absolute; left: 0px; top: 28%; width: 100%; height: 73%; z-index: -1; background-color: rgb(12, 116, 235); "></div>
+<div style="position: absolute; left: 0px; top: 30%; width: 100%; height: 71%; z-index: -1; background-color: rgb(11, 115, 234); "></div>
+<div style="position: absolute; left: 0px; top: 32%; width: 100%; height: 69%; z-index: -1; background-color: rgb(11, 114, 233); "></div>
+<div style="position: absolute; left: 0px; top: 34%; width: 100%; height: 67%; z-index: -1; background-color: rgb(11, 112, 231); "></div>
+<div style="position: absolute; left: 0px; top: 36%; width: 100%; height: 65%; z-index: -1; background-color: rgb(10, 111, 230); "></div>
+<div style="position: absolute; left: 0px; top: 38%; width: 100%; height: 63%; z-index: -1; background-color: rgb(10, 110, 229); "></div>
+<div style="position: absolute; left: 0px; top: 40%; width: 100%; height: 61%; z-index: -1; background-color: rgb(10, 108, 227); "></div>
+<div style="position: absolute; left: 0px; top: 42%; width: 100%; height: 59%; z-index: -1; background-color: rgb(9, 107, 226); "></div>
+<div style="position: absolute; left: 0px; top: 44%; width: 100%; height: 57%; z-index: -1; background-color: rgb(9, 106, 225); "></div>
+<div style="position: absolute; left: 0px; top: 46%; width: 100%; height: 55%; z-index: -1; background-color: rgb(9, 104, 223); "></div>
+<div style="position: absolute; left: 0px; top: 48%; width: 100%; height: 53%; z-index: -1; background-color: rgb(8, 103, 222); "></div>
+<div style="position: absolute; left: 0px; top: 50%; width: 100%; height: 51%; z-index: -1; background-color: rgb(8, 102, 221); "></div>
+<div style="position: absolute; left: 0px; top: 52%; width: 100%; height: 49%; z-index: -1; background-color: rgb(8, 100, 219); "></div>
+<div style="position: absolute; left: 0px; top: 54%; width: 100%; height: 47%; z-index: -1; background-color: rgb(7, 99, 218); "></div>
+<div style="position: absolute; left: 0px; top: 56%; width: 100%; height: 45%; z-index: -1; background-color: rgb(7, 97, 216); "></div>
+<div style="position: absolute; left: 0px; top: 58%; width: 100%; height: 43%; z-index: -1; background-color: rgb(7, 96, 215); "></div>
+<div style="position: absolute; left: 0px; top: 60%; width: 100%; height: 41%; z-index: -1; background-color: rgb(6, 95, 214); "></div>
+<div style="position: absolute; left: 0px; top: 62%; width: 100%; height: 39%; z-index: -1; background-color: rgb(6, 93, 212); "></div>
+<div style="position: absolute; left: 0px; top: 64%; width: 100%; height: 37%; z-index: -1; background-color: rgb(6, 92, 211); "></div>
+<div style="position: absolute; left: 0px; top: 66%; width: 100%; height: 35%; z-index: -1; background-color: rgb(5, 91, 210); "></div>
+<div style="position: absolute; left: 0px; top: 68%; width: 100%; height: 33%; z-index: -1; background-color: rgb(5, 89, 208); "></div>
+<div style="position: absolute; left: 0px; top: 70%; width: 100%; height: 31%; z-index: -1; background-color: rgb(5, 88, 207); "></div>
+<div style="position: absolute; left: 0px; top: 72%; width: 100%; height: 29%; z-index: -1; background-color: rgb(4, 87, 206); "></div>
+<div style="position: absolute; left: 0px; top: 74%; width: 100%; height: 27%; z-index: -1; background-color: rgb(4, 85, 204); "></div>
+<div style="position: absolute; left: 0px; top: 76%; width: 100%; height: 25%; z-index: -1; background-color: rgb(4, 84, 203); "></div>
+<div style="position: absolute; left: 0px; top: 78%; width: 100%; height: 23%; z-index: -1; background-color: rgb(3, 82, 201); "></div>
+<div style="position: absolute; left: 0px; top: 80%; width: 100%; height: 21%; z-index: -1; background-color: rgb(3, 81, 200); "></div>
+<div style="position: absolute; left: 0px; top: 82%; width: 100%; height: 19%; z-index: -1; background-color: rgb(3, 80, 199); "></div>
+<div style="position: absolute; left: 0px; top: 84%; width: 100%; height: 17%; z-index: -1; background-color: rgb(2, 78, 197); "></div>
+<div style="position: absolute; left: 0px; top: 86%; width: 100%; height: 15%; z-index: -1; background-color: rgb(2, 77, 196); "></div>
+<div style="position: absolute; left: 0px; top: 88%; width: 100%; height: 13%; z-index: -1; background-color: rgb(2, 76, 195); "></div>
+<div style="position: absolute; left: 0px; top: 90%; width: 100%; height: 11%; z-index: -1; background-color: rgb(1, 74, 193); "></div>
+<div style="position: absolute; left: 0px; top: 92%; width: 100%; height: 9%; z-index: -1; background-color: rgb(1, 73, 192); "></div>
+<div style="position: absolute; left: 0px; top: 94%; width: 100%; height: 7%; z-index: -1; background-color: rgb(1, 72, 191); "></div>
+<div style="position: absolute; left: 0px; top: 96%; width: 100%; height: 5%; z-index: -1; background-color: rgb(0, 70, 189); "></div>
+<div style="position: absolute; left: 0px; top: 98%; width: 100%; height: 3%; z-index: -1; background-color: rgb(0, 69, 188); "></div>
+<div style="position: absolute; left: 0px; top: 100%; width: 100%; height: 1%; z-index: -1; background-color: rgb(0, 68, 187); "></div>
+</div>
+
+<div class="body_content"><br />
+<table class="file_summary"><tr><td class="h">Filename</td><td align="left"><a href="file:///Users/edenc/perl5/lib/perl5/darwin-multi-2level/JSON/XS.pm">/Users/edenc/perl5/lib/perl5/darwin-multi-2level/JSON/XS.pm</a></td></tr>
+<tr><td class="h">Statements</td><td align="left">Executed 15 statements in 2.66ms</td></tr></table>
+
+ <table id="subs_table" border="1" cellpadding="0" class="tablesorter">
+ <caption>Subroutines</caption>
+ <thead>
+ <tr>
+ <th>Calls</th>
+ <th><span title="Number of Places sub is called from">P</span></th>
+ <th><span title="Number of Files sub is called from">F</span></th>
+ <th>Exclusive<br />Time</th>
+ <th>Inclusive<br />Time</th>
+ <th>Subroutine</th>
+ </tr>
+ </thead>
+ <tbody>
+<tr><td class="c0">2326</td><td class="c3">1</td><td class="c3">1</td><td class="c0"><span title="0.1%">41.3ms</span></td><td class="c0"><span title="0.1%">41.3ms</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::encode</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#JSON__XS__encode">encode</a> (xsub)</span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c0"><span title="0.0%">554µs</span></td><td class="c0"><span title="0.0%">572µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::BEGIN@104</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#104">BEGIN@104</a></span></td></tr>
+<tr><td class="c0">4</td><td class="c3">1</td><td class="c3">1</td><td class="c1"><span title="0.0%">34µs</span></td><td class="c3"><span title="0.0%">34µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::new</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#JSON__XS__new">new</a> (xsub)</span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c3"><span title="0.0%">20µs</span></td><td class="c2"><span title="0.0%">72µs</span></td><td class="sub_name"><span style="display: none;">DynaLoader::::BEGIN@98</span> DynaLoader::<a href="JSON-XS-pm-679-block.html#98">BEGIN@98</a></span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c3"><span title="0.0%">17µs</span></td><td class="c2"><span title="0.0%">76µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::Boolean::::BEGIN@1477</span>JSON::XS::Boolean::<a href="JSON-XS-pm-679-block.html#1477">BEGIN@1477</a></span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c3"><span title="0.0%">16µs</span></td><td class="c3"><span title="0.0%">47µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::BEGIN@121</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#121">BEGIN@121</a></span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c3"><span title="0.0%">9µs</span></td><td class="c3"><span title="0.0%">9µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::BEGIN@122</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#122">BEGIN@122</a></span></td></tr>
+<tr><td class="c1">2</td><td class="c1">2</td><td class="c1">2</td><td class="c3"><span title="0.0%">4µs</span></td><td class="c3"><span title="0.0%">4µs</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::utf8</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#JSON__XS__utf8">utf8</a> (xsub)</span></td></tr>
+<tr><td class="c3">1</td><td class="c3">1</td><td class="c3">1</td><td class="c3"><span title="0.0%">900ns</span></td><td class="c3"><span title="0.0%">900ns</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::canonical</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#JSON__XS__canonical">canonical</a> (xsub)</span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::Boolean::::__ANON__[:1477]</span>JSON::XS::Boolean::<a href="JSON-XS-pm-679-block.html#1477">__ANON__[:1477]</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::Boolean::::__ANON__[:1478]</span>JSON::XS::Boolean::<a href="JSON-XS-pm-679-block.html#1478">__ANON__[:1478]</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::Boolean::::__ANON__[:1479]</span>JSON::XS::Boolean::<a href="JSON-XS-pm-679-block.html#1479">__ANON__[:1479]</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::false</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#1465">false</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::from_json</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#116">from_json</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::is_bool</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#1467">is_bool</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::to_json</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#111">to_json</a></span></td></tr>
+<tr><td class="c3">0</td><td class="c3">0</td><td class="c3">0</td><td class="c3"><span title="0.0%">0s</span></td><td class="c3"><span title="0.0%">0s</span></td><td class="sub_name"><span style="display: none;">JSON::XS::::true</span> JSON::XS::<a href="JSON-XS-pm-679-block.html#1464">true</a></span></td></tr>
+</tbody></table>
+ Call graph for these subroutines as a
+ <a href="http://en.wikipedia.org/wiki/Graphviz">Graphviz</a>
+ <a href="Users-edenc-perl5-lib-perl5-darwin-multi-2level-JSON-XS-pm.dot">dot language file</a>.
+
+ <table border="1" cellpadding="0">
+ <thead>
+ <tr><th>Line</th>
+ <th><span title="Number of statements executed">State<br />ments</span></th>
+ <th><span title="Time spend executing statements on the line,
+ excluding time spent executing statements in any called subroutines">Time<br />on line</span></th>
+ <th><span title="Number of subroutines calls">Calls</span></th>
+ <th><span title="Time spent in subroutines called (inclusive)">Time<br />in subs</span></th>
+ <th class="left_indent_header">Code</th>
+ </tr>
+
+ </thead>
+ <tbody>
+ <tr><td class="h"><a name="0"></a>0</td><td></td><td></td><td class="c3">1</td><td class="c3">72µs</td><td class="s">Profile data that couldn't be associated with a specific line:<div class="calls"><div class="calls_out"># spent 72µs making 1 call to <a href="JSON-XS-pm-679-block.html#98">DynaLoader::BEGIN@98</a></div></div></td></tr>
+<tr><td class="h"><a name="1"></a>1</td><td class="c3">1</td><td class="c3"><span title="Avg 56µs">56µs</span></td><td></td><td></td><td class="s">=head1 NAME</td></tr>
+<tr><td class="h"><a name="2"></a>2</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="3"></a>3</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS - JSON serialising/deserialising, done correctly and fast</td></tr>
+<tr><td class="h"><a name="4"></a>4</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="5"></a>5</td><td></td><td></td><td></td><td></td><td class="s">=encoding utf-8</td></tr>
+<tr><td class="h"><a name="6"></a>6</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="7"></a>7</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ</td></tr>
+<tr><td class="h"><a name="8"></a>8</td><td></td><td></td><td></td><td></td><td class="s"> (http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html)</td></tr>
+<tr><td class="h"><a name="9"></a>9</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="10"></a>10</td><td></td><td></td><td></td><td></td><td class="s">=head1 SYNOPSIS</td></tr>
+<tr><td class="h"><a name="11"></a>11</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="12"></a>12</td><td></td><td></td><td></td><td></td><td class="s"> use JSON::XS;</td></tr>
+<tr><td class="h"><a name="13"></a>13</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="14"></a>14</td><td></td><td></td><td></td><td></td><td class="s"> # exported functions, they croak on error</td></tr>
+<tr><td class="h"><a name="15"></a>15</td><td></td><td></td><td></td><td></td><td class="s"> # and expect/generate UTF-8</td></tr>
+<tr><td class="h"><a name="16"></a>16</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="17"></a>17</td><td></td><td></td><td></td><td></td><td class="s"> $utf8_encoded_json_text = encode_json $perl_hash_or_arrayref;</td></tr>
+<tr><td class="h"><a name="18"></a>18</td><td></td><td></td><td></td><td></td><td class="s"> $perl_hash_or_arrayref = decode_json $utf8_encoded_json_text;</td></tr>
+<tr><td class="h"><a name="19"></a>19</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="20"></a>20</td><td></td><td></td><td></td><td></td><td class="s"> # OO-interface</td></tr>
+<tr><td class="h"><a name="21"></a>21</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="22"></a>22</td><td></td><td></td><td></td><td></td><td class="s"> $coder = JSON::XS->new->ascii->pretty->allow_nonref;</td></tr>
+<tr><td class="h"><a name="23"></a>23</td><td></td><td></td><td></td><td></td><td class="s"> $pretty_printed_unencoded = $coder->encode ($perl_scalar);</td></tr>
+<tr><td class="h"><a name="24"></a>24</td><td></td><td></td><td></td><td></td><td class="s"> $perl_scalar = $coder->decode ($unicode_json_text);</td></tr>
+<tr><td class="h"><a name="25"></a>25</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="26"></a>26</td><td></td><td></td><td></td><td></td><td class="s"> # Note that JSON version 2.0 and above will automatically use JSON::XS</td></tr>
+<tr><td class="h"><a name="27"></a>27</td><td></td><td></td><td></td><td></td><td class="s"> # if available, at virtually no speed overhead either, so you should</td></tr>
+<tr><td class="h"><a name="28"></a>28</td><td></td><td></td><td></td><td></td><td class="s"> # be able to just:</td></tr>
+<tr><td class="h"><a name="29"></a>29</td><td></td><td></td><td></td><td></td><td class="s"> </td></tr>
+<tr><td class="h"><a name="30"></a>30</td><td></td><td></td><td></td><td></td><td class="s"> use JSON;</td></tr>
+<tr><td class="h"><a name="31"></a>31</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="32"></a>32</td><td></td><td></td><td></td><td></td><td class="s"> # and do the same things, except that you have a pure-perl fallback now.</td></tr>
+<tr><td class="h"><a name="33"></a>33</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="34"></a>34</td><td></td><td></td><td></td><td></td><td class="s">=head1 DESCRIPTION</td></tr>
+<tr><td class="h"><a name="35"></a>35</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="36"></a>36</td><td></td><td></td><td></td><td></td><td class="s">This module converts Perl data structures to JSON and vice versa. Its</td></tr>
+<tr><td class="h"><a name="37"></a>37</td><td></td><td></td><td></td><td></td><td class="s">primary goal is to be I<correct> and its secondary goal is to be</td></tr>
+<tr><td class="h"><a name="38"></a>38</td><td></td><td></td><td></td><td></td><td class="s">I<fast>. To reach the latter goal it was written in C.</td></tr>
+<tr><td class="h"><a name="39"></a>39</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="40"></a>40</td><td></td><td></td><td></td><td></td><td class="s">Beginning with version 2.0 of the JSON module, when both JSON and</td></tr>
+<tr><td class="h"><a name="41"></a>41</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS are installed, then JSON will fall back on JSON::XS (this can be</td></tr>
+<tr><td class="h"><a name="42"></a>42</td><td></td><td></td><td></td><td></td><td class="s">overridden) with no overhead due to emulation (by inheriting constructor</td></tr>
+<tr><td class="h"><a name="43"></a>43</td><td></td><td></td><td></td><td></td><td class="s">and methods). If JSON::XS is not available, it will fall back to the</td></tr>
+<tr><td class="h"><a name="44"></a>44</td><td></td><td></td><td></td><td></td><td class="s">compatible JSON::PP module as backend, so using JSON instead of JSON::XS</td></tr>
+<tr><td class="h"><a name="45"></a>45</td><td></td><td></td><td></td><td></td><td class="s">gives you a portable JSON API that can be fast when you need and doesn't</td></tr>
+<tr><td class="h"><a name="46"></a>46</td><td></td><td></td><td></td><td></td><td class="s">require a C compiler when that is a problem.</td></tr>
+<tr><td class="h"><a name="47"></a>47</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="48"></a>48</td><td></td><td></td><td></td><td></td><td class="s">As this is the n-th-something JSON module on CPAN, what was the reason</td></tr>
+<tr><td class="h"><a name="49"></a>49</td><td></td><td></td><td></td><td></td><td class="s">to write yet another JSON module? While it seems there are many JSON</td></tr>
+<tr><td class="h"><a name="50"></a>50</td><td></td><td></td><td></td><td></td><td class="s">modules, none of them correctly handle all corner cases, and in most cases</td></tr>
+<tr><td class="h"><a name="51"></a>51</td><td></td><td></td><td></td><td></td><td class="s">their maintainers are unresponsive, gone missing, or not listening to bug</td></tr>
+<tr><td class="h"><a name="52"></a>52</td><td></td><td></td><td></td><td></td><td class="s">reports for other reasons.</td></tr>
+<tr><td class="h"><a name="53"></a>53</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="54"></a>54</td><td></td><td></td><td></td><td></td><td class="s">See MAPPING, below, on how JSON::XS maps perl values to JSON values and</td></tr>
+<tr><td class="h"><a name="55"></a>55</td><td></td><td></td><td></td><td></td><td class="s">vice versa.</td></tr>
+<tr><td class="h"><a name="56"></a>56</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="57"></a>57</td><td></td><td></td><td></td><td></td><td class="s">=head2 FEATURES</td></tr>
+<tr><td class="h"><a name="58"></a>58</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="59"></a>59</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="60"></a>60</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="61"></a>61</td><td></td><td></td><td></td><td></td><td class="s">=item * correct Unicode handling</td></tr>
+<tr><td class="h"><a name="62"></a>62</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="63"></a>63</td><td></td><td></td><td></td><td></td><td class="s">This module knows how to handle Unicode, documents how and when it does</td></tr>
+<tr><td class="h"><a name="64"></a>64</td><td></td><td></td><td></td><td></td><td class="s">so, and even documents what "correct" means.</td></tr>
+<tr><td class="h"><a name="65"></a>65</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="66"></a>66</td><td></td><td></td><td></td><td></td><td class="s">=item * round-trip integrity</td></tr>
+<tr><td class="h"><a name="67"></a>67</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="68"></a>68</td><td></td><td></td><td></td><td></td><td class="s">When you serialise a perl data structure using only data types supported</td></tr>
+<tr><td class="h"><a name="69"></a>69</td><td></td><td></td><td></td><td></td><td class="s">by JSON and Perl, the deserialised data structure is identical on the Perl</td></tr>
+<tr><td class="h"><a name="70"></a>70</td><td></td><td></td><td></td><td></td><td class="s">level. (e.g. the string "2.0" doesn't suddenly become "2" just because</td></tr>
+<tr><td class="h"><a name="71"></a>71</td><td></td><td></td><td></td><td></td><td class="s">it looks like a number). There I<are> minor exceptions to this, read the</td></tr>
+<tr><td class="h"><a name="72"></a>72</td><td></td><td></td><td></td><td></td><td class="s">MAPPING section below to learn about those.</td></tr>
+<tr><td class="h"><a name="73"></a>73</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="74"></a>74</td><td></td><td></td><td></td><td></td><td class="s">=item * strict checking of JSON correctness</td></tr>
+<tr><td class="h"><a name="75"></a>75</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="76"></a>76</td><td></td><td></td><td></td><td></td><td class="s">There is no guessing, no generating of illegal JSON texts by default,</td></tr>
+<tr><td class="h"><a name="77"></a>77</td><td></td><td></td><td></td><td></td><td class="s">and only JSON is accepted as input by default (the latter is a security</td></tr>
+<tr><td class="h"><a name="78"></a>78</td><td></td><td></td><td></td><td></td><td class="s">feature).</td></tr>
+<tr><td class="h"><a name="79"></a>79</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="80"></a>80</td><td></td><td></td><td></td><td></td><td class="s">=item * fast</td></tr>
+<tr><td class="h"><a name="81"></a>81</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="82"></a>82</td><td></td><td></td><td></td><td></td><td class="s">Compared to other JSON modules and other serialisers such as Storable,</td></tr>
+<tr><td class="h"><a name="83"></a>83</td><td></td><td></td><td></td><td></td><td class="s">this module usually compares favourably in terms of speed, too.</td></tr>
+<tr><td class="h"><a name="84"></a>84</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="85"></a>85</td><td></td><td></td><td></td><td></td><td class="s">=item * simple to use</td></tr>
+<tr><td class="h"><a name="86"></a>86</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="87"></a>87</td><td></td><td></td><td></td><td></td><td class="s">This module has both a simple functional interface as well as an object</td></tr>
+<tr><td class="h"><a name="88"></a>88</td><td></td><td></td><td></td><td></td><td class="s">oriented interface interface.</td></tr>
+<tr><td class="h"><a name="89"></a>89</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="90"></a>90</td><td></td><td></td><td></td><td></td><td class="s">=item * reasonably versatile output formats</td></tr>
+<tr><td class="h"><a name="91"></a>91</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="92"></a>92</td><td></td><td></td><td></td><td></td><td class="s">You can choose between the most compact guaranteed-single-line format</td></tr>
+<tr><td class="h"><a name="93"></a>93</td><td></td><td></td><td></td><td></td><td class="s">possible (nice for simple line-based protocols), a pure-ASCII format</td></tr>
+<tr><td class="h"><a name="94"></a>94</td><td></td><td></td><td></td><td></td><td class="s">(for when your transport is not 8-bit clean, still supports the whole</td></tr>
+<tr><td class="h"><a name="95"></a>95</td><td></td><td></td><td></td><td></td><td class="s">Unicode range), or a pretty-printed format (for when you want to read that</td></tr>
+<tr><td class="h"><a name="96"></a>96</td><td></td><td></td><td></td><td></td><td class="s">stuff). Or you can combine those features in whatever way you like.</td></tr>
+<tr><td class="h"><a name="97"></a>97</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="98"></a>98</td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 72µs (20+52) within DynaLoader::BEGIN@98 which was called:
+# once (20µs+52µs) by XSLoader::load at <a href="JSON-XS-pm-679-block.html#0">line 0</a></div></div>=back</td></tr>
+<tr><td class="h"><a name="99"></a>99</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="100"></a>100</td><td></td><td></td><td></td><td></td><td class="s">=cut</td></tr>
+<tr><td class="h"><a name="101"></a>101</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="102"></a>102</td><td></td><td></td><td></td><td></td><td class="s">package JSON::XS;</td></tr>
+<tr><td class="h"><a name="103"></a>103</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="104"></a>104</td><td class="c1">2</td><td class="c0"><span title="Avg 348µs">696µs</span></td><td class="c3">2</td><td class="c0">590µs</td><td class="s"><div class="calls"><div class="calls_in"># spent 572µs (554+18) within JSON::XS::BEGIN@104 which was called:
+# once (554µs+18µs) by JSON::BEGIN@2 at <a href="JSON-XS-pm-679-block.html#104">line 104</a></div></div>use common::sense;<div class="calls"><div class="calls_out"># spent 572µs making 1 call to <a href="JSON-XS-pm-679-block.html#104">JSON::XS::BEGIN@104</a>
+# spent 18µs making 1 call to <a href="common-sense-pm-680-block.html#235">common::sense::import</a></div></div></td></tr>
+<tr><td class="h"><a name="105"></a>105</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="106"></a>106</td><td class="c3">1</td><td class="c3"><span title="Avg 1µs">1µs</span></td><td></td><td></td><td class="s">our $VERSION = '2.32';</td></tr>
+<tr><td class="h"><a name="107"></a>107</td><td class="c3">1</td><td class="c3"><span title="Avg 9µs">9µs</span></td><td></td><td></td><td class="s">our @ISA = qw(Exporter);</td></tr>
+<tr><td class="h"><a name="108"></a>108</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="109"></a>109</td><td class="c3">1</td><td class="c3"><span title="Avg 2µs">2µs</span></td><td></td><td></td><td class="s">our @EXPORT = qw(encode_json decode_json to_json from_json);</td></tr>
+<tr><td class="h"><a name="110"></a>110</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="111"></a>111</td><td></td><td></td><td></td><td></td><td class="s">sub to_json($) {</td></tr>
+<tr><td class="h"><a name="112"></a>112</td><td></td><td></td><td></td><td></td><td class="s"> require Carp;</td></tr>
+<tr><td class="h"><a name="113"></a>113</td><td></td><td></td><td></td><td></td><td class="s"> Carp::croak ("JSON::XS::to_json has been renamed to encode_json, either downgrade to pre-2.0 versions of JSON::XS or rename the call");</td></tr>
+<tr><td class="h"><a name="114"></a>114</td><td></td><td></td><td></td><td></td><td class="s">}</td></tr>
+<tr><td class="h"><a name="115"></a>115</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="116"></a>116</td><td></td><td></td><td></td><td></td><td class="s">sub from_json($) {</td></tr>
+<tr><td class="h"><a name="117"></a>117</td><td></td><td></td><td></td><td></td><td class="s"> require Carp;</td></tr>
+<tr><td class="h"><a name="118"></a>118</td><td></td><td></td><td></td><td></td><td class="s"> Carp::croak ("JSON::XS::from_json has been renamed to decode_json, either downgrade to pre-2.0 versions of JSON::XS or rename the call");</td></tr>
+<tr><td class="h"><a name="119"></a>119</td><td></td><td></td><td></td><td></td><td class="s">}</td></tr>
+<tr><td class="h"><a name="120"></a>120</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="121"></a>121</td><td class="c1">2</td><td class="c3"><span title="Avg 20µs">40µs</span></td><td class="c3">2</td><td class="c3">77µs</td><td class="s"><div class="calls"><div class="calls_in"># spent 47µs (16+31) within JSON::XS::BEGIN@121 which was called:
+# once (16µs+31µs) by JSON::BEGIN@2 at <a href="JSON-XS-pm-679-block.html#121">line 121</a></div></div>use Exporter;<div class="calls"><div class="calls_out"># spent 47µs making 1 call to <a href="JSON-XS-pm-679-block.html#121">JSON::XS::BEGIN@121</a>
+# spent 31µs making 1 call to <a href="Exporter-pm-8-block.html#28">Exporter::import</a></div></div></td></tr>
+<tr><td class="h"><a name="122"></a>122</td><td class="c1">2</td><td class="c0"><span title="Avg 616µs">1.23ms</span></td><td class="c3">1</td><td class="c3">9µs</td><td class="s"><div class="calls"><div class="calls_in"># spent 9µs within JSON::XS::BEGIN@122 which was called:
+# once (9µs+0s) by JSON::BEGIN@2 at <a href="JSON-XS-pm-679-block.html#122">line 122</a></div></div>use XSLoader;<div class="calls"><div class="calls_out"># spent 9µs making 1 call to <a href="JSON-XS-pm-679-block.html#122">JSON::XS::BEGIN@122</a></div></div></td></tr>
+<tr><td class="h"><a name="123"></a>123</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="124"></a>124</td><td></td><td></td><td></td><td></td><td class="s">=head1 FUNCTIONAL INTERFACE</td></tr>
+<tr><td class="h"><a name="125"></a>125</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="126"></a>126</td><td></td><td></td><td></td><td></td><td class="s">The following convenience methods are provided by this module. They are</td></tr>
+<tr><td class="h"><a name="127"></a>127</td><td></td><td></td><td></td><td></td><td class="s">exported by default:</td></tr>
+<tr><td class="h"><a name="128"></a>128</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="129"></a>129</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="130"></a>130</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="131"></a>131</td><td></td><td></td><td></td><td></td><td class="s">=item $json_text = encode_json $perl_scalar</td></tr>
+<tr><td class="h"><a name="132"></a>132</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="133"></a>133</td><td></td><td></td><td></td><td></td><td class="s">Converts the given Perl data structure to a UTF-8 encoded, binary string</td></tr>
+<tr><td class="h"><a name="134"></a>134</td><td></td><td></td><td></td><td></td><td class="s">(that is, the string contains octets only). Croaks on error.</td></tr>
+<tr><td class="h"><a name="135"></a>135</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="136"></a>136</td><td></td><td></td><td></td><td></td><td class="s">This function call is functionally identical to:</td></tr>
+<tr><td class="h"><a name="137"></a>137</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="138"></a>138</td><td></td><td></td><td></td><td></td><td class="s"> $json_text = JSON::XS->new->utf8->encode ($perl_scalar)</td></tr>
+<tr><td class="h"><a name="139"></a>139</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="140"></a>140</td><td></td><td></td><td></td><td></td><td class="s">Except being faster.</td></tr>
+<tr><td class="h"><a name="141"></a>141</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="142"></a>142</td><td></td><td></td><td></td><td></td><td class="s">=item $perl_scalar = decode_json $json_text</td></tr>
+<tr><td class="h"><a name="143"></a>143</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="144"></a>144</td><td></td><td></td><td></td><td></td><td class="s">The opposite of C<encode_json>: expects an UTF-8 (binary) string and tries</td></tr>
+<tr><td class="h"><a name="145"></a>145</td><td></td><td></td><td></td><td></td><td class="s">to parse that as an UTF-8 encoded JSON text, returning the resulting</td></tr>
+<tr><td class="h"><a name="146"></a>146</td><td></td><td></td><td></td><td></td><td class="s">reference. Croaks on error.</td></tr>
+<tr><td class="h"><a name="147"></a>147</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="148"></a>148</td><td></td><td></td><td></td><td></td><td class="s">This function call is functionally identical to:</td></tr>
+<tr><td class="h"><a name="149"></a>149</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="150"></a>150</td><td></td><td></td><td></td><td></td><td class="s"> $perl_scalar = JSON::XS->new->utf8->decode ($json_text)</td></tr>
+<tr><td class="h"><a name="151"></a>151</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="152"></a>152</td><td></td><td></td><td></td><td></td><td class="s">Except being faster.</td></tr>
+<tr><td class="h"><a name="153"></a>153</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="154"></a>154</td><td></td><td></td><td></td><td></td><td class="s">=item $is_boolean = JSON::XS::is_bool $scalar</td></tr>
+<tr><td class="h"><a name="155"></a>155</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="156"></a>156</td><td></td><td></td><td></td><td></td><td class="s">Returns true if the passed scalar represents either JSON::XS::true or</td></tr>
+<tr><td class="h"><a name="157"></a>157</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS::false, two constants that act like C<1> and C<0>, respectively</td></tr>
+<tr><td class="h"><a name="158"></a>158</td><td></td><td></td><td></td><td></td><td class="s">and are used to represent JSON C<true> and C<false> values in Perl.</td></tr>
+<tr><td class="h"><a name="159"></a>159</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="160"></a>160</td><td></td><td></td><td></td><td></td><td class="s">See MAPPING, below, for more information on how JSON values are mapped to</td></tr>
+<tr><td class="h"><a name="161"></a>161</td><td></td><td></td><td></td><td></td><td class="s">Perl.</td></tr>
+<tr><td class="h"><a name="162"></a>162</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="163"></a>163</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="164"></a>164</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="165"></a>165</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="166"></a>166</td><td></td><td></td><td></td><td></td><td class="s">=head1 A FEW NOTES ON UNICODE AND PERL</td></tr>
+<tr><td class="h"><a name="167"></a>167</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="168"></a>168</td><td></td><td></td><td></td><td></td><td class="s">Since this often leads to confusion, here are a few very clear words on</td></tr>
+<tr><td class="h"><a name="169"></a>169</td><td></td><td></td><td></td><td></td><td class="s">how Unicode works in Perl, modulo bugs.</td></tr>
+<tr><td class="h"><a name="170"></a>170</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="171"></a>171</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="172"></a>172</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="173"></a>173</td><td></td><td></td><td></td><td></td><td class="s">=item 1. Perl strings can store characters with ordinal values > 255.</td></tr>
+<tr><td class="h"><a name="174"></a>174</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="175"></a>175</td><td></td><td></td><td></td><td></td><td class="s">This enables you to store Unicode characters as single characters in a</td></tr>
+<tr><td class="h"><a name="176"></a>176</td><td></td><td></td><td></td><td></td><td class="s">Perl string - very natural.</td></tr>
+<tr><td class="h"><a name="177"></a>177</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="178"></a>178</td><td></td><td></td><td></td><td></td><td class="s">=item 2. Perl does I<not> associate an encoding with your strings.</td></tr>
+<tr><td class="h"><a name="179"></a>179</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="180"></a>180</td><td></td><td></td><td></td><td></td><td class="s">... until you force it to, e.g. when matching it against a regex, or</td></tr>
+<tr><td class="h"><a name="181"></a>181</td><td></td><td></td><td></td><td></td><td class="s">printing the scalar to a file, in which case Perl either interprets your</td></tr>
+<tr><td class="h"><a name="182"></a>182</td><td></td><td></td><td></td><td></td><td class="s">string as locale-encoded text, octets/binary, or as Unicode, depending</td></tr>
+<tr><td class="h"><a name="183"></a>183</td><td></td><td></td><td></td><td></td><td class="s">on various settings. In no case is an encoding stored together with your</td></tr>
+<tr><td class="h"><a name="184"></a>184</td><td></td><td></td><td></td><td></td><td class="s">data, it is I<use> that decides encoding, not any magical meta data.</td></tr>
+<tr><td class="h"><a name="185"></a>185</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="186"></a>186</td><td></td><td></td><td></td><td></td><td class="s">=item 3. The internal utf-8 flag has no meaning with regards to the</td></tr>
+<tr><td class="h"><a name="187"></a>187</td><td></td><td></td><td></td><td></td><td class="s">encoding of your string.</td></tr>
+<tr><td class="h"><a name="188"></a>188</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="189"></a>189</td><td></td><td></td><td></td><td></td><td class="s">Just ignore that flag unless you debug a Perl bug, a module written in</td></tr>
+<tr><td class="h"><a name="190"></a>190</td><td></td><td></td><td></td><td></td><td class="s">XS or want to dive into the internals of perl. Otherwise it will only</td></tr>
+<tr><td class="h"><a name="191"></a>191</td><td></td><td></td><td></td><td></td><td class="s">confuse you, as, despite the name, it says nothing about how your string</td></tr>
+<tr><td class="h"><a name="192"></a>192</td><td></td><td></td><td></td><td></td><td class="s">is encoded. You can have Unicode strings with that flag set, with that</td></tr>
+<tr><td class="h"><a name="193"></a>193</td><td></td><td></td><td></td><td></td><td class="s">flag clear, and you can have binary data with that flag set and that flag</td></tr>
+<tr><td class="h"><a name="194"></a>194</td><td></td><td></td><td></td><td></td><td class="s">clear. Other possibilities exist, too.</td></tr>
+<tr><td class="h"><a name="195"></a>195</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="196"></a>196</td><td></td><td></td><td></td><td></td><td class="s">If you didn't know about that flag, just the better, pretend it doesn't</td></tr>
+<tr><td class="h"><a name="197"></a>197</td><td></td><td></td><td></td><td></td><td class="s">exist.</td></tr>
+<tr><td class="h"><a name="198"></a>198</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="199"></a>199</td><td></td><td></td><td></td><td></td><td class="s">=item 4. A "Unicode String" is simply a string where each character can be</td></tr>
+<tr><td class="h"><a name="200"></a>200</td><td></td><td></td><td></td><td></td><td class="s">validly interpreted as a Unicode code point.</td></tr>
+<tr><td class="h"><a name="201"></a>201</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="202"></a>202</td><td></td><td></td><td></td><td></td><td class="s">If you have UTF-8 encoded data, it is no longer a Unicode string, but a</td></tr>
+<tr><td class="h"><a name="203"></a>203</td><td></td><td></td><td></td><td></td><td class="s">Unicode string encoded in UTF-8, giving you a binary string.</td></tr>
+<tr><td class="h"><a name="204"></a>204</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="205"></a>205</td><td></td><td></td><td></td><td></td><td class="s">=item 5. A string containing "high" (> 255) character values is I<not> a UTF-8 string.</td></tr>
+<tr><td class="h"><a name="206"></a>206</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="207"></a>207</td><td></td><td></td><td></td><td></td><td class="s">It's a fact. Learn to live with it.</td></tr>
+<tr><td class="h"><a name="208"></a>208</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="209"></a>209</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="210"></a>210</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="211"></a>211</td><td></td><td></td><td></td><td></td><td class="s">I hope this helps :)</td></tr>
+<tr><td class="h"><a name="212"></a>212</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="213"></a>213</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="214"></a>214</td><td></td><td></td><td></td><td></td><td class="s">=head1 OBJECT-ORIENTED INTERFACE</td></tr>
+<tr><td class="h"><a name="215"></a>215</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="216"></a>216</td><td></td><td></td><td></td><td></td><td class="s">The object oriented interface lets you configure your own encoding or</td></tr>
+<tr><td class="h"><a name="217"></a>217</td><td></td><td></td><td></td><td></td><td class="s">decoding style, within the limits of supported formats.</td></tr>
+<tr><td class="h"><a name="218"></a>218</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="219"></a>219</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="220"></a>220</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="221"></a>221</td><td></td><td></td><td></td><td></td><td class="s">=item $json = new JSON::XS</td></tr>
+<tr><td class="h"><a name="222"></a>222</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="223"></a>223</td><td></td><td></td><td></td><td></td><td class="s">Creates a new JSON::XS object that can be used to de/encode JSON</td></tr>
+<tr><td class="h"><a name="224"></a>224</td><td></td><td></td><td></td><td></td><td class="s">strings. All boolean flags described below are by default I<disabled>.</td></tr>
+<tr><td class="h"><a name="225"></a>225</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="226"></a>226</td><td></td><td></td><td></td><td></td><td class="s">The mutators for flags all return the JSON object again and thus calls can</td></tr>
+<tr><td class="h"><a name="227"></a>227</td><td></td><td></td><td></td><td></td><td class="s">be chained:</td></tr>
+<tr><td class="h"><a name="228"></a>228</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="229"></a>229</td><td></td><td></td><td></td><td></td><td class="s"> my $json = JSON::XS->new->utf8->space_after->encode ({a => [1,2]})</td></tr>
+<tr><td class="h"><a name="230"></a>230</td><td></td><td></td><td></td><td></td><td class="s"> => {"a": [1, 2]}</td></tr>
+<tr><td class="h"><a name="231"></a>231</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="232"></a>232</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->ascii ([$enable])</td></tr>
+<tr><td class="h"><a name="233"></a>233</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="234"></a>234</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_ascii</td></tr>
+<tr><td class="h"><a name="235"></a>235</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="236"></a>236</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will not</td></tr>
+<tr><td class="h"><a name="237"></a>237</td><td></td><td></td><td></td><td></td><td class="s">generate characters outside the code range C<0..127> (which is ASCII). Any</td></tr>
+<tr><td class="h"><a name="238"></a>238</td><td></td><td></td><td></td><td></td><td class="s">Unicode characters outside that range will be escaped using either a</td></tr>
+<tr><td class="h"><a name="239"></a>239</td><td></td><td></td><td></td><td></td><td class="s">single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence,</td></tr>
+<tr><td class="h"><a name="240"></a>240</td><td></td><td></td><td></td><td></td><td class="s">as per RFC4627. The resulting encoded JSON text can be treated as a native</td></tr>
+<tr><td class="h"><a name="241"></a>241</td><td></td><td></td><td></td><td></td><td class="s">Unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string,</td></tr>
+<tr><td class="h"><a name="242"></a>242</td><td></td><td></td><td></td><td></td><td class="s">or any other superset of ASCII.</td></tr>
+<tr><td class="h"><a name="243"></a>243</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="244"></a>244</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will not escape Unicode</td></tr>
+<tr><td class="h"><a name="245"></a>245</td><td></td><td></td><td></td><td></td><td class="s">characters unless required by the JSON syntax or other flags. This results</td></tr>
+<tr><td class="h"><a name="246"></a>246</td><td></td><td></td><td></td><td></td><td class="s">in a faster and more compact format.</td></tr>
+<tr><td class="h"><a name="247"></a>247</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="248"></a>248</td><td></td><td></td><td></td><td></td><td class="s">See also the section I<ENCODING/CODESET FLAG NOTES> later in this</td></tr>
+<tr><td class="h"><a name="249"></a>249</td><td></td><td></td><td></td><td></td><td class="s">document.</td></tr>
+<tr><td class="h"><a name="250"></a>250</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="251"></a>251</td><td></td><td></td><td></td><td></td><td class="s">The main use for this flag is to produce JSON texts that can be</td></tr>
+<tr><td class="h"><a name="252"></a>252</td><td></td><td></td><td></td><td></td><td class="s">transmitted over a 7-bit channel, as the encoded JSON texts will not</td></tr>
+<tr><td class="h"><a name="253"></a>253</td><td></td><td></td><td></td><td></td><td class="s">contain any 8 bit characters.</td></tr>
+<tr><td class="h"><a name="254"></a>254</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="255"></a>255</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS->new->ascii (1)->encode ([chr 0x10401])</td></tr>
+<tr><td class="h"><a name="256"></a>256</td><td></td><td></td><td></td><td></td><td class="s"> => ["\ud801\udc01"]</td></tr>
+<tr><td class="h"><a name="257"></a>257</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="258"></a>258</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->latin1 ([$enable])</td></tr>
+<tr><td class="h"><a name="259"></a>259</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="260"></a>260</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_latin1</td></tr>
+<tr><td class="h"><a name="261"></a>261</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="262"></a>262</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will encode</td></tr>
+<tr><td class="h"><a name="263"></a>263</td><td></td><td></td><td></td><td></td><td class="s">the resulting JSON text as latin1 (or iso-8859-1), escaping any characters</td></tr>
+<tr><td class="h"><a name="264"></a>264</td><td></td><td></td><td></td><td></td><td class="s">outside the code range C<0..255>. The resulting string can be treated as a</td></tr>
+<tr><td class="h"><a name="265"></a>265</td><td></td><td></td><td></td><td></td><td class="s">latin1-encoded JSON text or a native Unicode string. The C<decode> method</td></tr>
+<tr><td class="h"><a name="266"></a>266</td><td></td><td></td><td></td><td></td><td class="s">will not be affected in any way by this flag, as C<decode> by default</td></tr>
+<tr><td class="h"><a name="267"></a>267</td><td></td><td></td><td></td><td></td><td class="s">expects Unicode, which is a strict superset of latin1.</td></tr>
+<tr><td class="h"><a name="268"></a>268</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="269"></a>269</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will not escape Unicode</td></tr>
+<tr><td class="h"><a name="270"></a>270</td><td></td><td></td><td></td><td></td><td class="s">characters unless required by the JSON syntax or other flags.</td></tr>
+<tr><td class="h"><a name="271"></a>271</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="272"></a>272</td><td></td><td></td><td></td><td></td><td class="s">See also the section I<ENCODING/CODESET FLAG NOTES> later in this</td></tr>
+<tr><td class="h"><a name="273"></a>273</td><td></td><td></td><td></td><td></td><td class="s">document.</td></tr>
+<tr><td class="h"><a name="274"></a>274</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="275"></a>275</td><td></td><td></td><td></td><td></td><td class="s">The main use for this flag is efficiently encoding binary data as JSON</td></tr>
+<tr><td class="h"><a name="276"></a>276</td><td></td><td></td><td></td><td></td><td class="s">text, as most octets will not be escaped, resulting in a smaller encoded</td></tr>
+<tr><td class="h"><a name="277"></a>277</td><td></td><td></td><td></td><td></td><td class="s">size. The disadvantage is that the resulting JSON text is encoded</td></tr>
+<tr><td class="h"><a name="278"></a>278</td><td></td><td></td><td></td><td></td><td class="s">in latin1 (and must correctly be treated as such when storing and</td></tr>
+<tr><td class="h"><a name="279"></a>279</td><td></td><td></td><td></td><td></td><td class="s">transferring), a rare encoding for JSON. It is therefore most useful when</td></tr>
+<tr><td class="h"><a name="280"></a>280</td><td></td><td></td><td></td><td></td><td class="s">you want to store data structures known to contain binary data efficiently</td></tr>
+<tr><td class="h"><a name="281"></a>281</td><td></td><td></td><td></td><td></td><td class="s">in files or databases, not when talking to other JSON encoders/decoders.</td></tr>
+<tr><td class="h"><a name="282"></a>282</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="283"></a>283</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS->new->latin1->encode (["\x{89}\x{abc}"]</td></tr>
+<tr><td class="h"><a name="284"></a>284</td><td></td><td></td><td></td><td></td><td class="s"> => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not)</td></tr>
+<tr><td class="h"><a name="285"></a>285</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="286"></a>286</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->utf8 ([$enable])</td></tr>
+<tr><td class="h"><a name="287"></a>287</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="288"></a>288</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_utf8</td></tr>
+<tr><td class="h"><a name="289"></a>289</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="290"></a>290</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will encode</td></tr>
+<tr><td class="h"><a name="291"></a>291</td><td></td><td></td><td></td><td></td><td class="s">the JSON result into UTF-8, as required by many protocols, while the</td></tr>
+<tr><td class="h"><a name="292"></a>292</td><td></td><td></td><td></td><td></td><td class="s">C<decode> method expects to be handled an UTF-8-encoded string. Please</td></tr>
+<tr><td class="h"><a name="293"></a>293</td><td></td><td></td><td></td><td></td><td class="s">note that UTF-8-encoded strings do not contain any characters outside the</td></tr>
+<tr><td class="h"><a name="294"></a>294</td><td></td><td></td><td></td><td></td><td class="s">range C<0..255>, they are thus useful for bytewise/binary I/O. In future</td></tr>
+<tr><td class="h"><a name="295"></a>295</td><td></td><td></td><td></td><td></td><td class="s">versions, enabling this option might enable autodetection of the UTF-16</td></tr>
+<tr><td class="h"><a name="296"></a>296</td><td></td><td></td><td></td><td></td><td class="s">and UTF-32 encoding families, as described in RFC4627.</td></tr>
+<tr><td class="h"><a name="297"></a>297</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="298"></a>298</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will return the JSON</td></tr>
+<tr><td class="h"><a name="299"></a>299</td><td></td><td></td><td></td><td></td><td class="s">string as a (non-encoded) Unicode string, while C<decode> expects thus a</td></tr>
+<tr><td class="h"><a name="300"></a>300</td><td></td><td></td><td></td><td></td><td class="s">Unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs</td></tr>
+<tr><td class="h"><a name="301"></a>301</td><td></td><td></td><td></td><td></td><td class="s">to be done yourself, e.g. using the Encode module.</td></tr>
+<tr><td class="h"><a name="302"></a>302</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="303"></a>303</td><td></td><td></td><td></td><td></td><td class="s">See also the section I<ENCODING/CODESET FLAG NOTES> later in this</td></tr>
+<tr><td class="h"><a name="304"></a>304</td><td></td><td></td><td></td><td></td><td class="s">document.</td></tr>
+<tr><td class="h"><a name="305"></a>305</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="306"></a>306</td><td></td><td></td><td></td><td></td><td class="s">Example, output UTF-16BE-encoded JSON:</td></tr>
+<tr><td class="h"><a name="307"></a>307</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="308"></a>308</td><td></td><td></td><td></td><td></td><td class="s"> use Encode;</td></tr>
+<tr><td class="h"><a name="309"></a>309</td><td></td><td></td><td></td><td></td><td class="s"> $jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object);</td></tr>
+<tr><td class="h"><a name="310"></a>310</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="311"></a>311</td><td></td><td></td><td></td><td></td><td class="s">Example, decode UTF-32LE-encoded JSON:</td></tr>
+<tr><td class="h"><a name="312"></a>312</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="313"></a>313</td><td></td><td></td><td></td><td></td><td class="s"> use Encode;</td></tr>
+<tr><td class="h"><a name="314"></a>314</td><td></td><td></td><td></td><td></td><td class="s"> $object = JSON::XS->new->decode (decode "UTF-32LE", $jsontext);</td></tr>
+<tr><td class="h"><a name="315"></a>315</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="316"></a>316</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->pretty ([$enable])</td></tr>
+<tr><td class="h"><a name="317"></a>317</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="318"></a>318</td><td></td><td></td><td></td><td></td><td class="s">This enables (or disables) all of the C<indent>, C<space_before> and</td></tr>
+<tr><td class="h"><a name="319"></a>319</td><td></td><td></td><td></td><td></td><td class="s">C<space_after> (and in the future possibly more) flags in one call to</td></tr>
+<tr><td class="h"><a name="320"></a>320</td><td></td><td></td><td></td><td></td><td class="s">generate the most readable (or most compact) form possible.</td></tr>
+<tr><td class="h"><a name="321"></a>321</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="322"></a>322</td><td></td><td></td><td></td><td></td><td class="s">Example, pretty-print some simple structure:</td></tr>
+<tr><td class="h"><a name="323"></a>323</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="324"></a>324</td><td></td><td></td><td></td><td></td><td class="s"> my $json = JSON::XS->new->pretty(1)->encode ({a => [1,2]})</td></tr>
+<tr><td class="h"><a name="325"></a>325</td><td></td><td></td><td></td><td></td><td class="s"> =></td></tr>
+<tr><td class="h"><a name="326"></a>326</td><td></td><td></td><td></td><td></td><td class="s"> {</td></tr>
+<tr><td class="h"><a name="327"></a>327</td><td></td><td></td><td></td><td></td><td class="s"> "a" : [</td></tr>
+<tr><td class="h"><a name="328"></a>328</td><td></td><td></td><td></td><td></td><td class="s"> 1,</td></tr>
+<tr><td class="h"><a name="329"></a>329</td><td></td><td></td><td></td><td></td><td class="s"> 2</td></tr>
+<tr><td class="h"><a name="330"></a>330</td><td></td><td></td><td></td><td></td><td class="s"> ]</td></tr>
+<tr><td class="h"><a name="331"></a>331</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="332"></a>332</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="333"></a>333</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->indent ([$enable])</td></tr>
+<tr><td class="h"><a name="334"></a>334</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="335"></a>335</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_indent</td></tr>
+<tr><td class="h"><a name="336"></a>336</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="337"></a>337</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will use a multiline</td></tr>
+<tr><td class="h"><a name="338"></a>338</td><td></td><td></td><td></td><td></td><td class="s">format as output, putting every array member or object/hash key-value pair</td></tr>
+<tr><td class="h"><a name="339"></a>339</td><td></td><td></td><td></td><td></td><td class="s">into its own line, indenting them properly.</td></tr>
+<tr><td class="h"><a name="340"></a>340</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="341"></a>341</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, no newlines or indenting will be produced, and the</td></tr>
+<tr><td class="h"><a name="342"></a>342</td><td></td><td></td><td></td><td></td><td class="s">resulting JSON text is guaranteed not to contain any C<newlines>.</td></tr>
+<tr><td class="h"><a name="343"></a>343</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="344"></a>344</td><td></td><td></td><td></td><td></td><td class="s">This setting has no effect when decoding JSON texts.</td></tr>
+<tr><td class="h"><a name="345"></a>345</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="346"></a>346</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->space_before ([$enable])</td></tr>
+<tr><td class="h"><a name="347"></a>347</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="348"></a>348</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_space_before</td></tr>
+<tr><td class="h"><a name="349"></a>349</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="350"></a>350</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will add an extra</td></tr>
+<tr><td class="h"><a name="351"></a>351</td><td></td><td></td><td></td><td></td><td class="s">optional space before the C<:> separating keys from values in JSON objects.</td></tr>
+<tr><td class="h"><a name="352"></a>352</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="353"></a>353</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will not add any extra</td></tr>
+<tr><td class="h"><a name="354"></a>354</td><td></td><td></td><td></td><td></td><td class="s">space at those places.</td></tr>
+<tr><td class="h"><a name="355"></a>355</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="356"></a>356</td><td></td><td></td><td></td><td></td><td class="s">This setting has no effect when decoding JSON texts. You will also</td></tr>
+<tr><td class="h"><a name="357"></a>357</td><td></td><td></td><td></td><td></td><td class="s">most likely combine this setting with C<space_after>.</td></tr>
+<tr><td class="h"><a name="358"></a>358</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="359"></a>359</td><td></td><td></td><td></td><td></td><td class="s">Example, space_before enabled, space_after and indent disabled:</td></tr>
+<tr><td class="h"><a name="360"></a>360</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="361"></a>361</td><td></td><td></td><td></td><td></td><td class="s"> {"key" :"value"}</td></tr>
+<tr><td class="h"><a name="362"></a>362</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="363"></a>363</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->space_after ([$enable])</td></tr>
+<tr><td class="h"><a name="364"></a>364</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="365"></a>365</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_space_after</td></tr>
+<tr><td class="h"><a name="366"></a>366</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="367"></a>367</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will add an extra</td></tr>
+<tr><td class="h"><a name="368"></a>368</td><td></td><td></td><td></td><td></td><td class="s">optional space after the C<:> separating keys from values in JSON objects</td></tr>
+<tr><td class="h"><a name="369"></a>369</td><td></td><td></td><td></td><td></td><td class="s">and extra whitespace after the C<,> separating key-value pairs and array</td></tr>
+<tr><td class="h"><a name="370"></a>370</td><td></td><td></td><td></td><td></td><td class="s">members.</td></tr>
+<tr><td class="h"><a name="371"></a>371</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="372"></a>372</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will not add any extra</td></tr>
+<tr><td class="h"><a name="373"></a>373</td><td></td><td></td><td></td><td></td><td class="s">space at those places.</td></tr>
+<tr><td class="h"><a name="374"></a>374</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="375"></a>375</td><td></td><td></td><td></td><td></td><td class="s">This setting has no effect when decoding JSON texts.</td></tr>
+<tr><td class="h"><a name="376"></a>376</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="377"></a>377</td><td></td><td></td><td></td><td></td><td class="s">Example, space_before and indent disabled, space_after enabled:</td></tr>
+<tr><td class="h"><a name="378"></a>378</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="379"></a>379</td><td></td><td></td><td></td><td></td><td class="s"> {"key": "value"}</td></tr>
+<tr><td class="h"><a name="380"></a>380</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="381"></a>381</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->relaxed ([$enable])</td></tr>
+<tr><td class="h"><a name="382"></a>382</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="383"></a>383</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_relaxed</td></tr>
+<tr><td class="h"><a name="384"></a>384</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="385"></a>385</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then C<decode> will accept some</td></tr>
+<tr><td class="h"><a name="386"></a>386</td><td></td><td></td><td></td><td></td><td class="s">extensions to normal JSON syntax (see below). C<encode> will not be</td></tr>
+<tr><td class="h"><a name="387"></a>387</td><td></td><td></td><td></td><td></td><td class="s">affected in anyway. I<Be aware that this option makes you accept invalid</td></tr>
+<tr><td class="h"><a name="388"></a>388</td><td></td><td></td><td></td><td></td><td class="s">JSON texts as if they were valid!>. I suggest only to use this option to</td></tr>
+<tr><td class="h"><a name="389"></a>389</td><td></td><td></td><td></td><td></td><td class="s">parse application-specific files written by humans (configuration files,</td></tr>
+<tr><td class="h"><a name="390"></a>390</td><td></td><td></td><td></td><td></td><td class="s">resource files etc.)</td></tr>
+<tr><td class="h"><a name="391"></a>391</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="392"></a>392</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false (the default), then C<decode> will only accept</td></tr>
+<tr><td class="h"><a name="393"></a>393</td><td></td><td></td><td></td><td></td><td class="s">valid JSON texts.</td></tr>
+<tr><td class="h"><a name="394"></a>394</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="395"></a>395</td><td></td><td></td><td></td><td></td><td class="s">Currently accepted extensions are:</td></tr>
+<tr><td class="h"><a name="396"></a>396</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="397"></a>397</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="398"></a>398</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="399"></a>399</td><td></td><td></td><td></td><td></td><td class="s">=item * list items can have an end-comma</td></tr>
+<tr><td class="h"><a name="400"></a>400</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="401"></a>401</td><td></td><td></td><td></td><td></td><td class="s">JSON I<separates> array elements and key-value pairs with commas. This</td></tr>
+<tr><td class="h"><a name="402"></a>402</td><td></td><td></td><td></td><td></td><td class="s">can be annoying if you write JSON texts manually and want to be able to</td></tr>
+<tr><td class="h"><a name="403"></a>403</td><td></td><td></td><td></td><td></td><td class="s">quickly append elements, so this extension accepts comma at the end of</td></tr>
+<tr><td class="h"><a name="404"></a>404</td><td></td><td></td><td></td><td></td><td class="s">such items not just between them:</td></tr>
+<tr><td class="h"><a name="405"></a>405</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="406"></a>406</td><td></td><td></td><td></td><td></td><td class="s"> [</td></tr>
+<tr><td class="h"><a name="407"></a>407</td><td></td><td></td><td></td><td></td><td class="s"> 1,</td></tr>
+<tr><td class="h"><a name="408"></a>408</td><td></td><td></td><td></td><td></td><td class="s"> 2, <- this comma not normally allowed</td></tr>
+<tr><td class="h"><a name="409"></a>409</td><td></td><td></td><td></td><td></td><td class="s"> ]</td></tr>
+<tr><td class="h"><a name="410"></a>410</td><td></td><td></td><td></td><td></td><td class="s"> {</td></tr>
+<tr><td class="h"><a name="411"></a>411</td><td></td><td></td><td></td><td></td><td class="s"> "k1": "v1",</td></tr>
+<tr><td class="h"><a name="412"></a>412</td><td></td><td></td><td></td><td></td><td class="s"> "k2": "v2", <- this comma not normally allowed</td></tr>
+<tr><td class="h"><a name="413"></a>413</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="414"></a>414</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="415"></a>415</td><td></td><td></td><td></td><td></td><td class="s">=item * shell-style '#'-comments</td></tr>
+<tr><td class="h"><a name="416"></a>416</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="417"></a>417</td><td></td><td></td><td></td><td></td><td class="s">Whenever JSON allows whitespace, shell-style comments are additionally</td></tr>
+<tr><td class="h"><a name="418"></a>418</td><td></td><td></td><td></td><td></td><td class="s">allowed. They are terminated by the first carriage-return or line-feed</td></tr>
+<tr><td class="h"><a name="419"></a>419</td><td></td><td></td><td></td><td></td><td class="s">character, after which more white-space and comments are allowed.</td></tr>
+<tr><td class="h"><a name="420"></a>420</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="421"></a>421</td><td></td><td></td><td></td><td></td><td class="s"> [</td></tr>
+<tr><td class="h"><a name="422"></a>422</td><td></td><td></td><td></td><td></td><td class="s"> 1, # this comment not allowed in JSON</td></tr>
+<tr><td class="h"><a name="423"></a>423</td><td></td><td></td><td></td><td></td><td class="s"> # neither this one...</td></tr>
+<tr><td class="h"><a name="424"></a>424</td><td></td><td></td><td></td><td></td><td class="s"> ]</td></tr>
+<tr><td class="h"><a name="425"></a>425</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="426"></a>426</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="427"></a>427</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="428"></a>428</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->canonical ([$enable])</td></tr>
+<tr><td class="h"><a name="429"></a>429</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="430"></a>430</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_canonical</td></tr>
+<tr><td class="h"><a name="431"></a>431</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="432"></a>432</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will output JSON objects</td></tr>
+<tr><td class="h"><a name="433"></a>433</td><td></td><td></td><td></td><td></td><td class="s">by sorting their keys. This is adding a comparatively high overhead.</td></tr>
+<tr><td class="h"><a name="434"></a>434</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="435"></a>435</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will output key-value</td></tr>
+<tr><td class="h"><a name="436"></a>436</td><td></td><td></td><td></td><td></td><td class="s">pairs in the order Perl stores them (which will likely change between runs</td></tr>
+<tr><td class="h"><a name="437"></a>437</td><td></td><td></td><td></td><td></td><td class="s">of the same script).</td></tr>
+<tr><td class="h"><a name="438"></a>438</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="439"></a>439</td><td></td><td></td><td></td><td></td><td class="s">This option is useful if you want the same data structure to be encoded as</td></tr>
+<tr><td class="h"><a name="440"></a>440</td><td></td><td></td><td></td><td></td><td class="s">the same JSON text (given the same overall settings). If it is disabled,</td></tr>
+<tr><td class="h"><a name="441"></a>441</td><td></td><td></td><td></td><td></td><td class="s">the same hash might be encoded differently even if contains the same data,</td></tr>
+<tr><td class="h"><a name="442"></a>442</td><td></td><td></td><td></td><td></td><td class="s">as key-value pairs have no inherent ordering in Perl.</td></tr>
+<tr><td class="h"><a name="443"></a>443</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="444"></a>444</td><td></td><td></td><td></td><td></td><td class="s">This setting has no effect when decoding JSON texts.</td></tr>
+<tr><td class="h"><a name="445"></a>445</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="446"></a>446</td><td></td><td></td><td></td><td></td><td class="s">This setting has currently no effect on tied hashes.</td></tr>
+<tr><td class="h"><a name="447"></a>447</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="448"></a>448</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->allow_nonref ([$enable])</td></tr>
+<tr><td class="h"><a name="449"></a>449</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="450"></a>450</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_allow_nonref</td></tr>
+<tr><td class="h"><a name="451"></a>451</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="452"></a>452</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method can convert a</td></tr>
+<tr><td class="h"><a name="453"></a>453</td><td></td><td></td><td></td><td></td><td class="s">non-reference into its corresponding string, number or null JSON value,</td></tr>
+<tr><td class="h"><a name="454"></a>454</td><td></td><td></td><td></td><td></td><td class="s">which is an extension to RFC4627. Likewise, C<decode> will accept those JSON</td></tr>
+<tr><td class="h"><a name="455"></a>455</td><td></td><td></td><td></td><td></td><td class="s">values instead of croaking.</td></tr>
+<tr><td class="h"><a name="456"></a>456</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="457"></a>457</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<encode> method will croak if it isn't</td></tr>
+<tr><td class="h"><a name="458"></a>458</td><td></td><td></td><td></td><td></td><td class="s">passed an arrayref or hashref, as JSON texts must either be an object</td></tr>
+<tr><td class="h"><a name="459"></a>459</td><td></td><td></td><td></td><td></td><td class="s">or array. Likewise, C<decode> will croak if given something that is not a</td></tr>
+<tr><td class="h"><a name="460"></a>460</td><td></td><td></td><td></td><td></td><td class="s">JSON object or array.</td></tr>
+<tr><td class="h"><a name="461"></a>461</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="462"></a>462</td><td></td><td></td><td></td><td></td><td class="s">Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>,</td></tr>
+<tr><td class="h"><a name="463"></a>463</td><td></td><td></td><td></td><td></td><td class="s">resulting in an invalid JSON text:</td></tr>
+<tr><td class="h"><a name="464"></a>464</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="465"></a>465</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS->new->allow_nonref->encode ("Hello, World!")</td></tr>
+<tr><td class="h"><a name="466"></a>466</td><td></td><td></td><td></td><td></td><td class="s"> => "Hello, World!"</td></tr>
+<tr><td class="h"><a name="467"></a>467</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="468"></a>468</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->allow_unknown ([$enable])</td></tr>
+<tr><td class="h"><a name="469"></a>469</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="470"></a>470</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_allow_unknown</td></tr>
+<tr><td class="h"><a name="471"></a>471</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="472"></a>472</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then C<encode> will I<not> throw an</td></tr>
+<tr><td class="h"><a name="473"></a>473</td><td></td><td></td><td></td><td></td><td class="s">exception when it encounters values it cannot represent in JSON (for</td></tr>
+<tr><td class="h"><a name="474"></a>474</td><td></td><td></td><td></td><td></td><td class="s">example, filehandles) but instead will encode a JSON C<null> value. Note</td></tr>
+<tr><td class="h"><a name="475"></a>475</td><td></td><td></td><td></td><td></td><td class="s">that blessed objects are not included here and are handled separately by</td></tr>
+<tr><td class="h"><a name="476"></a>476</td><td></td><td></td><td></td><td></td><td class="s">c<allow_nonref>.</td></tr>
+<tr><td class="h"><a name="477"></a>477</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="478"></a>478</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false (the default), then C<encode> will throw an</td></tr>
+<tr><td class="h"><a name="479"></a>479</td><td></td><td></td><td></td><td></td><td class="s">exception when it encounters anything it cannot encode as JSON.</td></tr>
+<tr><td class="h"><a name="480"></a>480</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="481"></a>481</td><td></td><td></td><td></td><td></td><td class="s">This option does not affect C<decode> in any way, and it is recommended to</td></tr>
+<tr><td class="h"><a name="482"></a>482</td><td></td><td></td><td></td><td></td><td class="s">leave it off unless you know your communications partner.</td></tr>
+<tr><td class="h"><a name="483"></a>483</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="484"></a>484</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->allow_blessed ([$enable])</td></tr>
+<tr><td class="h"><a name="485"></a>485</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="486"></a>486</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_allow_blessed</td></tr>
+<tr><td class="h"><a name="487"></a>487</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="488"></a>488</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then the C<encode> method will not</td></tr>
+<tr><td class="h"><a name="489"></a>489</td><td></td><td></td><td></td><td></td><td class="s">barf when it encounters a blessed reference. Instead, the value of the</td></tr>
+<tr><td class="h"><a name="490"></a>490</td><td></td><td></td><td></td><td></td><td class="s">B<convert_blessed> option will decide whether C<null> (C<convert_blessed></td></tr>
+<tr><td class="h"><a name="491"></a>491</td><td></td><td></td><td></td><td></td><td class="s">disabled or no C<TO_JSON> method found) or a representation of the</td></tr>
+<tr><td class="h"><a name="492"></a>492</td><td></td><td></td><td></td><td></td><td class="s">object (C<convert_blessed> enabled and C<TO_JSON> method found) is being</td></tr>
+<tr><td class="h"><a name="493"></a>493</td><td></td><td></td><td></td><td></td><td class="s">encoded. Has no effect on C<decode>.</td></tr>
+<tr><td class="h"><a name="494"></a>494</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="495"></a>495</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false (the default), then C<encode> will throw an</td></tr>
+<tr><td class="h"><a name="496"></a>496</td><td></td><td></td><td></td><td></td><td class="s">exception when it encounters a blessed object.</td></tr>
+<tr><td class="h"><a name="497"></a>497</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="498"></a>498</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->convert_blessed ([$enable])</td></tr>
+<tr><td class="h"><a name="499"></a>499</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="500"></a>500</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_convert_blessed</td></tr>
+<tr><td class="h"><a name="501"></a>501</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="502"></a>502</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), then C<encode>, upon encountering a</td></tr>
+<tr><td class="h"><a name="503"></a>503</td><td></td><td></td><td></td><td></td><td class="s">blessed object, will check for the availability of the C<TO_JSON> method</td></tr>
+<tr><td class="h"><a name="504"></a>504</td><td></td><td></td><td></td><td></td><td class="s">on the object's class. If found, it will be called in scalar context</td></tr>
+<tr><td class="h"><a name="505"></a>505</td><td></td><td></td><td></td><td></td><td class="s">and the resulting scalar will be encoded instead of the object. If no</td></tr>
+<tr><td class="h"><a name="506"></a>506</td><td></td><td></td><td></td><td></td><td class="s">C<TO_JSON> method is found, the value of C<allow_blessed> will decide what</td></tr>
+<tr><td class="h"><a name="507"></a>507</td><td></td><td></td><td></td><td></td><td class="s">to do.</td></tr>
+<tr><td class="h"><a name="508"></a>508</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="509"></a>509</td><td></td><td></td><td></td><td></td><td class="s">The C<TO_JSON> method may safely call die if it wants. If C<TO_JSON></td></tr>
+<tr><td class="h"><a name="510"></a>510</td><td></td><td></td><td></td><td></td><td class="s">returns other blessed objects, those will be handled in the same</td></tr>
+<tr><td class="h"><a name="511"></a>511</td><td></td><td></td><td></td><td></td><td class="s">way. C<TO_JSON> must take care of not causing an endless recursion cycle</td></tr>
+<tr><td class="h"><a name="512"></a>512</td><td></td><td></td><td></td><td></td><td class="s">(== crash) in this case. The name of C<TO_JSON> was chosen because other</td></tr>
+<tr><td class="h"><a name="513"></a>513</td><td></td><td></td><td></td><td></td><td class="s">methods called by the Perl core (== not by the user of the object) are</td></tr>
+<tr><td class="h"><a name="514"></a>514</td><td></td><td></td><td></td><td></td><td class="s">usually in upper case letters and to avoid collisions with any C<to_json></td></tr>
+<tr><td class="h"><a name="515"></a>515</td><td></td><td></td><td></td><td></td><td class="s">function or method.</td></tr>
+<tr><td class="h"><a name="516"></a>516</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="517"></a>517</td><td></td><td></td><td></td><td></td><td class="s">This setting does not yet influence C<decode> in any way, but in the</td></tr>
+<tr><td class="h"><a name="518"></a>518</td><td></td><td></td><td></td><td></td><td class="s">future, global hooks might get installed that influence C<decode> and are</td></tr>
+<tr><td class="h"><a name="519"></a>519</td><td></td><td></td><td></td><td></td><td class="s">enabled by this setting.</td></tr>
+<tr><td class="h"><a name="520"></a>520</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="521"></a>521</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the C<allow_blessed> setting will decide what</td></tr>
+<tr><td class="h"><a name="522"></a>522</td><td></td><td></td><td></td><td></td><td class="s">to do when a blessed object is found.</td></tr>
+<tr><td class="h"><a name="523"></a>523</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="524"></a>524</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->filter_json_object ([$coderef->($hashref)])</td></tr>
+<tr><td class="h"><a name="525"></a>525</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="526"></a>526</td><td></td><td></td><td></td><td></td><td class="s">When C<$coderef> is specified, it will be called from C<decode> each</td></tr>
+<tr><td class="h"><a name="527"></a>527</td><td></td><td></td><td></td><td></td><td class="s">time it decodes a JSON object. The only argument is a reference to the</td></tr>
+<tr><td class="h"><a name="528"></a>528</td><td></td><td></td><td></td><td></td><td class="s">newly-created hash. If the code references returns a single scalar (which</td></tr>
+<tr><td class="h"><a name="529"></a>529</td><td></td><td></td><td></td><td></td><td class="s">need not be a reference), this value (i.e. a copy of that scalar to avoid</td></tr>
+<tr><td class="h"><a name="530"></a>530</td><td></td><td></td><td></td><td></td><td class="s">aliasing) is inserted into the deserialised data structure. If it returns</td></tr>
+<tr><td class="h"><a name="531"></a>531</td><td></td><td></td><td></td><td></td><td class="s">an empty list (NOTE: I<not> C<undef>, which is a valid scalar), the</td></tr>
+<tr><td class="h"><a name="532"></a>532</td><td></td><td></td><td></td><td></td><td class="s">original deserialised hash will be inserted. This setting can slow down</td></tr>
+<tr><td class="h"><a name="533"></a>533</td><td></td><td></td><td></td><td></td><td class="s">decoding considerably.</td></tr>
+<tr><td class="h"><a name="534"></a>534</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="535"></a>535</td><td></td><td></td><td></td><td></td><td class="s">When C<$coderef> is omitted or undefined, any existing callback will</td></tr>
+<tr><td class="h"><a name="536"></a>536</td><td></td><td></td><td></td><td></td><td class="s">be removed and C<decode> will not change the deserialised hash in any</td></tr>
+<tr><td class="h"><a name="537"></a>537</td><td></td><td></td><td></td><td></td><td class="s">way.</td></tr>
+<tr><td class="h"><a name="538"></a>538</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="539"></a>539</td><td></td><td></td><td></td><td></td><td class="s">Example, convert all JSON objects into the integer 5:</td></tr>
+<tr><td class="h"><a name="540"></a>540</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="541"></a>541</td><td></td><td></td><td></td><td></td><td class="s"> my $js = JSON::XS->new->filter_json_object (sub { 5 });</td></tr>
+<tr><td class="h"><a name="542"></a>542</td><td></td><td></td><td></td><td></td><td class="s"> # returns [5]</td></tr>
+<tr><td class="h"><a name="543"></a>543</td><td></td><td></td><td></td><td></td><td class="s"> $js->decode ('[{}]')</td></tr>
+<tr><td class="h"><a name="544"></a>544</td><td></td><td></td><td></td><td></td><td class="s"> # throw an exception because allow_nonref is not enabled</td></tr>
+<tr><td class="h"><a name="545"></a>545</td><td></td><td></td><td></td><td></td><td class="s"> # so a lone 5 is not allowed.</td></tr>
+<tr><td class="h"><a name="546"></a>546</td><td></td><td></td><td></td><td></td><td class="s"> $js->decode ('{"a":1, "b":2}');</td></tr>
+<tr><td class="h"><a name="547"></a>547</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="548"></a>548</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->filter_json_single_key_object ($key [=> $coderef->($value)])</td></tr>
+<tr><td class="h"><a name="549"></a>549</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="550"></a>550</td><td></td><td></td><td></td><td></td><td class="s">Works remotely similar to C<filter_json_object>, but is only called for</td></tr>
+<tr><td class="h"><a name="551"></a>551</td><td></td><td></td><td></td><td></td><td class="s">JSON objects having a single key named C<$key>.</td></tr>
+<tr><td class="h"><a name="552"></a>552</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="553"></a>553</td><td></td><td></td><td></td><td></td><td class="s">This C<$coderef> is called before the one specified via</td></tr>
+<tr><td class="h"><a name="554"></a>554</td><td></td><td></td><td></td><td></td><td class="s">C<filter_json_object>, if any. It gets passed the single value in the JSON</td></tr>
+<tr><td class="h"><a name="555"></a>555</td><td></td><td></td><td></td><td></td><td class="s">object. If it returns a single value, it will be inserted into the data</td></tr>
+<tr><td class="h"><a name="556"></a>556</td><td></td><td></td><td></td><td></td><td class="s">structure. If it returns nothing (not even C<undef> but the empty list),</td></tr>
+<tr><td class="h"><a name="557"></a>557</td><td></td><td></td><td></td><td></td><td class="s">the callback from C<filter_json_object> will be called next, as if no</td></tr>
+<tr><td class="h"><a name="558"></a>558</td><td></td><td></td><td></td><td></td><td class="s">single-key callback were specified.</td></tr>
+<tr><td class="h"><a name="559"></a>559</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="560"></a>560</td><td></td><td></td><td></td><td></td><td class="s">If C<$coderef> is omitted or undefined, the corresponding callback will be</td></tr>
+<tr><td class="h"><a name="561"></a>561</td><td></td><td></td><td></td><td></td><td class="s">disabled. There can only ever be one callback for a given key.</td></tr>
+<tr><td class="h"><a name="562"></a>562</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="563"></a>563</td><td></td><td></td><td></td><td></td><td class="s">As this callback gets called less often then the C<filter_json_object></td></tr>
+<tr><td class="h"><a name="564"></a>564</td><td></td><td></td><td></td><td></td><td class="s">one, decoding speed will not usually suffer as much. Therefore, single-key</td></tr>
+<tr><td class="h"><a name="565"></a>565</td><td></td><td></td><td></td><td></td><td class="s">objects make excellent targets to serialise Perl objects into, especially</td></tr>
+<tr><td class="h"><a name="566"></a>566</td><td></td><td></td><td></td><td></td><td class="s">as single-key JSON objects are as close to the type-tagged value concept</td></tr>
+<tr><td class="h"><a name="567"></a>567</td><td></td><td></td><td></td><td></td><td class="s">as JSON gets (it's basically an ID/VALUE tuple). Of course, JSON does not</td></tr>
+<tr><td class="h"><a name="568"></a>568</td><td></td><td></td><td></td><td></td><td class="s">support this in any way, so you need to make sure your data never looks</td></tr>
+<tr><td class="h"><a name="569"></a>569</td><td></td><td></td><td></td><td></td><td class="s">like a serialised Perl hash.</td></tr>
+<tr><td class="h"><a name="570"></a>570</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="571"></a>571</td><td></td><td></td><td></td><td></td><td class="s">Typical names for the single object key are C<__class_whatever__>, or</td></tr>
+<tr><td class="h"><a name="572"></a>572</td><td></td><td></td><td></td><td></td><td class="s">C<$__dollars_are_rarely_used__$> or C<}ugly_brace_placement>, or even</td></tr>
+<tr><td class="h"><a name="573"></a>573</td><td></td><td></td><td></td><td></td><td class="s">things like C<__class_md5sum(classname)__>, to reduce the risk of clashing</td></tr>
+<tr><td class="h"><a name="574"></a>574</td><td></td><td></td><td></td><td></td><td class="s">with real hashes.</td></tr>
+<tr><td class="h"><a name="575"></a>575</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="576"></a>576</td><td></td><td></td><td></td><td></td><td class="s">Example, decode JSON objects of the form C<< { "__widget__" => <id> } >></td></tr>
+<tr><td class="h"><a name="577"></a>577</td><td></td><td></td><td></td><td></td><td class="s">into the corresponding C<< $WIDGET{<id>} >> object:</td></tr>
+<tr><td class="h"><a name="578"></a>578</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="579"></a>579</td><td></td><td></td><td></td><td></td><td class="s"> # return whatever is in $WIDGET{5}:</td></tr>
+<tr><td class="h"><a name="580"></a>580</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS</td></tr>
+<tr><td class="h"><a name="581"></a>581</td><td></td><td></td><td></td><td></td><td class="s"> ->new</td></tr>
+<tr><td class="h"><a name="582"></a>582</td><td></td><td></td><td></td><td></td><td class="s"> ->filter_json_single_key_object (__widget__ => sub {</td></tr>
+<tr><td class="h"><a name="583"></a>583</td><td></td><td></td><td></td><td></td><td class="s"> $WIDGET{ $_[0] }</td></tr>
+<tr><td class="h"><a name="584"></a>584</td><td></td><td></td><td></td><td></td><td class="s"> })</td></tr>
+<tr><td class="h"><a name="585"></a>585</td><td></td><td></td><td></td><td></td><td class="s"> ->decode ('{"__widget__": 5')</td></tr>
+<tr><td class="h"><a name="586"></a>586</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="587"></a>587</td><td></td><td></td><td></td><td></td><td class="s"> # this can be used with a TO_JSON method in some "widget" class</td></tr>
+<tr><td class="h"><a name="588"></a>588</td><td></td><td></td><td></td><td></td><td class="s"> # for serialisation to json:</td></tr>
+<tr><td class="h"><a name="589"></a>589</td><td></td><td></td><td></td><td></td><td class="s"> sub WidgetBase::TO_JSON {</td></tr>
+<tr><td class="h"><a name="590"></a>590</td><td></td><td></td><td></td><td></td><td class="s"> my ($self) = @_;</td></tr>
+<tr><td class="h"><a name="591"></a>591</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="592"></a>592</td><td></td><td></td><td></td><td></td><td class="s"> unless ($self->{id}) {</td></tr>
+<tr><td class="h"><a name="593"></a>593</td><td></td><td></td><td></td><td></td><td class="s"> $self->{id} = ..get..some..id..;</td></tr>
+<tr><td class="h"><a name="594"></a>594</td><td></td><td></td><td></td><td></td><td class="s"> $WIDGET{$self->{id}} = $self;</td></tr>
+<tr><td class="h"><a name="595"></a>595</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="596"></a>596</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="597"></a>597</td><td></td><td></td><td></td><td></td><td class="s"> { __widget__ => $self->{id} }</td></tr>
+<tr><td class="h"><a name="598"></a>598</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="599"></a>599</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="600"></a>600</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->shrink ([$enable])</td></tr>
+<tr><td class="h"><a name="601"></a>601</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="602"></a>602</td><td></td><td></td><td></td><td></td><td class="s">=item $enabled = $json->get_shrink</td></tr>
+<tr><td class="h"><a name="603"></a>603</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="604"></a>604</td><td></td><td></td><td></td><td></td><td class="s">Perl usually over-allocates memory a bit when allocating space for</td></tr>
+<tr><td class="h"><a name="605"></a>605</td><td></td><td></td><td></td><td></td><td class="s">strings. This flag optionally resizes strings generated by either</td></tr>
+<tr><td class="h"><a name="606"></a>606</td><td></td><td></td><td></td><td></td><td class="s">C<encode> or C<decode> to their minimum size possible. This can save</td></tr>
+<tr><td class="h"><a name="607"></a>607</td><td></td><td></td><td></td><td></td><td class="s">memory when your JSON texts are either very very long or you have many</td></tr>
+<tr><td class="h"><a name="608"></a>608</td><td></td><td></td><td></td><td></td><td class="s">short strings. It will also try to downgrade any strings to octet-form</td></tr>
+<tr><td class="h"><a name="609"></a>609</td><td></td><td></td><td></td><td></td><td class="s">if possible: perl stores strings internally either in an encoding called</td></tr>
+<tr><td class="h"><a name="610"></a>610</td><td></td><td></td><td></td><td></td><td class="s">UTF-X or in octet-form. The latter cannot store everything but uses less</td></tr>
+<tr><td class="h"><a name="611"></a>611</td><td></td><td></td><td></td><td></td><td class="s">space in general (and some buggy Perl or C code might even rely on that</td></tr>
+<tr><td class="h"><a name="612"></a>612</td><td></td><td></td><td></td><td></td><td class="s">internal representation being used).</td></tr>
+<tr><td class="h"><a name="613"></a>613</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="614"></a>614</td><td></td><td></td><td></td><td></td><td class="s">The actual definition of what shrink does might change in future versions,</td></tr>
+<tr><td class="h"><a name="615"></a>615</td><td></td><td></td><td></td><td></td><td class="s">but it will always try to save space at the expense of time.</td></tr>
+<tr><td class="h"><a name="616"></a>616</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="617"></a>617</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is true (or missing), the string returned by C<encode> will</td></tr>
+<tr><td class="h"><a name="618"></a>618</td><td></td><td></td><td></td><td></td><td class="s">be shrunk-to-fit, while all strings generated by C<decode> will also be</td></tr>
+<tr><td class="h"><a name="619"></a>619</td><td></td><td></td><td></td><td></td><td class="s">shrunk-to-fit.</td></tr>
+<tr><td class="h"><a name="620"></a>620</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="621"></a>621</td><td></td><td></td><td></td><td></td><td class="s">If C<$enable> is false, then the normal perl allocation algorithms are used.</td></tr>
+<tr><td class="h"><a name="622"></a>622</td><td></td><td></td><td></td><td></td><td class="s">If you work with your data, then this is likely to be faster.</td></tr>
+<tr><td class="h"><a name="623"></a>623</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="624"></a>624</td><td></td><td></td><td></td><td></td><td class="s">In the future, this setting might control other things, such as converting</td></tr>
+<tr><td class="h"><a name="625"></a>625</td><td></td><td></td><td></td><td></td><td class="s">strings that look like integers or floats into integers or floats</td></tr>
+<tr><td class="h"><a name="626"></a>626</td><td></td><td></td><td></td><td></td><td class="s">internally (there is no difference on the Perl level), saving space.</td></tr>
+<tr><td class="h"><a name="627"></a>627</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="628"></a>628</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->max_depth ([$maximum_nesting_depth])</td></tr>
+<tr><td class="h"><a name="629"></a>629</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="630"></a>630</td><td></td><td></td><td></td><td></td><td class="s">=item $max_depth = $json->get_max_depth</td></tr>
+<tr><td class="h"><a name="631"></a>631</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="632"></a>632</td><td></td><td></td><td></td><td></td><td class="s">Sets the maximum nesting level (default C<512>) accepted while encoding</td></tr>
+<tr><td class="h"><a name="633"></a>633</td><td></td><td></td><td></td><td></td><td class="s">or decoding. If a higher nesting level is detected in JSON text or a Perl</td></tr>
+<tr><td class="h"><a name="634"></a>634</td><td></td><td></td><td></td><td></td><td class="s">data structure, then the encoder and decoder will stop and croak at that</td></tr>
+<tr><td class="h"><a name="635"></a>635</td><td></td><td></td><td></td><td></td><td class="s">point.</td></tr>
+<tr><td class="h"><a name="636"></a>636</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="637"></a>637</td><td></td><td></td><td></td><td></td><td class="s">Nesting level is defined by number of hash- or arrayrefs that the encoder</td></tr>
+<tr><td class="h"><a name="638"></a>638</td><td></td><td></td><td></td><td></td><td class="s">needs to traverse to reach a given point or the number of C<{> or C<[></td></tr>
+<tr><td class="h"><a name="639"></a>639</td><td></td><td></td><td></td><td></td><td class="s">characters without their matching closing parenthesis crossed to reach a</td></tr>
+<tr><td class="h"><a name="640"></a>640</td><td></td><td></td><td></td><td></td><td class="s">given character in a string.</td></tr>
+<tr><td class="h"><a name="641"></a>641</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="642"></a>642</td><td></td><td></td><td></td><td></td><td class="s">Setting the maximum depth to one disallows any nesting, so that ensures</td></tr>
+<tr><td class="h"><a name="643"></a>643</td><td></td><td></td><td></td><td></td><td class="s">that the object is only a single hash/object or array.</td></tr>
+<tr><td class="h"><a name="644"></a>644</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="645"></a>645</td><td></td><td></td><td></td><td></td><td class="s">If no argument is given, the highest possible setting will be used, which</td></tr>
+<tr><td class="h"><a name="646"></a>646</td><td></td><td></td><td></td><td></td><td class="s">is rarely useful.</td></tr>
+<tr><td class="h"><a name="647"></a>647</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="648"></a>648</td><td></td><td></td><td></td><td></td><td class="s">Note that nesting is implemented by recursion in C. The default value has</td></tr>
+<tr><td class="h"><a name="649"></a>649</td><td></td><td></td><td></td><td></td><td class="s">been chosen to be as large as typical operating systems allow without</td></tr>
+<tr><td class="h"><a name="650"></a>650</td><td></td><td></td><td></td><td></td><td class="s">crashing.</td></tr>
+<tr><td class="h"><a name="651"></a>651</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="652"></a>652</td><td></td><td></td><td></td><td></td><td class="s">See SECURITY CONSIDERATIONS, below, for more info on why this is useful.</td></tr>
+<tr><td class="h"><a name="653"></a>653</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="654"></a>654</td><td></td><td></td><td></td><td></td><td class="s">=item $json = $json->max_size ([$maximum_string_size])</td></tr>
+<tr><td class="h"><a name="655"></a>655</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="656"></a>656</td><td></td><td></td><td></td><td></td><td class="s">=item $max_size = $json->get_max_size</td></tr>
+<tr><td class="h"><a name="657"></a>657</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="658"></a>658</td><td></td><td></td><td></td><td></td><td class="s">Set the maximum length a JSON text may have (in bytes) where decoding is</td></tr>
+<tr><td class="h"><a name="659"></a>659</td><td></td><td></td><td></td><td></td><td class="s">being attempted. The default is C<0>, meaning no limit. When C<decode></td></tr>
+<tr><td class="h"><a name="660"></a>660</td><td></td><td></td><td></td><td></td><td class="s">is called on a string that is longer then this many bytes, it will not</td></tr>
+<tr><td class="h"><a name="661"></a>661</td><td></td><td></td><td></td><td></td><td class="s">attempt to decode the string but throw an exception. This setting has no</td></tr>
+<tr><td class="h"><a name="662"></a>662</td><td></td><td></td><td></td><td></td><td class="s">effect on C<encode> (yet).</td></tr>
+<tr><td class="h"><a name="663"></a>663</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="664"></a>664</td><td></td><td></td><td></td><td></td><td class="s">If no argument is given, the limit check will be deactivated (same as when</td></tr>
+<tr><td class="h"><a name="665"></a>665</td><td></td><td></td><td></td><td></td><td class="s">C<0> is specified).</td></tr>
+<tr><td class="h"><a name="666"></a>666</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="667"></a>667</td><td></td><td></td><td></td><td></td><td class="s">See SECURITY CONSIDERATIONS, below, for more info on why this is useful.</td></tr>
+<tr><td class="h"><a name="668"></a>668</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="669"></a>669</td><td></td><td></td><td></td><td></td><td class="s">=item $json_text = $json->encode ($perl_scalar)</td></tr>
+<tr><td class="h"><a name="670"></a>670</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="671"></a>671</td><td></td><td></td><td></td><td></td><td class="s">Converts the given Perl data structure (a simple scalar or a reference</td></tr>
+<tr><td class="h"><a name="672"></a>672</td><td></td><td></td><td></td><td></td><td class="s">to a hash or array) to its JSON representation. Simple scalars will be</td></tr>
+<tr><td class="h"><a name="673"></a>673</td><td></td><td></td><td></td><td></td><td class="s">converted into JSON string or number sequences, while references to arrays</td></tr>
+<tr><td class="h"><a name="674"></a>674</td><td></td><td></td><td></td><td></td><td class="s">become JSON arrays and references to hashes become JSON objects. Undefined</td></tr>
+<tr><td class="h"><a name="675"></a>675</td><td></td><td></td><td></td><td></td><td class="s">Perl values (e.g. C<undef>) become JSON C<null> values. Neither C<true></td></tr>
+<tr><td class="h"><a name="676"></a>676</td><td></td><td></td><td></td><td></td><td class="s">nor C<false> values will be generated.</td></tr>
+<tr><td class="h"><a name="677"></a>677</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="678"></a>678</td><td></td><td></td><td></td><td></td><td class="s">=item $perl_scalar = $json->decode ($json_text)</td></tr>
+<tr><td class="h"><a name="679"></a>679</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="680"></a>680</td><td></td><td></td><td></td><td></td><td class="s">The opposite of C<encode>: expects a JSON text and tries to parse it,</td></tr>
+<tr><td class="h"><a name="681"></a>681</td><td></td><td></td><td></td><td></td><td class="s">returning the resulting simple scalar or reference. Croaks on error.</td></tr>
+<tr><td class="h"><a name="682"></a>682</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="683"></a>683</td><td></td><td></td><td></td><td></td><td class="s">JSON numbers and strings become simple Perl scalars. JSON arrays become</td></tr>
+<tr><td class="h"><a name="684"></a>684</td><td></td><td></td><td></td><td></td><td class="s">Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes</td></tr>
+<tr><td class="h"><a name="685"></a>685</td><td></td><td></td><td></td><td></td><td class="s">C<1>, C<false> becomes C<0> and C<null> becomes C<undef>.</td></tr>
+<tr><td class="h"><a name="686"></a>686</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="687"></a>687</td><td></td><td></td><td></td><td></td><td class="s">=item ($perl_scalar, $characters) = $json->decode_prefix ($json_text)</td></tr>
+<tr><td class="h"><a name="688"></a>688</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="689"></a>689</td><td></td><td></td><td></td><td></td><td class="s">This works like the C<decode> method, but instead of raising an exception</td></tr>
+<tr><td class="h"><a name="690"></a>690</td><td></td><td></td><td></td><td></td><td class="s">when there is trailing garbage after the first JSON object, it will</td></tr>
+<tr><td class="h"><a name="691"></a>691</td><td></td><td></td><td></td><td></td><td class="s">silently stop parsing there and return the number of characters consumed</td></tr>
+<tr><td class="h"><a name="692"></a>692</td><td></td><td></td><td></td><td></td><td class="s">so far.</td></tr>
+<tr><td class="h"><a name="693"></a>693</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="694"></a>694</td><td></td><td></td><td></td><td></td><td class="s">This is useful if your JSON texts are not delimited by an outer protocol</td></tr>
+<tr><td class="h"><a name="695"></a>695</td><td></td><td></td><td></td><td></td><td class="s">(which is not the brightest thing to do in the first place) and you need</td></tr>
+<tr><td class="h"><a name="696"></a>696</td><td></td><td></td><td></td><td></td><td class="s">to know where the JSON text ends.</td></tr>
+<tr><td class="h"><a name="697"></a>697</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="698"></a>698</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS->new->decode_prefix ("[1] the tail")</td></tr>
+<tr><td class="h"><a name="699"></a>699</td><td></td><td></td><td></td><td></td><td class="s"> => ([], 3)</td></tr>
+<tr><td class="h"><a name="700"></a>700</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="701"></a>701</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="702"></a>702</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="703"></a>703</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="704"></a>704</td><td></td><td></td><td></td><td></td><td class="s">=head1 INCREMENTAL PARSING</td></tr>
+<tr><td class="h"><a name="705"></a>705</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="706"></a>706</td><td></td><td></td><td></td><td></td><td class="s">In some cases, there is the need for incremental parsing of JSON</td></tr>
+<tr><td class="h"><a name="707"></a>707</td><td></td><td></td><td></td><td></td><td class="s">texts. While this module always has to keep both JSON text and resulting</td></tr>
+<tr><td class="h"><a name="708"></a>708</td><td></td><td></td><td></td><td></td><td class="s">Perl data structure in memory at one time, it does allow you to parse a</td></tr>
+<tr><td class="h"><a name="709"></a>709</td><td></td><td></td><td></td><td></td><td class="s">JSON stream incrementally. It does so by accumulating text until it has</td></tr>
+<tr><td class="h"><a name="710"></a>710</td><td></td><td></td><td></td><td></td><td class="s">a full JSON object, which it then can decode. This process is similar to</td></tr>
+<tr><td class="h"><a name="711"></a>711</td><td></td><td></td><td></td><td></td><td class="s">using C<decode_prefix> to see if a full JSON object is available, but</td></tr>
+<tr><td class="h"><a name="712"></a>712</td><td></td><td></td><td></td><td></td><td class="s">is much more efficient (and can be implemented with a minimum of method</td></tr>
+<tr><td class="h"><a name="713"></a>713</td><td></td><td></td><td></td><td></td><td class="s">calls).</td></tr>
+<tr><td class="h"><a name="714"></a>714</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="715"></a>715</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS will only attempt to parse the JSON text once it is sure it</td></tr>
+<tr><td class="h"><a name="716"></a>716</td><td></td><td></td><td></td><td></td><td class="s">has enough text to get a decisive result, using a very simple but</td></tr>
+<tr><td class="h"><a name="717"></a>717</td><td></td><td></td><td></td><td></td><td class="s">truly incremental parser. This means that it sometimes won't stop as</td></tr>
+<tr><td class="h"><a name="718"></a>718</td><td></td><td></td><td></td><td></td><td class="s">early as the full parser, for example, it doesn't detect mismatched</td></tr>
+<tr><td class="h"><a name="719"></a>719</td><td></td><td></td><td></td><td></td><td class="s">parentheses. The only thing it guarantees is that it starts decoding as</td></tr>
+<tr><td class="h"><a name="720"></a>720</td><td></td><td></td><td></td><td></td><td class="s">soon as a syntactically valid JSON text has been seen. This means you need</td></tr>
+<tr><td class="h"><a name="721"></a>721</td><td></td><td></td><td></td><td></td><td class="s">to set resource limits (e.g. C<max_size>) to ensure the parser will stop</td></tr>
+<tr><td class="h"><a name="722"></a>722</td><td></td><td></td><td></td><td></td><td class="s">parsing in the presence if syntax errors.</td></tr>
+<tr><td class="h"><a name="723"></a>723</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="724"></a>724</td><td></td><td></td><td></td><td></td><td class="s">The following methods implement this incremental parser.</td></tr>
+<tr><td class="h"><a name="725"></a>725</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="726"></a>726</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="727"></a>727</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="728"></a>728</td><td></td><td></td><td></td><td></td><td class="s">=item [void, scalar or list context] = $json->incr_parse ([$string])</td></tr>
+<tr><td class="h"><a name="729"></a>729</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="730"></a>730</td><td></td><td></td><td></td><td></td><td class="s">This is the central parsing function. It can both append new text and</td></tr>
+<tr><td class="h"><a name="731"></a>731</td><td></td><td></td><td></td><td></td><td class="s">extract objects from the stream accumulated so far (both of these</td></tr>
+<tr><td class="h"><a name="732"></a>732</td><td></td><td></td><td></td><td></td><td class="s">functions are optional).</td></tr>
+<tr><td class="h"><a name="733"></a>733</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="734"></a>734</td><td></td><td></td><td></td><td></td><td class="s">If C<$string> is given, then this string is appended to the already</td></tr>
+<tr><td class="h"><a name="735"></a>735</td><td></td><td></td><td></td><td></td><td class="s">existing JSON fragment stored in the C<$json> object.</td></tr>
+<tr><td class="h"><a name="736"></a>736</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="737"></a>737</td><td></td><td></td><td></td><td></td><td class="s">After that, if the function is called in void context, it will simply</td></tr>
+<tr><td class="h"><a name="738"></a>738</td><td></td><td></td><td></td><td></td><td class="s">return without doing anything further. This can be used to add more text</td></tr>
+<tr><td class="h"><a name="739"></a>739</td><td></td><td></td><td></td><td></td><td class="s">in as many chunks as you want.</td></tr>
+<tr><td class="h"><a name="740"></a>740</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="741"></a>741</td><td></td><td></td><td></td><td></td><td class="s">If the method is called in scalar context, then it will try to extract</td></tr>
+<tr><td class="h"><a name="742"></a>742</td><td></td><td></td><td></td><td></td><td class="s">exactly I<one> JSON object. If that is successful, it will return this</td></tr>
+<tr><td class="h"><a name="743"></a>743</td><td></td><td></td><td></td><td></td><td class="s">object, otherwise it will return C<undef>. If there is a parse error,</td></tr>
+<tr><td class="h"><a name="744"></a>744</td><td></td><td></td><td></td><td></td><td class="s">this method will croak just as C<decode> would do (one can then use</td></tr>
+<tr><td class="h"><a name="745"></a>745</td><td></td><td></td><td></td><td></td><td class="s">C<incr_skip> to skip the errornous part). This is the most common way of</td></tr>
+<tr><td class="h"><a name="746"></a>746</td><td></td><td></td><td></td><td></td><td class="s">using the method.</td></tr>
+<tr><td class="h"><a name="747"></a>747</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="748"></a>748</td><td></td><td></td><td></td><td></td><td class="s">And finally, in list context, it will try to extract as many objects</td></tr>
+<tr><td class="h"><a name="749"></a>749</td><td></td><td></td><td></td><td></td><td class="s">from the stream as it can find and return them, or the empty list</td></tr>
+<tr><td class="h"><a name="750"></a>750</td><td></td><td></td><td></td><td></td><td class="s">otherwise. For this to work, there must be no separators between the JSON</td></tr>
+<tr><td class="h"><a name="751"></a>751</td><td></td><td></td><td></td><td></td><td class="s">objects or arrays, instead they must be concatenated back-to-back. If</td></tr>
+<tr><td class="h"><a name="752"></a>752</td><td></td><td></td><td></td><td></td><td class="s">an error occurs, an exception will be raised as in the scalar context</td></tr>
+<tr><td class="h"><a name="753"></a>753</td><td></td><td></td><td></td><td></td><td class="s">case. Note that in this case, any previously-parsed JSON texts will be</td></tr>
+<tr><td class="h"><a name="754"></a>754</td><td></td><td></td><td></td><td></td><td class="s">lost.</td></tr>
+<tr><td class="h"><a name="755"></a>755</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="756"></a>756</td><td></td><td></td><td></td><td></td><td class="s">Example: Parse some JSON arrays/objects in a given string and return</td></tr>
+<tr><td class="h"><a name="757"></a>757</td><td></td><td></td><td></td><td></td><td class="s">them.</td></tr>
+<tr><td class="h"><a name="758"></a>758</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="759"></a>759</td><td></td><td></td><td></td><td></td><td class="s"> my @objs = JSON::XS->new->incr_parse ("[5][7][1,2]");</td></tr>
+<tr><td class="h"><a name="760"></a>760</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="761"></a>761</td><td></td><td></td><td></td><td></td><td class="s">=item $lvalue_string = $json->incr_text</td></tr>
+<tr><td class="h"><a name="762"></a>762</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="763"></a>763</td><td></td><td></td><td></td><td></td><td class="s">This method returns the currently stored JSON fragment as an lvalue, that</td></tr>
+<tr><td class="h"><a name="764"></a>764</td><td></td><td></td><td></td><td></td><td class="s">is, you can manipulate it. This I<only> works when a preceding call to</td></tr>
+<tr><td class="h"><a name="765"></a>765</td><td></td><td></td><td></td><td></td><td class="s">C<incr_parse> in I<scalar context> successfully returned an object. Under</td></tr>
+<tr><td class="h"><a name="766"></a>766</td><td></td><td></td><td></td><td></td><td class="s">all other circumstances you must not call this function (I mean it.</td></tr>
+<tr><td class="h"><a name="767"></a>767</td><td></td><td></td><td></td><td></td><td class="s">although in simple tests it might actually work, it I<will> fail under</td></tr>
+<tr><td class="h"><a name="768"></a>768</td><td></td><td></td><td></td><td></td><td class="s">real world conditions). As a special exception, you can also call this</td></tr>
+<tr><td class="h"><a name="769"></a>769</td><td></td><td></td><td></td><td></td><td class="s">method before having parsed anything.</td></tr>
+<tr><td class="h"><a name="770"></a>770</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="771"></a>771</td><td></td><td></td><td></td><td></td><td class="s">This function is useful in two cases: a) finding the trailing text after a</td></tr>
+<tr><td class="h"><a name="772"></a>772</td><td></td><td></td><td></td><td></td><td class="s">JSON object or b) parsing multiple JSON objects separated by non-JSON text</td></tr>
+<tr><td class="h"><a name="773"></a>773</td><td></td><td></td><td></td><td></td><td class="s">(such as commas).</td></tr>
+<tr><td class="h"><a name="774"></a>774</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="775"></a>775</td><td></td><td></td><td></td><td></td><td class="s">=item $json->incr_skip</td></tr>
+<tr><td class="h"><a name="776"></a>776</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="777"></a>777</td><td></td><td></td><td></td><td></td><td class="s">This will reset the state of the incremental parser and will remove</td></tr>
+<tr><td class="h"><a name="778"></a>778</td><td></td><td></td><td></td><td></td><td class="s">the parsed text from the input buffer so far. This is useful after</td></tr>
+<tr><td class="h"><a name="779"></a>779</td><td></td><td></td><td></td><td></td><td class="s">C<incr_parse> died, in which case the input buffer and incremental parser</td></tr>
+<tr><td class="h"><a name="780"></a>780</td><td></td><td></td><td></td><td></td><td class="s">state is left unchanged, to skip the text parsed so far and to reset the</td></tr>
+<tr><td class="h"><a name="781"></a>781</td><td></td><td></td><td></td><td></td><td class="s">parse state.</td></tr>
+<tr><td class="h"><a name="782"></a>782</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="783"></a>783</td><td></td><td></td><td></td><td></td><td class="s">The difference to C<incr_reset> is that only text until the parse error</td></tr>
+<tr><td class="h"><a name="784"></a>784</td><td></td><td></td><td></td><td></td><td class="s">occured is removed.</td></tr>
+<tr><td class="h"><a name="785"></a>785</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="786"></a>786</td><td></td><td></td><td></td><td></td><td class="s">=item $json->incr_reset</td></tr>
+<tr><td class="h"><a name="787"></a>787</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="788"></a>788</td><td></td><td></td><td></td><td></td><td class="s">This completely resets the incremental parser, that is, after this call,</td></tr>
+<tr><td class="h"><a name="789"></a>789</td><td></td><td></td><td></td><td></td><td class="s">it will be as if the parser had never parsed anything.</td></tr>
+<tr><td class="h"><a name="790"></a>790</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="791"></a>791</td><td></td><td></td><td></td><td></td><td class="s">This is useful if you want to repeatedly parse JSON objects and want to</td></tr>
+<tr><td class="h"><a name="792"></a>792</td><td></td><td></td><td></td><td></td><td class="s">ignore any trailing data, which means you have to reset the parser after</td></tr>
+<tr><td class="h"><a name="793"></a>793</td><td></td><td></td><td></td><td></td><td class="s">each successful decode.</td></tr>
+<tr><td class="h"><a name="794"></a>794</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="795"></a>795</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="796"></a>796</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="797"></a>797</td><td></td><td></td><td></td><td></td><td class="s">=head2 LIMITATIONS</td></tr>
+<tr><td class="h"><a name="798"></a>798</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="799"></a>799</td><td></td><td></td><td></td><td></td><td class="s">All options that affect decoding are supported, except</td></tr>
+<tr><td class="h"><a name="800"></a>800</td><td></td><td></td><td></td><td></td><td class="s">C<allow_nonref>. The reason for this is that it cannot be made to</td></tr>
+<tr><td class="h"><a name="801"></a>801</td><td></td><td></td><td></td><td></td><td class="s">work sensibly: JSON objects and arrays are self-delimited, i.e. you can concatenate</td></tr>
+<tr><td class="h"><a name="802"></a>802</td><td></td><td></td><td></td><td></td><td class="s">them back to back and still decode them perfectly. This does not hold true</td></tr>
+<tr><td class="h"><a name="803"></a>803</td><td></td><td></td><td></td><td></td><td class="s">for JSON numbers, however.</td></tr>
+<tr><td class="h"><a name="804"></a>804</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="805"></a>805</td><td></td><td></td><td></td><td></td><td class="s">For example, is the string C<1> a single JSON number, or is it simply the</td></tr>
+<tr><td class="h"><a name="806"></a>806</td><td></td><td></td><td></td><td></td><td class="s">start of C<12>? Or is C<12> a single JSON number, or the concatenation</td></tr>
+<tr><td class="h"><a name="807"></a>807</td><td></td><td></td><td></td><td></td><td class="s">of C<1> and C<2>? In neither case you can tell, and this is why JSON::XS</td></tr>
+<tr><td class="h"><a name="808"></a>808</td><td></td><td></td><td></td><td></td><td class="s">takes the conservative route and disallows this case.</td></tr>
+<tr><td class="h"><a name="809"></a>809</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="810"></a>810</td><td></td><td></td><td></td><td></td><td class="s">=head2 EXAMPLES</td></tr>
+<tr><td class="h"><a name="811"></a>811</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="812"></a>812</td><td></td><td></td><td></td><td></td><td class="s">Some examples will make all this clearer. First, a simple example that</td></tr>
+<tr><td class="h"><a name="813"></a>813</td><td></td><td></td><td></td><td></td><td class="s">works similarly to C<decode_prefix>: We want to decode the JSON object at</td></tr>
+<tr><td class="h"><a name="814"></a>814</td><td></td><td></td><td></td><td></td><td class="s">the start of a string and identify the portion after the JSON object:</td></tr>
+<tr><td class="h"><a name="815"></a>815</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="816"></a>816</td><td></td><td></td><td></td><td></td><td class="s"> my $text = "[1,2,3] hello";</td></tr>
+<tr><td class="h"><a name="817"></a>817</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="818"></a>818</td><td></td><td></td><td></td><td></td><td class="s"> my $json = new JSON::XS;</td></tr>
+<tr><td class="h"><a name="819"></a>819</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="820"></a>820</td><td></td><td></td><td></td><td></td><td class="s"> my $obj = $json->incr_parse ($text)</td></tr>
+<tr><td class="h"><a name="821"></a>821</td><td></td><td></td><td></td><td></td><td class="s"> or die "expected JSON object or array at beginning of string";</td></tr>
+<tr><td class="h"><a name="822"></a>822</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="823"></a>823</td><td></td><td></td><td></td><td></td><td class="s"> my $tail = $json->incr_text;</td></tr>
+<tr><td class="h"><a name="824"></a>824</td><td></td><td></td><td></td><td></td><td class="s"> # $tail now contains " hello"</td></tr>
+<tr><td class="h"><a name="825"></a>825</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="826"></a>826</td><td></td><td></td><td></td><td></td><td class="s">Easy, isn't it?</td></tr>
+<tr><td class="h"><a name="827"></a>827</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="828"></a>828</td><td></td><td></td><td></td><td></td><td class="s">Now for a more complicated example: Imagine a hypothetical protocol where</td></tr>
+<tr><td class="h"><a name="829"></a>829</td><td></td><td></td><td></td><td></td><td class="s">you read some requests from a TCP stream, and each request is a JSON</td></tr>
+<tr><td class="h"><a name="830"></a>830</td><td></td><td></td><td></td><td></td><td class="s">array, without any separation between them (in fact, it is often useful to</td></tr>
+<tr><td class="h"><a name="831"></a>831</td><td></td><td></td><td></td><td></td><td class="s">use newlines as "separators", as these get interpreted as whitespace at</td></tr>
+<tr><td class="h"><a name="832"></a>832</td><td></td><td></td><td></td><td></td><td class="s">the start of the JSON text, which makes it possible to test said protocol</td></tr>
+<tr><td class="h"><a name="833"></a>833</td><td></td><td></td><td></td><td></td><td class="s">with C<telnet>...).</td></tr>
+<tr><td class="h"><a name="834"></a>834</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="835"></a>835</td><td></td><td></td><td></td><td></td><td class="s">Here is how you'd do it (it is trivial to write this in an event-based</td></tr>
+<tr><td class="h"><a name="836"></a>836</td><td></td><td></td><td></td><td></td><td class="s">manner):</td></tr>
+<tr><td class="h"><a name="837"></a>837</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="838"></a>838</td><td></td><td></td><td></td><td></td><td class="s"> my $json = new JSON::XS;</td></tr>
+<tr><td class="h"><a name="839"></a>839</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="840"></a>840</td><td></td><td></td><td></td><td></td><td class="s"> # read some data from the socket</td></tr>
+<tr><td class="h"><a name="841"></a>841</td><td></td><td></td><td></td><td></td><td class="s"> while (sysread $socket, my $buf, 4096) {</td></tr>
+<tr><td class="h"><a name="842"></a>842</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="843"></a>843</td><td></td><td></td><td></td><td></td><td class="s"> # split and decode as many requests as possible</td></tr>
+<tr><td class="h"><a name="844"></a>844</td><td></td><td></td><td></td><td></td><td class="s"> for my $request ($json->incr_parse ($buf)) {</td></tr>
+<tr><td class="h"><a name="845"></a>845</td><td></td><td></td><td></td><td></td><td class="s"> # act on the $request</td></tr>
+<tr><td class="h"><a name="846"></a>846</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="847"></a>847</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="848"></a>848</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="849"></a>849</td><td></td><td></td><td></td><td></td><td class="s">Another complicated example: Assume you have a string with JSON objects</td></tr>
+<tr><td class="h"><a name="850"></a>850</td><td></td><td></td><td></td><td></td><td class="s">or arrays, all separated by (optional) comma characters (e.g. C<[1],[2],</td></tr>
+<tr><td class="h"><a name="851"></a>851</td><td></td><td></td><td></td><td></td><td class="s">[3]>). To parse them, we have to skip the commas between the JSON texts,</td></tr>
+<tr><td class="h"><a name="852"></a>852</td><td></td><td></td><td></td><td></td><td class="s">and here is where the lvalue-ness of C<incr_text> comes in useful:</td></tr>
+<tr><td class="h"><a name="853"></a>853</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="854"></a>854</td><td></td><td></td><td></td><td></td><td class="s"> my $text = "[1],[2], [3]";</td></tr>
+<tr><td class="h"><a name="855"></a>855</td><td></td><td></td><td></td><td></td><td class="s"> my $json = new JSON::XS;</td></tr>
+<tr><td class="h"><a name="856"></a>856</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="857"></a>857</td><td></td><td></td><td></td><td></td><td class="s"> # void context, so no parsing done</td></tr>
+<tr><td class="h"><a name="858"></a>858</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_parse ($text);</td></tr>
+<tr><td class="h"><a name="859"></a>859</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="860"></a>860</td><td></td><td></td><td></td><td></td><td class="s"> # now extract as many objects as possible. note the</td></tr>
+<tr><td class="h"><a name="861"></a>861</td><td></td><td></td><td></td><td></td><td class="s"> # use of scalar context so incr_text can be called.</td></tr>
+<tr><td class="h"><a name="862"></a>862</td><td></td><td></td><td></td><td></td><td class="s"> while (my $obj = $json->incr_parse) {</td></tr>
+<tr><td class="h"><a name="863"></a>863</td><td></td><td></td><td></td><td></td><td class="s"> # do something with $obj</td></tr>
+<tr><td class="h"><a name="864"></a>864</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="865"></a>865</td><td></td><td></td><td></td><td></td><td class="s"> # now skip the optional comma</td></tr>
+<tr><td class="h"><a name="866"></a>866</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_text =~ s/^ \s* , //x;</td></tr>
+<tr><td class="h"><a name="867"></a>867</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="868"></a>868</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="869"></a>869</td><td></td><td></td><td></td><td></td><td class="s">Now lets go for a very complex example: Assume that you have a gigantic</td></tr>
+<tr><td class="h"><a name="870"></a>870</td><td></td><td></td><td></td><td></td><td class="s">JSON array-of-objects, many gigabytes in size, and you want to parse it,</td></tr>
+<tr><td class="h"><a name="871"></a>871</td><td></td><td></td><td></td><td></td><td class="s">but you cannot load it into memory fully (this has actually happened in</td></tr>
+<tr><td class="h"><a name="872"></a>872</td><td></td><td></td><td></td><td></td><td class="s">the real world :).</td></tr>
+<tr><td class="h"><a name="873"></a>873</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="874"></a>874</td><td></td><td></td><td></td><td></td><td class="s">Well, you lost, you have to implement your own JSON parser. But JSON::XS</td></tr>
+<tr><td class="h"><a name="875"></a>875</td><td></td><td></td><td></td><td></td><td class="s">can still help you: You implement a (very simple) array parser and let</td></tr>
+<tr><td class="h"><a name="876"></a>876</td><td></td><td></td><td></td><td></td><td class="s">JSON decode the array elements, which are all full JSON objects on their</td></tr>
+<tr><td class="h"><a name="877"></a>877</td><td></td><td></td><td></td><td></td><td class="s">own (this wouldn't work if the array elements could be JSON numbers, for</td></tr>
+<tr><td class="h"><a name="878"></a>878</td><td></td><td></td><td></td><td></td><td class="s">example):</td></tr>
+<tr><td class="h"><a name="879"></a>879</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="880"></a>880</td><td></td><td></td><td></td><td></td><td class="s"> my $json = new JSON::XS;</td></tr>
+<tr><td class="h"><a name="881"></a>881</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="882"></a>882</td><td></td><td></td><td></td><td></td><td class="s"> # open the monster</td></tr>
+<tr><td class="h"><a name="883"></a>883</td><td></td><td></td><td></td><td></td><td class="s"> open my $fh, "<bigfile.json"</td></tr>
+<tr><td class="h"><a name="884"></a>884</td><td></td><td></td><td></td><td></td><td class="s"> or die "bigfile: $!";</td></tr>
+<tr><td class="h"><a name="885"></a>885</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="886"></a>886</td><td></td><td></td><td></td><td></td><td class="s"> # first parse the initial "["</td></tr>
+<tr><td class="h"><a name="887"></a>887</td><td></td><td></td><td></td><td></td><td class="s"> for (;;) {</td></tr>
+<tr><td class="h"><a name="888"></a>888</td><td></td><td></td><td></td><td></td><td class="s"> sysread $fh, my $buf, 65536</td></tr>
+<tr><td class="h"><a name="889"></a>889</td><td></td><td></td><td></td><td></td><td class="s"> or die "read error: $!";</td></tr>
+<tr><td class="h"><a name="890"></a>890</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_parse ($buf); # void context, so no parsing</td></tr>
+<tr><td class="h"><a name="891"></a>891</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="892"></a>892</td><td></td><td></td><td></td><td></td><td class="s"> # Exit the loop once we found and removed(!) the initial "[".</td></tr>
+<tr><td class="h"><a name="893"></a>893</td><td></td><td></td><td></td><td></td><td class="s"> # In essence, we are (ab-)using the $json object as a simple scalar</td></tr>
+<tr><td class="h"><a name="894"></a>894</td><td></td><td></td><td></td><td></td><td class="s"> # we append data to.</td></tr>
+<tr><td class="h"><a name="895"></a>895</td><td></td><td></td><td></td><td></td><td class="s"> last if $json->incr_text =~ s/^ \s* \[ //x;</td></tr>
+<tr><td class="h"><a name="896"></a>896</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="897"></a>897</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="898"></a>898</td><td></td><td></td><td></td><td></td><td class="s"> # now we have the skipped the initial "[", so continue</td></tr>
+<tr><td class="h"><a name="899"></a>899</td><td></td><td></td><td></td><td></td><td class="s"> # parsing all the elements.</td></tr>
+<tr><td class="h"><a name="900"></a>900</td><td></td><td></td><td></td><td></td><td class="s"> for (;;) {</td></tr>
+<tr><td class="h"><a name="901"></a>901</td><td></td><td></td><td></td><td></td><td class="s"> # in this loop we read data until we got a single JSON object</td></tr>
+<tr><td class="h"><a name="902"></a>902</td><td></td><td></td><td></td><td></td><td class="s"> for (;;) {</td></tr>
+<tr><td class="h"><a name="903"></a>903</td><td></td><td></td><td></td><td></td><td class="s"> if (my $obj = $json->incr_parse) {</td></tr>
+<tr><td class="h"><a name="904"></a>904</td><td></td><td></td><td></td><td></td><td class="s"> # do something with $obj</td></tr>
+<tr><td class="h"><a name="905"></a>905</td><td></td><td></td><td></td><td></td><td class="s"> last;</td></tr>
+<tr><td class="h"><a name="906"></a>906</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="907"></a>907</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="908"></a>908</td><td></td><td></td><td></td><td></td><td class="s"> # add more data</td></tr>
+<tr><td class="h"><a name="909"></a>909</td><td></td><td></td><td></td><td></td><td class="s"> sysread $fh, my $buf, 65536</td></tr>
+<tr><td class="h"><a name="910"></a>910</td><td></td><td></td><td></td><td></td><td class="s"> or die "read error: $!";</td></tr>
+<tr><td class="h"><a name="911"></a>911</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_parse ($buf); # void context, so no parsing</td></tr>
+<tr><td class="h"><a name="912"></a>912</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="913"></a>913</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="914"></a>914</td><td></td><td></td><td></td><td></td><td class="s"> # in this loop we read data until we either found and parsed the</td></tr>
+<tr><td class="h"><a name="915"></a>915</td><td></td><td></td><td></td><td></td><td class="s"> # separating "," between elements, or the final "]"</td></tr>
+<tr><td class="h"><a name="916"></a>916</td><td></td><td></td><td></td><td></td><td class="s"> for (;;) {</td></tr>
+<tr><td class="h"><a name="917"></a>917</td><td></td><td></td><td></td><td></td><td class="s"> # first skip whitespace</td></tr>
+<tr><td class="h"><a name="918"></a>918</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_text =~ s/^\s*//;</td></tr>
+<tr><td class="h"><a name="919"></a>919</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="920"></a>920</td><td></td><td></td><td></td><td></td><td class="s"> # if we find "]", we are done</td></tr>
+<tr><td class="h"><a name="921"></a>921</td><td></td><td></td><td></td><td></td><td class="s"> if ($json->incr_text =~ s/^\]//) {</td></tr>
+<tr><td class="h"><a name="922"></a>922</td><td></td><td></td><td></td><td></td><td class="s"> print "finished.\n";</td></tr>
+<tr><td class="h"><a name="923"></a>923</td><td></td><td></td><td></td><td></td><td class="s"> exit;</td></tr>
+<tr><td class="h"><a name="924"></a>924</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="925"></a>925</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="926"></a>926</td><td></td><td></td><td></td><td></td><td class="s"> # if we find ",", we can continue with the next element</td></tr>
+<tr><td class="h"><a name="927"></a>927</td><td></td><td></td><td></td><td></td><td class="s"> if ($json->incr_text =~ s/^,//) {</td></tr>
+<tr><td class="h"><a name="928"></a>928</td><td></td><td></td><td></td><td></td><td class="s"> last;</td></tr>
+<tr><td class="h"><a name="929"></a>929</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="930"></a>930</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="931"></a>931</td><td></td><td></td><td></td><td></td><td class="s"> # if we find anything else, we have a parse error!</td></tr>
+<tr><td class="h"><a name="932"></a>932</td><td></td><td></td><td></td><td></td><td class="s"> if (length $json->incr_text) {</td></tr>
+<tr><td class="h"><a name="933"></a>933</td><td></td><td></td><td></td><td></td><td class="s"> die "parse error near ", $json->incr_text;</td></tr>
+<tr><td class="h"><a name="934"></a>934</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="935"></a>935</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="936"></a>936</td><td></td><td></td><td></td><td></td><td class="s"> # else add more data</td></tr>
+<tr><td class="h"><a name="937"></a>937</td><td></td><td></td><td></td><td></td><td class="s"> sysread $fh, my $buf, 65536</td></tr>
+<tr><td class="h"><a name="938"></a>938</td><td></td><td></td><td></td><td></td><td class="s"> or die "read error: $!";</td></tr>
+<tr><td class="h"><a name="939"></a>939</td><td></td><td></td><td></td><td></td><td class="s"> $json->incr_parse ($buf); # void context, so no parsing</td></tr>
+<tr><td class="h"><a name="940"></a>940</td><td></td><td></td><td></td><td></td><td class="s"> }</td></tr>
+<tr><td class="h"><a name="941"></a>941</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="942"></a>942</td><td></td><td></td><td></td><td></td><td class="s">This is a complex example, but most of the complexity comes from the fact</td></tr>
+<tr><td class="h"><a name="943"></a>943</td><td></td><td></td><td></td><td></td><td class="s">that we are trying to be correct (bear with me if I am wrong, I never ran</td></tr>
+<tr><td class="h"><a name="944"></a>944</td><td></td><td></td><td></td><td></td><td class="s">the above example :).</td></tr>
+<tr><td class="h"><a name="945"></a>945</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="- -"></a>- -</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="948"></a>948</td><td></td><td></td><td></td><td></td><td class="s">=head1 MAPPING</td></tr>
+<tr><td class="h"><a name="949"></a>949</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="950"></a>950</td><td></td><td></td><td></td><td></td><td class="s">This section describes how JSON::XS maps Perl values to JSON values and</td></tr>
+<tr><td class="h"><a name="951"></a>951</td><td></td><td></td><td></td><td></td><td class="s">vice versa. These mappings are designed to "do the right thing" in most</td></tr>
+<tr><td class="h"><a name="952"></a>952</td><td></td><td></td><td></td><td></td><td class="s">circumstances automatically, preserving round-tripping characteristics</td></tr>
+<tr><td class="h"><a name="953"></a>953</td><td></td><td></td><td></td><td></td><td class="s">(what you put in comes out as something equivalent).</td></tr>
+<tr><td class="h"><a name="954"></a>954</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="955"></a>955</td><td></td><td></td><td></td><td></td><td class="s">For the more enlightened: note that in the following descriptions,</td></tr>
+<tr><td class="h"><a name="956"></a>956</td><td></td><td></td><td></td><td></td><td class="s">lowercase I<perl> refers to the Perl interpreter, while uppercase I<Perl></td></tr>
+<tr><td class="h"><a name="957"></a>957</td><td></td><td></td><td></td><td></td><td class="s">refers to the abstract Perl language itself.</td></tr>
+<tr><td class="h"><a name="958"></a>958</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="959"></a>959</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="960"></a>960</td><td></td><td></td><td></td><td></td><td class="s">=head2 JSON -> PERL</td></tr>
+<tr><td class="h"><a name="961"></a>961</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="962"></a>962</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="963"></a>963</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="964"></a>964</td><td></td><td></td><td></td><td></td><td class="s">=item object</td></tr>
+<tr><td class="h"><a name="965"></a>965</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="966"></a>966</td><td></td><td></td><td></td><td></td><td class="s">A JSON object becomes a reference to a hash in Perl. No ordering of object</td></tr>
+<tr><td class="h"><a name="967"></a>967</td><td></td><td></td><td></td><td></td><td class="s">keys is preserved (JSON does not preserve object key ordering itself).</td></tr>
+<tr><td class="h"><a name="968"></a>968</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="969"></a>969</td><td></td><td></td><td></td><td></td><td class="s">=item array</td></tr>
+<tr><td class="h"><a name="970"></a>970</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="971"></a>971</td><td></td><td></td><td></td><td></td><td class="s">A JSON array becomes a reference to an array in Perl.</td></tr>
+<tr><td class="h"><a name="972"></a>972</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="973"></a>973</td><td></td><td></td><td></td><td></td><td class="s">=item string</td></tr>
+<tr><td class="h"><a name="974"></a>974</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="975"></a>975</td><td></td><td></td><td></td><td></td><td class="s">A JSON string becomes a string scalar in Perl - Unicode codepoints in JSON</td></tr>
+<tr><td class="h"><a name="976"></a>976</td><td></td><td></td><td></td><td></td><td class="s">are represented by the same codepoints in the Perl string, so no manual</td></tr>
+<tr><td class="h"><a name="977"></a>977</td><td></td><td></td><td></td><td></td><td class="s">decoding is necessary.</td></tr>
+<tr><td class="h"><a name="978"></a>978</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="979"></a>979</td><td></td><td></td><td></td><td></td><td class="s">=item number</td></tr>
+<tr><td class="h"><a name="980"></a>980</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="981"></a>981</td><td></td><td></td><td></td><td></td><td class="s">A JSON number becomes either an integer, numeric (floating point) or</td></tr>
+<tr><td class="h"><a name="982"></a>982</td><td></td><td></td><td></td><td></td><td class="s">string scalar in perl, depending on its range and any fractional parts. On</td></tr>
+<tr><td class="h"><a name="983"></a>983</td><td></td><td></td><td></td><td></td><td class="s">the Perl level, there is no difference between those as Perl handles all</td></tr>
+<tr><td class="h"><a name="984"></a>984</td><td></td><td></td><td></td><td></td><td class="s">the conversion details, but an integer may take slightly less memory and</td></tr>
+<tr><td class="h"><a name="985"></a>985</td><td></td><td></td><td></td><td></td><td class="s">might represent more values exactly than floating point numbers.</td></tr>
+<tr><td class="h"><a name="986"></a>986</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="987"></a>987</td><td></td><td></td><td></td><td></td><td class="s">If the number consists of digits only, JSON::XS will try to represent</td></tr>
+<tr><td class="h"><a name="988"></a>988</td><td></td><td></td><td></td><td></td><td class="s">it as an integer value. If that fails, it will try to represent it as</td></tr>
+<tr><td class="h"><a name="989"></a>989</td><td></td><td></td><td></td><td></td><td class="s">a numeric (floating point) value if that is possible without loss of</td></tr>
+<tr><td class="h"><a name="990"></a>990</td><td></td><td></td><td></td><td></td><td class="s">precision. Otherwise it will preserve the number as a string value (in</td></tr>
+<tr><td class="h"><a name="991"></a>991</td><td></td><td></td><td></td><td></td><td class="s">which case you lose roundtripping ability, as the JSON number will be</td></tr>
+<tr><td class="h"><a name="992"></a>992</td><td></td><td></td><td></td><td></td><td class="s">re-encoded toa JSON string).</td></tr>
+<tr><td class="h"><a name="993"></a>993</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="994"></a>994</td><td></td><td></td><td></td><td></td><td class="s">Numbers containing a fractional or exponential part will always be</td></tr>
+<tr><td class="h"><a name="995"></a>995</td><td></td><td></td><td></td><td></td><td class="s">represented as numeric (floating point) values, possibly at a loss of</td></tr>
+<tr><td class="h"><a name="996"></a>996</td><td></td><td></td><td></td><td></td><td class="s">precision (in which case you might lose perfect roundtripping ability, but</td></tr>
+<tr><td class="h"><a name="997"></a>997</td><td></td><td></td><td></td><td></td><td class="s">the JSON number will still be re-encoded as a JSON number).</td></tr>
+<tr><td class="h"><a name="998"></a>998</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="999"></a>999</td><td></td><td></td><td></td><td></td><td class="s">Note that precision is not accuracy - binary floating point values cannot</td></tr>
+<tr><td class="h"><a name="1000"></a>1000</td><td></td><td></td><td></td><td></td><td class="s">represent most decimal fractions exactly, and when converting from and to</td></tr>
+<tr><td class="h"><a name="1001"></a>1001</td><td></td><td></td><td></td><td></td><td class="s">floating point, JSON::XS only guarantees precision up to but not including</td></tr>
+<tr><td class="h"><a name="1002"></a>1002</td><td></td><td></td><td></td><td></td><td class="s">the leats significant bit.</td></tr>
+<tr><td class="h"><a name="1003"></a>1003</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1004"></a>1004</td><td></td><td></td><td></td><td></td><td class="s">=item true, false</td></tr>
+<tr><td class="h"><a name="1005"></a>1005</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1006"></a>1006</td><td></td><td></td><td></td><td></td><td class="s">These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>,</td></tr>
+<tr><td class="h"><a name="1007"></a>1007</td><td></td><td></td><td></td><td></td><td class="s">respectively. They are overloaded to act almost exactly like the numbers</td></tr>
+<tr><td class="h"><a name="1008"></a>1008</td><td></td><td></td><td></td><td></td><td class="s">C<1> and C<0>. You can check whether a scalar is a JSON boolean by using</td></tr>
+<tr><td class="h"><a name="1009"></a>1009</td><td></td><td></td><td></td><td></td><td class="s">the C<JSON::XS::is_bool> function.</td></tr>
+<tr><td class="h"><a name="1010"></a>1010</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1011"></a>1011</td><td></td><td></td><td></td><td></td><td class="s">=item null</td></tr>
+<tr><td class="h"><a name="1012"></a>1012</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1013"></a>1013</td><td></td><td></td><td></td><td></td><td class="s">A JSON null atom becomes C<undef> in Perl.</td></tr>
+<tr><td class="h"><a name="1014"></a>1014</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1015"></a>1015</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="1016"></a>1016</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1017"></a>1017</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1018"></a>1018</td><td></td><td></td><td></td><td></td><td class="s">=head2 PERL -> JSON</td></tr>
+<tr><td class="h"><a name="1019"></a>1019</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1020"></a>1020</td><td></td><td></td><td></td><td></td><td class="s">The mapping from Perl to JSON is slightly more difficult, as Perl is a</td></tr>
+<tr><td class="h"><a name="1021"></a>1021</td><td></td><td></td><td></td><td></td><td class="s">truly typeless language, so we can only guess which JSON type is meant by</td></tr>
+<tr><td class="h"><a name="1022"></a>1022</td><td></td><td></td><td></td><td></td><td class="s">a Perl value.</td></tr>
+<tr><td class="h"><a name="1023"></a>1023</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1024"></a>1024</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="1025"></a>1025</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1026"></a>1026</td><td></td><td></td><td></td><td></td><td class="s">=item hash references</td></tr>
+<tr><td class="h"><a name="1027"></a>1027</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1028"></a>1028</td><td></td><td></td><td></td><td></td><td class="s">Perl hash references become JSON objects. As there is no inherent ordering</td></tr>
+<tr><td class="h"><a name="1029"></a>1029</td><td></td><td></td><td></td><td></td><td class="s">in hash keys (or JSON objects), they will usually be encoded in a</td></tr>
+<tr><td class="h"><a name="1030"></a>1030</td><td></td><td></td><td></td><td></td><td class="s">pseudo-random order that can change between runs of the same program but</td></tr>
+<tr><td class="h"><a name="1031"></a>1031</td><td></td><td></td><td></td><td></td><td class="s">stays generally the same within a single run of a program. JSON::XS can</td></tr>
+<tr><td class="h"><a name="1032"></a>1032</td><td></td><td></td><td></td><td></td><td class="s">optionally sort the hash keys (determined by the I<canonical> flag), so</td></tr>
+<tr><td class="h"><a name="1033"></a>1033</td><td></td><td></td><td></td><td></td><td class="s">the same datastructure will serialise to the same JSON text (given same</td></tr>
+<tr><td class="h"><a name="1034"></a>1034</td><td></td><td></td><td></td><td></td><td class="s">settings and version of JSON::XS), but this incurs a runtime overhead</td></tr>
+<tr><td class="h"><a name="1035"></a>1035</td><td></td><td></td><td></td><td></td><td class="s">and is only rarely useful, e.g. when you want to compare some JSON text</td></tr>
+<tr><td class="h"><a name="1036"></a>1036</td><td></td><td></td><td></td><td></td><td class="s">against another for equality.</td></tr>
+<tr><td class="h"><a name="1037"></a>1037</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1038"></a>1038</td><td></td><td></td><td></td><td></td><td class="s">=item array references</td></tr>
+<tr><td class="h"><a name="1039"></a>1039</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1040"></a>1040</td><td></td><td></td><td></td><td></td><td class="s">Perl array references become JSON arrays.</td></tr>
+<tr><td class="h"><a name="1041"></a>1041</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1042"></a>1042</td><td></td><td></td><td></td><td></td><td class="s">=item other references</td></tr>
+<tr><td class="h"><a name="1043"></a>1043</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1044"></a>1044</td><td></td><td></td><td></td><td></td><td class="s">Other unblessed references are generally not allowed and will cause an</td></tr>
+<tr><td class="h"><a name="1045"></a>1045</td><td></td><td></td><td></td><td></td><td class="s">exception to be thrown, except for references to the integers C<0> and</td></tr>
+<tr><td class="h"><a name="1046"></a>1046</td><td></td><td></td><td></td><td></td><td class="s">C<1>, which get turned into C<false> and C<true> atoms in JSON. You can</td></tr>
+<tr><td class="h"><a name="1047"></a>1047</td><td></td><td></td><td></td><td></td><td class="s">also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability.</td></tr>
+<tr><td class="h"><a name="1048"></a>1048</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1049"></a>1049</td><td></td><td></td><td></td><td></td><td class="s"> encode_json [\0, JSON::XS::true] # yields [false,true]</td></tr>
+<tr><td class="h"><a name="1050"></a>1050</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1051"></a>1051</td><td></td><td></td><td></td><td></td><td class="s">=item JSON::XS::true, JSON::XS::false</td></tr>
+<tr><td class="h"><a name="1052"></a>1052</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1053"></a>1053</td><td></td><td></td><td></td><td></td><td class="s">These special values become JSON true and JSON false values,</td></tr>
+<tr><td class="h"><a name="1054"></a>1054</td><td></td><td></td><td></td><td></td><td class="s">respectively. You can also use C<\1> and C<\0> directly if you want.</td></tr>
+<tr><td class="h"><a name="1055"></a>1055</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1056"></a>1056</td><td></td><td></td><td></td><td></td><td class="s">=item blessed objects</td></tr>
+<tr><td class="h"><a name="1057"></a>1057</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1058"></a>1058</td><td></td><td></td><td></td><td></td><td class="s">Blessed objects are not directly representable in JSON. See the</td></tr>
+<tr><td class="h"><a name="1059"></a>1059</td><td></td><td></td><td></td><td></td><td class="s">C<allow_blessed> and C<convert_blessed> methods on various options on</td></tr>
+<tr><td class="h"><a name="1060"></a>1060</td><td></td><td></td><td></td><td></td><td class="s">how to deal with this: basically, you can choose between throwing an</td></tr>
+<tr><td class="h"><a name="1061"></a>1061</td><td></td><td></td><td></td><td></td><td class="s">exception, encoding the reference as if it weren't blessed, or provide</td></tr>
+<tr><td class="h"><a name="1062"></a>1062</td><td></td><td></td><td></td><td></td><td class="s">your own serialiser method.</td></tr>
+<tr><td class="h"><a name="1063"></a>1063</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1064"></a>1064</td><td></td><td></td><td></td><td></td><td class="s">=item simple scalars</td></tr>
+<tr><td class="h"><a name="1065"></a>1065</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1066"></a>1066</td><td></td><td></td><td></td><td></td><td class="s">Simple Perl scalars (any scalar that is not a reference) are the most</td></tr>
+<tr><td class="h"><a name="1067"></a>1067</td><td></td><td></td><td></td><td></td><td class="s">difficult objects to encode: JSON::XS will encode undefined scalars as</td></tr>
+<tr><td class="h"><a name="1068"></a>1068</td><td></td><td></td><td></td><td></td><td class="s">JSON C<null> values, scalars that have last been used in a string context</td></tr>
+<tr><td class="h"><a name="1069"></a>1069</td><td></td><td></td><td></td><td></td><td class="s">before encoding as JSON strings, and anything else as number value:</td></tr>
+<tr><td class="h"><a name="1070"></a>1070</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1071"></a>1071</td><td></td><td></td><td></td><td></td><td class="s"> # dump as number</td></tr>
+<tr><td class="h"><a name="1072"></a>1072</td><td></td><td></td><td></td><td></td><td class="s"> encode_json [2] # yields [2]</td></tr>
+<tr><td class="h"><a name="1073"></a>1073</td><td></td><td></td><td></td><td></td><td class="s"> encode_json [-3.0e17] # yields [-3e+17]</td></tr>
+<tr><td class="h"><a name="1074"></a>1074</td><td></td><td></td><td></td><td></td><td class="s"> my $value = 5; encode_json [$value] # yields [5]</td></tr>
+<tr><td class="h"><a name="1075"></a>1075</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1076"></a>1076</td><td></td><td></td><td></td><td></td><td class="s"> # used as string, so dump as string</td></tr>
+<tr><td class="h"><a name="1077"></a>1077</td><td></td><td></td><td></td><td></td><td class="s"> print $value;</td></tr>
+<tr><td class="h"><a name="1078"></a>1078</td><td></td><td></td><td></td><td></td><td class="s"> encode_json [$value] # yields ["5"]</td></tr>
+<tr><td class="h"><a name="1079"></a>1079</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1080"></a>1080</td><td></td><td></td><td></td><td></td><td class="s"> # undef becomes null</td></tr>
+<tr><td class="h"><a name="1081"></a>1081</td><td></td><td></td><td></td><td></td><td class="s"> encode_json [undef] # yields [null]</td></tr>
+<tr><td class="h"><a name="1082"></a>1082</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1083"></a>1083</td><td></td><td></td><td></td><td></td><td class="s">You can force the type to be a JSON string by stringifying it:</td></tr>
+<tr><td class="h"><a name="1084"></a>1084</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1085"></a>1085</td><td></td><td></td><td></td><td></td><td class="s"> my $x = 3.1; # some variable containing a number</td></tr>
+<tr><td class="h"><a name="1086"></a>1086</td><td></td><td></td><td></td><td></td><td class="s"> "$x"; # stringified</td></tr>
+<tr><td class="h"><a name="1087"></a>1087</td><td></td><td></td><td></td><td></td><td class="s"> $x .= ""; # another, more awkward way to stringify</td></tr>
+<tr><td class="h"><a name="1088"></a>1088</td><td></td><td></td><td></td><td></td><td class="s"> print $x; # perl does it for you, too, quite often</td></tr>
+<tr><td class="h"><a name="1089"></a>1089</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1090"></a>1090</td><td></td><td></td><td></td><td></td><td class="s">You can force the type to be a JSON number by numifying it:</td></tr>
+<tr><td class="h"><a name="1091"></a>1091</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1092"></a>1092</td><td></td><td></td><td></td><td></td><td class="s"> my $x = "3"; # some variable containing a string</td></tr>
+<tr><td class="h"><a name="1093"></a>1093</td><td></td><td></td><td></td><td></td><td class="s"> $x += 0; # numify it, ensuring it will be dumped as a number</td></tr>
+<tr><td class="h"><a name="1094"></a>1094</td><td></td><td></td><td></td><td></td><td class="s"> $x *= 1; # same thing, the choice is yours.</td></tr>
+<tr><td class="h"><a name="1095"></a>1095</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1096"></a>1096</td><td></td><td></td><td></td><td></td><td class="s">You can not currently force the type in other, less obscure, ways. Tell me</td></tr>
+<tr><td class="h"><a name="1097"></a>1097</td><td></td><td></td><td></td><td></td><td class="s">if you need this capability (but don't forget to explain why it's needed</td></tr>
+<tr><td class="h"><a name="1098"></a>1098</td><td></td><td></td><td></td><td></td><td class="s">:).</td></tr>
+<tr><td class="h"><a name="1099"></a>1099</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1100"></a>1100</td><td></td><td></td><td></td><td></td><td class="s">Note that numerical precision has the same meaning as under Perl (so</td></tr>
+<tr><td class="h"><a name="1101"></a>1101</td><td></td><td></td><td></td><td></td><td class="s">binary to decimal conversion follows the same rules as in Perl, which</td></tr>
+<tr><td class="h"><a name="1102"></a>1102</td><td></td><td></td><td></td><td></td><td class="s">can differ to other languages). Also, your perl interpreter might expose</td></tr>
+<tr><td class="h"><a name="1103"></a>1103</td><td></td><td></td><td></td><td></td><td class="s">extensions to the floating point numbers of your platform, such as</td></tr>
+<tr><td class="h"><a name="1104"></a>1104</td><td></td><td></td><td></td><td></td><td class="s">infinities or NaN's - these cannot be represented in JSON, and it is an</td></tr>
+<tr><td class="h"><a name="1105"></a>1105</td><td></td><td></td><td></td><td></td><td class="s">error to pass those in.</td></tr>
+<tr><td class="h"><a name="1106"></a>1106</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1107"></a>1107</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="1108"></a>1108</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1109"></a>1109</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1110"></a>1110</td><td></td><td></td><td></td><td></td><td class="s">=head1 ENCODING/CODESET FLAG NOTES</td></tr>
+<tr><td class="h"><a name="1111"></a>1111</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1112"></a>1112</td><td></td><td></td><td></td><td></td><td class="s">The interested reader might have seen a number of flags that signify</td></tr>
+<tr><td class="h"><a name="1113"></a>1113</td><td></td><td></td><td></td><td></td><td class="s">encodings or codesets - C<utf8>, C<latin1> and C<ascii>. There seems to be</td></tr>
+<tr><td class="h"><a name="1114"></a>1114</td><td></td><td></td><td></td><td></td><td class="s">some confusion on what these do, so here is a short comparison:</td></tr>
+<tr><td class="h"><a name="1115"></a>1115</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1116"></a>1116</td><td></td><td></td><td></td><td></td><td class="s">C<utf8> controls whether the JSON text created by C<encode> (and expected</td></tr>
+<tr><td class="h"><a name="1117"></a>1117</td><td></td><td></td><td></td><td></td><td class="s">by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only</td></tr>
+<tr><td class="h"><a name="1118"></a>1118</td><td></td><td></td><td></td><td></td><td class="s">control whether C<encode> escapes character values outside their respective</td></tr>
+<tr><td class="h"><a name="1119"></a>1119</td><td></td><td></td><td></td><td></td><td class="s">codeset range. Neither of these flags conflict with each other, although</td></tr>
+<tr><td class="h"><a name="1120"></a>1120</td><td></td><td></td><td></td><td></td><td class="s">some combinations make less sense than others.</td></tr>
+<tr><td class="h"><a name="1121"></a>1121</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1122"></a>1122</td><td></td><td></td><td></td><td></td><td class="s">Care has been taken to make all flags symmetrical with respect to</td></tr>
+<tr><td class="h"><a name="1123"></a>1123</td><td></td><td></td><td></td><td></td><td class="s">C<encode> and C<decode>, that is, texts encoded with any combination of</td></tr>
+<tr><td class="h"><a name="1124"></a>1124</td><td></td><td></td><td></td><td></td><td class="s">these flag values will be correctly decoded when the same flags are used</td></tr>
+<tr><td class="h"><a name="1125"></a>1125</td><td></td><td></td><td></td><td></td><td class="s">- in general, if you use different flag settings while encoding vs. when</td></tr>
+<tr><td class="h"><a name="1126"></a>1126</td><td></td><td></td><td></td><td></td><td class="s">decoding you likely have a bug somewhere.</td></tr>
+<tr><td class="h"><a name="1127"></a>1127</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1128"></a>1128</td><td></td><td></td><td></td><td></td><td class="s">Below comes a verbose discussion of these flags. Note that a "codeset" is</td></tr>
+<tr><td class="h"><a name="1129"></a>1129</td><td></td><td></td><td></td><td></td><td class="s">simply an abstract set of character-codepoint pairs, while an encoding</td></tr>
+<tr><td class="h"><a name="1130"></a>1130</td><td></td><td></td><td></td><td></td><td class="s">takes those codepoint numbers and I<encodes> them, in our case into</td></tr>
+<tr><td class="h"><a name="1131"></a>1131</td><td></td><td></td><td></td><td></td><td class="s">octets. Unicode is (among other things) a codeset, UTF-8 is an encoding,</td></tr>
+<tr><td class="h"><a name="1132"></a>1132</td><td></td><td></td><td></td><td></td><td class="s">and ISO-8859-1 (= latin 1) and ASCII are both codesets I<and> encodings at</td></tr>
+<tr><td class="h"><a name="1133"></a>1133</td><td></td><td></td><td></td><td></td><td class="s">the same time, which can be confusing.</td></tr>
+<tr><td class="h"><a name="1134"></a>1134</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1135"></a>1135</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="1136"></a>1136</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1137"></a>1137</td><td></td><td></td><td></td><td></td><td class="s">=item C<utf8> flag disabled</td></tr>
+<tr><td class="h"><a name="1138"></a>1138</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1139"></a>1139</td><td></td><td></td><td></td><td></td><td class="s">When C<utf8> is disabled (the default), then C<encode>/C<decode> generate</td></tr>
+<tr><td class="h"><a name="1140"></a>1140</td><td></td><td></td><td></td><td></td><td class="s">and expect Unicode strings, that is, characters with high ordinal Unicode</td></tr>
+<tr><td class="h"><a name="1141"></a>1141</td><td></td><td></td><td></td><td></td><td class="s">values (> 255) will be encoded as such characters, and likewise such</td></tr>
+<tr><td class="h"><a name="1142"></a>1142</td><td></td><td></td><td></td><td></td><td class="s">characters are decoded as-is, no canges to them will be done, except</td></tr>
+<tr><td class="h"><a name="1143"></a>1143</td><td></td><td></td><td></td><td></td><td class="s">"(re-)interpreting" them as Unicode codepoints or Unicode characters,</td></tr>
+<tr><td class="h"><a name="1144"></a>1144</td><td></td><td></td><td></td><td></td><td class="s">respectively (to Perl, these are the same thing in strings unless you do</td></tr>
+<tr><td class="h"><a name="1145"></a>1145</td><td></td><td></td><td></td><td></td><td class="s">funny/weird/dumb stuff).</td></tr>
+<tr><td class="h"><a name="1146"></a>1146</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1147"></a>1147</td><td></td><td></td><td></td><td></td><td class="s">This is useful when you want to do the encoding yourself (e.g. when you</td></tr>
+<tr><td class="h"><a name="1148"></a>1148</td><td></td><td></td><td></td><td></td><td class="s">want to have UTF-16 encoded JSON texts) or when some other layer does</td></tr>
+<tr><td class="h"><a name="1149"></a>1149</td><td></td><td></td><td></td><td></td><td class="s">the encoding for you (for example, when printing to a terminal using a</td></tr>
+<tr><td class="h"><a name="1150"></a>1150</td><td></td><td></td><td></td><td></td><td class="s">filehandle that transparently encodes to UTF-8 you certainly do NOT want</td></tr>
+<tr><td class="h"><a name="1151"></a>1151</td><td></td><td></td><td></td><td></td><td class="s">to UTF-8 encode your data first and have Perl encode it another time).</td></tr>
+<tr><td class="h"><a name="1152"></a>1152</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1153"></a>1153</td><td></td><td></td><td></td><td></td><td class="s">=item C<utf8> flag enabled</td></tr>
+<tr><td class="h"><a name="1154"></a>1154</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1155"></a>1155</td><td></td><td></td><td></td><td></td><td class="s">If the C<utf8>-flag is enabled, C<encode>/C<decode> will encode all</td></tr>
+<tr><td class="h"><a name="1156"></a>1156</td><td></td><td></td><td></td><td></td><td class="s">characters using the corresponding UTF-8 multi-byte sequence, and will</td></tr>
+<tr><td class="h"><a name="1157"></a>1157</td><td></td><td></td><td></td><td></td><td class="s">expect your input strings to be encoded as UTF-8, that is, no "character"</td></tr>
+<tr><td class="h"><a name="1158"></a>1158</td><td></td><td></td><td></td><td></td><td class="s">of the input string must have any value > 255, as UTF-8 does not allow</td></tr>
+<tr><td class="h"><a name="1159"></a>1159</td><td></td><td></td><td></td><td></td><td class="s">that.</td></tr>
+<tr><td class="h"><a name="1160"></a>1160</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1161"></a>1161</td><td></td><td></td><td></td><td></td><td class="s">The C<utf8> flag therefore switches between two modes: disabled means you</td></tr>
+<tr><td class="h"><a name="1162"></a>1162</td><td></td><td></td><td></td><td></td><td class="s">will get a Unicode string in Perl, enabled means you get an UTF-8 encoded</td></tr>
+<tr><td class="h"><a name="1163"></a>1163</td><td></td><td></td><td></td><td></td><td class="s">octet/binary string in Perl.</td></tr>
+<tr><td class="h"><a name="1164"></a>1164</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1165"></a>1165</td><td></td><td></td><td></td><td></td><td class="s">=item C<latin1> or C<ascii> flags enabled</td></tr>
+<tr><td class="h"><a name="1166"></a>1166</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1167"></a>1167</td><td></td><td></td><td></td><td></td><td class="s">With C<latin1> (or C<ascii>) enabled, C<encode> will escape characters</td></tr>
+<tr><td class="h"><a name="1168"></a>1168</td><td></td><td></td><td></td><td></td><td class="s">with ordinal values > 255 (> 127 with C<ascii>) and encode the remaining</td></tr>
+<tr><td class="h"><a name="1169"></a>1169</td><td></td><td></td><td></td><td></td><td class="s">characters as specified by the C<utf8> flag.</td></tr>
+<tr><td class="h"><a name="1170"></a>1170</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1171"></a>1171</td><td></td><td></td><td></td><td></td><td class="s">If C<utf8> is disabled, then the result is also correctly encoded in those</td></tr>
+<tr><td class="h"><a name="1172"></a>1172</td><td></td><td></td><td></td><td></td><td class="s">character sets (as both are proper subsets of Unicode, meaning that a</td></tr>
+<tr><td class="h"><a name="1173"></a>1173</td><td></td><td></td><td></td><td></td><td class="s">Unicode string with all character values < 256 is the same thing as a</td></tr>
+<tr><td class="h"><a name="1174"></a>1174</td><td></td><td></td><td></td><td></td><td class="s">ISO-8859-1 string, and a Unicode string with all character values < 128 is</td></tr>
+<tr><td class="h"><a name="1175"></a>1175</td><td></td><td></td><td></td><td></td><td class="s">the same thing as an ASCII string in Perl).</td></tr>
+<tr><td class="h"><a name="1176"></a>1176</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1177"></a>1177</td><td></td><td></td><td></td><td></td><td class="s">If C<utf8> is enabled, you still get a correct UTF-8-encoded string,</td></tr>
+<tr><td class="h"><a name="1178"></a>1178</td><td></td><td></td><td></td><td></td><td class="s">regardless of these flags, just some more characters will be escaped using</td></tr>
+<tr><td class="h"><a name="1179"></a>1179</td><td></td><td></td><td></td><td></td><td class="s">C<\uXXXX> then before.</td></tr>
+<tr><td class="h"><a name="1180"></a>1180</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1181"></a>1181</td><td></td><td></td><td></td><td></td><td class="s">Note that ISO-8859-1-I<encoded> strings are not compatible with UTF-8</td></tr>
+<tr><td class="h"><a name="1182"></a>1182</td><td></td><td></td><td></td><td></td><td class="s">encoding, while ASCII-encoded strings are. That is because the ISO-8859-1</td></tr>
+<tr><td class="h"><a name="1183"></a>1183</td><td></td><td></td><td></td><td></td><td class="s">encoding is NOT a subset of UTF-8 (despite the ISO-8859-1 I<codeset> being</td></tr>
+<tr><td class="h"><a name="1184"></a>1184</td><td></td><td></td><td></td><td></td><td class="s">a subset of Unicode), while ASCII is.</td></tr>
+<tr><td class="h"><a name="1185"></a>1185</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1186"></a>1186</td><td></td><td></td><td></td><td></td><td class="s">Surprisingly, C<decode> will ignore these flags and so treat all input</td></tr>
+<tr><td class="h"><a name="1187"></a>1187</td><td></td><td></td><td></td><td></td><td class="s">values as governed by the C<utf8> flag. If it is disabled, this allows you</td></tr>
+<tr><td class="h"><a name="1188"></a>1188</td><td></td><td></td><td></td><td></td><td class="s">to decode ISO-8859-1- and ASCII-encoded strings, as both strict subsets of</td></tr>
+<tr><td class="h"><a name="1189"></a>1189</td><td></td><td></td><td></td><td></td><td class="s">Unicode. If it is enabled, you can correctly decode UTF-8 encoded strings.</td></tr>
+<tr><td class="h"><a name="1190"></a>1190</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1191"></a>1191</td><td></td><td></td><td></td><td></td><td class="s">So neither C<latin1> nor C<ascii> are incompatible with the C<utf8> flag -</td></tr>
+<tr><td class="h"><a name="1192"></a>1192</td><td></td><td></td><td></td><td></td><td class="s">they only govern when the JSON output engine escapes a character or not.</td></tr>
+<tr><td class="h"><a name="1193"></a>1193</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1194"></a>1194</td><td></td><td></td><td></td><td></td><td class="s">The main use for C<latin1> is to relatively efficiently store binary data</td></tr>
+<tr><td class="h"><a name="1195"></a>1195</td><td></td><td></td><td></td><td></td><td class="s">as JSON, at the expense of breaking compatibility with most JSON decoders.</td></tr>
+<tr><td class="h"><a name="1196"></a>1196</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1197"></a>1197</td><td></td><td></td><td></td><td></td><td class="s">The main use for C<ascii> is to force the output to not contain characters</td></tr>
+<tr><td class="h"><a name="1198"></a>1198</td><td></td><td></td><td></td><td></td><td class="s">with values > 127, which means you can interpret the resulting string</td></tr>
+<tr><td class="h"><a name="1199"></a>1199</td><td></td><td></td><td></td><td></td><td class="s">as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and</td></tr>
+<tr><td class="h"><a name="1200"></a>1200</td><td></td><td></td><td></td><td></td><td class="s">8-bit-encoding, and still get the same data structure back. This is useful</td></tr>
+<tr><td class="h"><a name="1201"></a>1201</td><td></td><td></td><td></td><td></td><td class="s">when your channel for JSON transfer is not 8-bit clean or the encoding</td></tr>
+<tr><td class="h"><a name="1202"></a>1202</td><td></td><td></td><td></td><td></td><td class="s">might be mangled in between (e.g. in mail), and works because ASCII is a</td></tr>
+<tr><td class="h"><a name="1203"></a>1203</td><td></td><td></td><td></td><td></td><td class="s">proper subset of most 8-bit and multibyte encodings in use in the world.</td></tr>
+<tr><td class="h"><a name="1204"></a>1204</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1205"></a>1205</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="1206"></a>1206</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1207"></a>1207</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1208"></a>1208</td><td></td><td></td><td></td><td></td><td class="s">=head2 JSON and ECMAscript</td></tr>
+<tr><td class="h"><a name="1209"></a>1209</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1210"></a>1210</td><td></td><td></td><td></td><td></td><td class="s">JSON syntax is based on how literals are represented in javascript (the</td></tr>
+<tr><td class="h"><a name="1211"></a>1211</td><td></td><td></td><td></td><td></td><td class="s">not-standardised predecessor of ECMAscript) which is presumably why it is</td></tr>
+<tr><td class="h"><a name="1212"></a>1212</td><td></td><td></td><td></td><td></td><td class="s">called "JavaScript Object Notation".</td></tr>
+<tr><td class="h"><a name="1213"></a>1213</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1214"></a>1214</td><td></td><td></td><td></td><td></td><td class="s">However, JSON is not a subset (and also not a superset of course) of</td></tr>
+<tr><td class="h"><a name="1215"></a>1215</td><td></td><td></td><td></td><td></td><td class="s">ECMAscript (the standard) or javascript (whatever browsers actually</td></tr>
+<tr><td class="h"><a name="1216"></a>1216</td><td></td><td></td><td></td><td></td><td class="s">implement).</td></tr>
+<tr><td class="h"><a name="1217"></a>1217</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1218"></a>1218</td><td></td><td></td><td></td><td></td><td class="s">If you want to use javascript's C<eval> function to "parse" JSON, you</td></tr>
+<tr><td class="h"><a name="1219"></a>1219</td><td></td><td></td><td></td><td></td><td class="s">might run into parse errors for valid JSON texts, or the resulting data</td></tr>
+<tr><td class="h"><a name="1220"></a>1220</td><td></td><td></td><td></td><td></td><td class="s">structure might not be queryable:</td></tr>
+<tr><td class="h"><a name="1221"></a>1221</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1222"></a>1222</td><td></td><td></td><td></td><td></td><td class="s">One of the problems is that U+2028 and U+2029 are valid characters inside</td></tr>
+<tr><td class="h"><a name="1223"></a>1223</td><td></td><td></td><td></td><td></td><td class="s">JSON strings, but are not allowed in ECMAscript string literals, so the</td></tr>
+<tr><td class="h"><a name="1224"></a>1224</td><td></td><td></td><td></td><td></td><td class="s">following Perl fragment will not output something that can be guaranteed</td></tr>
+<tr><td class="h"><a name="1225"></a>1225</td><td></td><td></td><td></td><td></td><td class="s">to be parsable by javascript's C<eval>:</td></tr>
+<tr><td class="h"><a name="1226"></a>1226</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1227"></a>1227</td><td></td><td></td><td></td><td></td><td class="s"> use JSON::XS;</td></tr>
+<tr><td class="h"><a name="1228"></a>1228</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1229"></a>1229</td><td></td><td></td><td></td><td></td><td class="s"> print encode_json [chr 0x2028];</td></tr>
+<tr><td class="h"><a name="1230"></a>1230</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1231"></a>1231</td><td></td><td></td><td></td><td></td><td class="s">The right fix for this is to use a proper JSON parser in your javascript</td></tr>
+<tr><td class="h"><a name="1232"></a>1232</td><td></td><td></td><td></td><td></td><td class="s">programs, and not rely on C<eval> (see for example Douglas Crockford's</td></tr>
+<tr><td class="h"><a name="1233"></a>1233</td><td></td><td></td><td></td><td></td><td class="s">F<json2.js> parser).</td></tr>
+<tr><td class="h"><a name="1234"></a>1234</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1235"></a>1235</td><td></td><td></td><td></td><td></td><td class="s">If this is not an option, you can, as a stop-gap measure, simply encode to</td></tr>
+<tr><td class="h"><a name="1236"></a>1236</td><td></td><td></td><td></td><td></td><td class="s">ASCII-only JSON:</td></tr>
+<tr><td class="h"><a name="1237"></a>1237</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1238"></a>1238</td><td></td><td></td><td></td><td></td><td class="s"> use JSON::XS;</td></tr>
+<tr><td class="h"><a name="1239"></a>1239</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1240"></a>1240</td><td></td><td></td><td></td><td></td><td class="s"> print JSON::XS->new->ascii->encode ([chr 0x2028]);</td></tr>
+<tr><td class="h"><a name="1241"></a>1241</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1242"></a>1242</td><td></td><td></td><td></td><td></td><td class="s">Note that this will enlarge the resulting JSON text quite a bit if you</td></tr>
+<tr><td class="h"><a name="1243"></a>1243</td><td></td><td></td><td></td><td></td><td class="s">have many non-ASCII characters. You might be tempted to run some regexes</td></tr>
+<tr><td class="h"><a name="1244"></a>1244</td><td></td><td></td><td></td><td></td><td class="s">to only escape U+2028 and U+2029, e.g.:</td></tr>
+<tr><td class="h"><a name="1245"></a>1245</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1246"></a>1246</td><td></td><td></td><td></td><td></td><td class="s"> # DO NOT USE THIS!</td></tr>
+<tr><td class="h"><a name="1247"></a>1247</td><td></td><td></td><td></td><td></td><td class="s"> my $json = JSON::XS->new->utf8->encode ([chr 0x2028]);</td></tr>
+<tr><td class="h"><a name="1248"></a>1248</td><td></td><td></td><td></td><td></td><td class="s"> $json =~ s/\xe2\x80\xa8/\\u2028/g; # escape U+2028</td></tr>
+<tr><td class="h"><a name="1249"></a>1249</td><td></td><td></td><td></td><td></td><td class="s"> $json =~ s/\xe2\x80\xa9/\\u2029/g; # escape U+2029</td></tr>
+<tr><td class="h"><a name="1250"></a>1250</td><td></td><td></td><td></td><td></td><td class="s"> print $json;</td></tr>
+<tr><td class="h"><a name="1251"></a>1251</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1252"></a>1252</td><td></td><td></td><td></td><td></td><td class="s">Note that I<this is a bad idea>: the above only works for U+2028 and</td></tr>
+<tr><td class="h"><a name="1253"></a>1253</td><td></td><td></td><td></td><td></td><td class="s">U+2029 and thus only for fully ECMAscript-compliant parsers. Many existing</td></tr>
+<tr><td class="h"><a name="1254"></a>1254</td><td></td><td></td><td></td><td></td><td class="s">javascript implementations, however, have issues with other characters as</td></tr>
+<tr><td class="h"><a name="1255"></a>1255</td><td></td><td></td><td></td><td></td><td class="s">well - using C<eval> naively simply I<will> cause problems.</td></tr>
+<tr><td class="h"><a name="1256"></a>1256</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1257"></a>1257</td><td></td><td></td><td></td><td></td><td class="s">Another problem is that some javascript implementations reserve</td></tr>
+<tr><td class="h"><a name="1258"></a>1258</td><td></td><td></td><td></td><td></td><td class="s">some property names for their own purposes (which probably makes</td></tr>
+<tr><td class="h"><a name="1259"></a>1259</td><td></td><td></td><td></td><td></td><td class="s">them non-ECMAscript-compliant). For example, Iceweasel reserves the</td></tr>
+<tr><td class="h"><a name="1260"></a>1260</td><td></td><td></td><td></td><td></td><td class="s">C<__proto__> property name for its own purposes.</td></tr>
+<tr><td class="h"><a name="1261"></a>1261</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1262"></a>1262</td><td></td><td></td><td></td><td></td><td class="s">If that is a problem, you could parse try to filter the resulting JSON</td></tr>
+<tr><td class="h"><a name="1263"></a>1263</td><td></td><td></td><td></td><td></td><td class="s">output for these property strings, e.g.:</td></tr>
+<tr><td class="h"><a name="1264"></a>1264</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1265"></a>1265</td><td></td><td></td><td></td><td></td><td class="s"> $json =~ s/"__proto__"\s*:/"__proto__renamed":/g;</td></tr>
+<tr><td class="h"><a name="1266"></a>1266</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1267"></a>1267</td><td></td><td></td><td></td><td></td><td class="s">This works because C<__proto__> is not valid outside of strings, so every</td></tr>
+<tr><td class="h"><a name="1268"></a>1268</td><td></td><td></td><td></td><td></td><td class="s">occurence of C<"__proto__"\s*:> must be a string used as property name.</td></tr>
+<tr><td class="h"><a name="1269"></a>1269</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1270"></a>1270</td><td></td><td></td><td></td><td></td><td class="s">If you know of other incompatibilities, please let me know.</td></tr>
+<tr><td class="h"><a name="1271"></a>1271</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1272"></a>1272</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1273"></a>1273</td><td></td><td></td><td></td><td></td><td class="s">=head2 JSON and YAML</td></tr>
+<tr><td class="h"><a name="1274"></a>1274</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1275"></a>1275</td><td></td><td></td><td></td><td></td><td class="s">You often hear that JSON is a subset of YAML. This is, however, a mass</td></tr>
+<tr><td class="h"><a name="1276"></a>1276</td><td></td><td></td><td></td><td></td><td class="s">hysteria(*) and very far from the truth (as of the time of this writing),</td></tr>
+<tr><td class="h"><a name="1277"></a>1277</td><td></td><td></td><td></td><td></td><td class="s">so let me state it clearly: I<in general, there is no way to configure</td></tr>
+<tr><td class="h"><a name="1278"></a>1278</td><td></td><td></td><td></td><td></td><td class="s">JSON::XS to output a data structure as valid YAML> that works in all</td></tr>
+<tr><td class="h"><a name="1279"></a>1279</td><td></td><td></td><td></td><td></td><td class="s">cases.</td></tr>
+<tr><td class="h"><a name="1280"></a>1280</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1281"></a>1281</td><td></td><td></td><td></td><td></td><td class="s">If you really must use JSON::XS to generate YAML, you should use this</td></tr>
+<tr><td class="h"><a name="1282"></a>1282</td><td></td><td></td><td></td><td></td><td class="s">algorithm (subject to change in future versions):</td></tr>
+<tr><td class="h"><a name="1283"></a>1283</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1284"></a>1284</td><td></td><td></td><td></td><td></td><td class="s"> my $to_yaml = JSON::XS->new->utf8->space_after (1);</td></tr>
+<tr><td class="h"><a name="1285"></a>1285</td><td></td><td></td><td></td><td></td><td class="s"> my $yaml = $to_yaml->encode ($ref) . "\n";</td></tr>
+<tr><td class="h"><a name="1286"></a>1286</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1287"></a>1287</td><td></td><td></td><td></td><td></td><td class="s">This will I<usually> generate JSON texts that also parse as valid</td></tr>
+<tr><td class="h"><a name="1288"></a>1288</td><td></td><td></td><td></td><td></td><td class="s">YAML. Please note that YAML has hardcoded limits on (simple) object key</td></tr>
+<tr><td class="h"><a name="1289"></a>1289</td><td></td><td></td><td></td><td></td><td class="s">lengths that JSON doesn't have and also has different and incompatible</td></tr>
+<tr><td class="h"><a name="1290"></a>1290</td><td></td><td></td><td></td><td></td><td class="s">unicode character escape syntax, so you should make sure that your hash</td></tr>
+<tr><td class="h"><a name="1291"></a>1291</td><td></td><td></td><td></td><td></td><td class="s">keys are noticeably shorter than the 1024 "stream characters" YAML allows</td></tr>
+<tr><td class="h"><a name="1292"></a>1292</td><td></td><td></td><td></td><td></td><td class="s">and that you do not have characters with codepoint values outside the</td></tr>
+<tr><td class="h"><a name="1293"></a>1293</td><td></td><td></td><td></td><td></td><td class="s">Unicode BMP (basic multilingual page). YAML also does not allow C<\/></td></tr>
+<tr><td class="h"><a name="1294"></a>1294</td><td></td><td></td><td></td><td></td><td class="s">sequences in strings (which JSON::XS does not I<currently> generate, but</td></tr>
+<tr><td class="h"><a name="1295"></a>1295</td><td></td><td></td><td></td><td></td><td class="s">other JSON generators might).</td></tr>
+<tr><td class="h"><a name="1296"></a>1296</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1297"></a>1297</td><td></td><td></td><td></td><td></td><td class="s">There might be other incompatibilities that I am not aware of (or the YAML</td></tr>
+<tr><td class="h"><a name="1298"></a>1298</td><td></td><td></td><td></td><td></td><td class="s">specification has been changed yet again - it does so quite often). In</td></tr>
+<tr><td class="h"><a name="1299"></a>1299</td><td></td><td></td><td></td><td></td><td class="s">general you should not try to generate YAML with a JSON generator or vice</td></tr>
+<tr><td class="h"><a name="1300"></a>1300</td><td></td><td></td><td></td><td></td><td class="s">versa, or try to parse JSON with a YAML parser or vice versa: chances are</td></tr>
+<tr><td class="h"><a name="1301"></a>1301</td><td></td><td></td><td></td><td></td><td class="s">high that you will run into severe interoperability problems when you</td></tr>
+<tr><td class="h"><a name="1302"></a>1302</td><td></td><td></td><td></td><td></td><td class="s">least expect it.</td></tr>
+<tr><td class="h"><a name="1303"></a>1303</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1304"></a>1304</td><td></td><td></td><td></td><td></td><td class="s">=over 4</td></tr>
+<tr><td class="h"><a name="1305"></a>1305</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1306"></a>1306</td><td></td><td></td><td></td><td></td><td class="s">=item (*)</td></tr>
+<tr><td class="h"><a name="1307"></a>1307</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1308"></a>1308</td><td></td><td></td><td></td><td></td><td class="s">I have been pressured multiple times by Brian Ingerson (one of the</td></tr>
+<tr><td class="h"><a name="1309"></a>1309</td><td></td><td></td><td></td><td></td><td class="s">authors of the YAML specification) to remove this paragraph, despite him</td></tr>
+<tr><td class="h"><a name="1310"></a>1310</td><td></td><td></td><td></td><td></td><td class="s">acknowledging that the actual incompatibilities exist. As I was personally</td></tr>
+<tr><td class="h"><a name="1311"></a>1311</td><td></td><td></td><td></td><td></td><td class="s">bitten by this "JSON is YAML" lie, I refused and said I will continue to</td></tr>
+<tr><td class="h"><a name="1312"></a>1312</td><td></td><td></td><td></td><td></td><td class="s">educate people about these issues, so others do not run into the same</td></tr>
+<tr><td class="h"><a name="1313"></a>1313</td><td></td><td></td><td></td><td></td><td class="s">problem again and again. After this, Brian called me a (quote)I<complete</td></tr>
+<tr><td class="h"><a name="1314"></a>1314</td><td></td><td></td><td></td><td></td><td class="s">and worthless idiot>(unquote).</td></tr>
+<tr><td class="h"><a name="1315"></a>1315</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1316"></a>1316</td><td></td><td></td><td></td><td></td><td class="s">In my opinion, instead of pressuring and insulting people who actually</td></tr>
+<tr><td class="h"><a name="1317"></a>1317</td><td></td><td></td><td></td><td></td><td class="s">clarify issues with YAML and the wrong statements of some of its</td></tr>
+<tr><td class="h"><a name="1318"></a>1318</td><td></td><td></td><td></td><td></td><td class="s">proponents, I would kindly suggest reading the JSON spec (which is not</td></tr>
+<tr><td class="h"><a name="1319"></a>1319</td><td></td><td></td><td></td><td></td><td class="s">that difficult or long) and finally make YAML compatible to it, and</td></tr>
+<tr><td class="h"><a name="1320"></a>1320</td><td></td><td></td><td></td><td></td><td class="s">educating users about the changes, instead of spreading lies about the</td></tr>
+<tr><td class="h"><a name="1321"></a>1321</td><td></td><td></td><td></td><td></td><td class="s">real compatibility for many I<years> and trying to silence people who</td></tr>
+<tr><td class="h"><a name="1322"></a>1322</td><td></td><td></td><td></td><td></td><td class="s">point out that it isn't true.</td></tr>
+<tr><td class="h"><a name="1323"></a>1323</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1324"></a>1324</td><td></td><td></td><td></td><td></td><td class="s">Addendum/2009: the YAML 1.2 spec is still incompatible with JSON, even</td></tr>
+<tr><td class="h"><a name="1325"></a>1325</td><td></td><td></td><td></td><td></td><td class="s">though the incompatibilities have been documented (and are known to Brian)</td></tr>
+<tr><td class="h"><a name="1326"></a>1326</td><td></td><td></td><td></td><td></td><td class="s">for many years and the spec makes explicit claims that YAML is a superset</td></tr>
+<tr><td class="h"><a name="1327"></a>1327</td><td></td><td></td><td></td><td></td><td class="s">of JSON. It would be so easy to fix, but apparently, bullying people and</td></tr>
+<tr><td class="h"><a name="1328"></a>1328</td><td></td><td></td><td></td><td></td><td class="s">corrupting userdata is so much easier.</td></tr>
+<tr><td class="h"><a name="1329"></a>1329</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1330"></a>1330</td><td></td><td></td><td></td><td></td><td class="s">=back</td></tr>
+<tr><td class="h"><a name="1331"></a>1331</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1332"></a>1332</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1333"></a>1333</td><td></td><td></td><td></td><td></td><td class="s">=head2 SPEED</td></tr>
+<tr><td class="h"><a name="1334"></a>1334</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1335"></a>1335</td><td></td><td></td><td></td><td></td><td class="s">It seems that JSON::XS is surprisingly fast, as shown in the following</td></tr>
+<tr><td class="h"><a name="1336"></a>1336</td><td></td><td></td><td></td><td></td><td class="s">tables. They have been generated with the help of the C<eg/bench> program</td></tr>
+<tr><td class="h"><a name="1337"></a>1337</td><td></td><td></td><td></td><td></td><td class="s">in the JSON::XS distribution, to make it easy to compare on your own</td></tr>
+<tr><td class="h"><a name="1338"></a>1338</td><td></td><td></td><td></td><td></td><td class="s">system.</td></tr>
+<tr><td class="h"><a name="1339"></a>1339</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1340"></a>1340</td><td></td><td></td><td></td><td></td><td class="s">First comes a comparison between various modules using</td></tr>
+<tr><td class="h"><a name="1341"></a>1341</td><td></td><td></td><td></td><td></td><td class="s">a very short single-line JSON string (also available at</td></tr>
+<tr><td class="h"><a name="1342"></a>1342</td><td></td><td></td><td></td><td></td><td class="s">L<http://dist.schmorp.de/misc/json/short.json>).</td></tr>
+<tr><td class="h"><a name="1343"></a>1343</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1344"></a>1344</td><td></td><td></td><td></td><td></td><td class="s"> {"method": "handleMessage", "params": ["user1",</td></tr>
+<tr><td class="h"><a name="1345"></a>1345</td><td></td><td></td><td></td><td></td><td class="s"> "we were just talking"], "id": null, "array":[1,11,234,-5,1e5,1e7,</td></tr>
+<tr><td class="h"><a name="1346"></a>1346</td><td></td><td></td><td></td><td></td><td class="s"> 1, 0]}</td></tr>
+<tr><td class="h"><a name="1347"></a>1347</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1348"></a>1348</td><td></td><td></td><td></td><td></td><td class="s">It shows the number of encodes/decodes per second (JSON::XS uses</td></tr>
+<tr><td class="h"><a name="1349"></a>1349</td><td></td><td></td><td></td><td></td><td class="s">the functional interface, while JSON::XS/2 uses the OO interface</td></tr>
+<tr><td class="h"><a name="1350"></a>1350</td><td></td><td></td><td></td><td></td><td class="s">with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables</td></tr>
+<tr><td class="h"><a name="1351"></a>1351</td><td></td><td></td><td></td><td></td><td class="s">shrink. JSON::DWIW/DS uses the deserialise function, while JSON::DWIW::FJ</td></tr>
+<tr><td class="h"><a name="1352"></a>1352</td><td></td><td></td><td></td><td></td><td class="s">uses the from_json method). Higher is better:</td></tr>
+<tr><td class="h"><a name="1353"></a>1353</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1354"></a>1354</td><td></td><td></td><td></td><td></td><td class="s"> module | encode | decode |</td></tr>
+<tr><td class="h"><a name="1355"></a>1355</td><td></td><td></td><td></td><td></td><td class="s"> --------------|------------|------------|</td></tr>
+<tr><td class="h"><a name="1356"></a>1356</td><td></td><td></td><td></td><td></td><td class="s"> JSON::DWIW/DS | 86302.551 | 102300.098 |</td></tr>
+<tr><td class="h"><a name="1357"></a>1357</td><td></td><td></td><td></td><td></td><td class="s"> JSON::DWIW/FJ | 86302.551 | 75983.768 |</td></tr>
+<tr><td class="h"><a name="1358"></a>1358</td><td></td><td></td><td></td><td></td><td class="s"> JSON::PP | 15827.562 | 6638.658 |</td></tr>
+<tr><td class="h"><a name="1359"></a>1359</td><td></td><td></td><td></td><td></td><td class="s"> JSON::Syck | 63358.066 | 47662.545 |</td></tr>
+<tr><td class="h"><a name="1360"></a>1360</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS | 511500.488 | 511500.488 |</td></tr>
+<tr><td class="h"><a name="1361"></a>1361</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS/2 | 291271.111 | 388361.481 |</td></tr>
+<tr><td class="h"><a name="1362"></a>1362</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS/3 | 361577.931 | 361577.931 |</td></tr>
+<tr><td class="h"><a name="1363"></a>1363</td><td></td><td></td><td></td><td></td><td class="s"> Storable | 66788.280 | 265462.278 |</td></tr>
+<tr><td class="h"><a name="1364"></a>1364</td><td></td><td></td><td></td><td></td><td class="s"> --------------+------------+------------+</td></tr>
+<tr><td class="h"><a name="1365"></a>1365</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1366"></a>1366</td><td></td><td></td><td></td><td></td><td class="s">That is, JSON::XS is almost six times faster than JSON::DWIW on encoding,</td></tr>
+<tr><td class="h"><a name="1367"></a>1367</td><td></td><td></td><td></td><td></td><td class="s">about five times faster on decoding, and over thirty to seventy times</td></tr>
+<tr><td class="h"><a name="1368"></a>1368</td><td></td><td></td><td></td><td></td><td class="s">faster than JSON's pure perl implementation. It also compares favourably</td></tr>
+<tr><td class="h"><a name="1369"></a>1369</td><td></td><td></td><td></td><td></td><td class="s">to Storable for small amounts of data.</td></tr>
+<tr><td class="h"><a name="1370"></a>1370</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1371"></a>1371</td><td></td><td></td><td></td><td></td><td class="s">Using a longer test string (roughly 18KB, generated from Yahoo! Locals</td></tr>
+<tr><td class="h"><a name="1372"></a>1372</td><td></td><td></td><td></td><td></td><td class="s">search API (L<http://dist.schmorp.de/misc/json/long.json>).</td></tr>
+<tr><td class="h"><a name="1373"></a>1373</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1374"></a>1374</td><td></td><td></td><td></td><td></td><td class="s"> module | encode | decode |</td></tr>
+<tr><td class="h"><a name="1375"></a>1375</td><td></td><td></td><td></td><td></td><td class="s"> --------------|------------|------------|</td></tr>
+<tr><td class="h"><a name="1376"></a>1376</td><td></td><td></td><td></td><td></td><td class="s"> JSON::DWIW/DS | 1647.927 | 2673.916 |</td></tr>
+<tr><td class="h"><a name="1377"></a>1377</td><td></td><td></td><td></td><td></td><td class="s"> JSON::DWIW/FJ | 1630.249 | 2596.128 |</td></tr>
+<tr><td class="h"><a name="1378"></a>1378</td><td></td><td></td><td></td><td></td><td class="s"> JSON::PP | 400.640 | 62.311 |</td></tr>
+<tr><td class="h"><a name="1379"></a>1379</td><td></td><td></td><td></td><td></td><td class="s"> JSON::Syck | 1481.040 | 1524.869 |</td></tr>
+<tr><td class="h"><a name="1380"></a>1380</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS | 20661.596 | 9541.183 |</td></tr>
+<tr><td class="h"><a name="1381"></a>1381</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS/2 | 10683.403 | 9416.938 |</td></tr>
+<tr><td class="h"><a name="1382"></a>1382</td><td></td><td></td><td></td><td></td><td class="s"> JSON::XS/3 | 20661.596 | 9400.054 |</td></tr>
+<tr><td class="h"><a name="1383"></a>1383</td><td></td><td></td><td></td><td></td><td class="s"> Storable | 19765.806 | 10000.725 |</td></tr>
+<tr><td class="h"><a name="1384"></a>1384</td><td></td><td></td><td></td><td></td><td class="s"> --------------+------------+------------+</td></tr>
+<tr><td class="h"><a name="1385"></a>1385</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1386"></a>1386</td><td></td><td></td><td></td><td></td><td class="s">Again, JSON::XS leads by far (except for Storable which non-surprisingly</td></tr>
+<tr><td class="h"><a name="1387"></a>1387</td><td></td><td></td><td></td><td></td><td class="s">decodes a bit faster).</td></tr>
+<tr><td class="h"><a name="1388"></a>1388</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1389"></a>1389</td><td></td><td></td><td></td><td></td><td class="s">On large strings containing lots of high Unicode characters, some modules</td></tr>
+<tr><td class="h"><a name="1390"></a>1390</td><td></td><td></td><td></td><td></td><td class="s">(such as JSON::PC) seem to decode faster than JSON::XS, but the result</td></tr>
+<tr><td class="h"><a name="1391"></a>1391</td><td></td><td></td><td></td><td></td><td class="s">will be broken due to missing (or wrong) Unicode handling. Others refuse</td></tr>
+<tr><td class="h"><a name="1392"></a>1392</td><td></td><td></td><td></td><td></td><td class="s">to decode or encode properly, so it was impossible to prepare a fair</td></tr>
+<tr><td class="h"><a name="1393"></a>1393</td><td></td><td></td><td></td><td></td><td class="s">comparison table for that case.</td></tr>
+<tr><td class="h"><a name="1394"></a>1394</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1395"></a>1395</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1396"></a>1396</td><td></td><td></td><td></td><td></td><td class="s">=head1 SECURITY CONSIDERATIONS</td></tr>
+<tr><td class="h"><a name="1397"></a>1397</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1398"></a>1398</td><td></td><td></td><td></td><td></td><td class="s">When you are using JSON in a protocol, talking to untrusted potentially</td></tr>
+<tr><td class="h"><a name="1399"></a>1399</td><td></td><td></td><td></td><td></td><td class="s">hostile creatures requires relatively few measures.</td></tr>
+<tr><td class="h"><a name="1400"></a>1400</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1401"></a>1401</td><td></td><td></td><td></td><td></td><td class="s">First of all, your JSON decoder should be secure, that is, should not have</td></tr>
+<tr><td class="h"><a name="1402"></a>1402</td><td></td><td></td><td></td><td></td><td class="s">any buffer overflows. Obviously, this module should ensure that and I am</td></tr>
+<tr><td class="h"><a name="1403"></a>1403</td><td></td><td></td><td></td><td></td><td class="s">trying hard on making that true, but you never know.</td></tr>
+<tr><td class="h"><a name="1404"></a>1404</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1405"></a>1405</td><td></td><td></td><td></td><td></td><td class="s">Second, you need to avoid resource-starving attacks. That means you should</td></tr>
+<tr><td class="h"><a name="1406"></a>1406</td><td></td><td></td><td></td><td></td><td class="s">limit the size of JSON texts you accept, or make sure then when your</td></tr>
+<tr><td class="h"><a name="1407"></a>1407</td><td></td><td></td><td></td><td></td><td class="s">resources run out, that's just fine (e.g. by using a separate process that</td></tr>
+<tr><td class="h"><a name="1408"></a>1408</td><td></td><td></td><td></td><td></td><td class="s">can crash safely). The size of a JSON text in octets or characters is</td></tr>
+<tr><td class="h"><a name="1409"></a>1409</td><td></td><td></td><td></td><td></td><td class="s">usually a good indication of the size of the resources required to decode</td></tr>
+<tr><td class="h"><a name="1410"></a>1410</td><td></td><td></td><td></td><td></td><td class="s">it into a Perl structure. While JSON::XS can check the size of the JSON</td></tr>
+<tr><td class="h"><a name="1411"></a>1411</td><td></td><td></td><td></td><td></td><td class="s">text, it might be too late when you already have it in memory, so you</td></tr>
+<tr><td class="h"><a name="1412"></a>1412</td><td></td><td></td><td></td><td></td><td class="s">might want to check the size before you accept the string.</td></tr>
+<tr><td class="h"><a name="1413"></a>1413</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1414"></a>1414</td><td></td><td></td><td></td><td></td><td class="s">Third, JSON::XS recurses using the C stack when decoding objects and</td></tr>
+<tr><td class="h"><a name="1415"></a>1415</td><td></td><td></td><td></td><td></td><td class="s">arrays. The C stack is a limited resource: for instance, on my amd64</td></tr>
+<tr><td class="h"><a name="1416"></a>1416</td><td></td><td></td><td></td><td></td><td class="s">machine with 8MB of stack size I can decode around 180k nested arrays but</td></tr>
+<tr><td class="h"><a name="1417"></a>1417</td><td></td><td></td><td></td><td></td><td class="s">only 14k nested JSON objects (due to perl itself recursing deeply on croak</td></tr>
+<tr><td class="h"><a name="1418"></a>1418</td><td></td><td></td><td></td><td></td><td class="s">to free the temporary). If that is exceeded, the program crashes. To be</td></tr>
+<tr><td class="h"><a name="1419"></a>1419</td><td></td><td></td><td></td><td></td><td class="s">conservative, the default nesting limit is set to 512. If your process</td></tr>
+<tr><td class="h"><a name="1420"></a>1420</td><td></td><td></td><td></td><td></td><td class="s">has a smaller stack, you should adjust this setting accordingly with the</td></tr>
+<tr><td class="h"><a name="1421"></a>1421</td><td></td><td></td><td></td><td></td><td class="s">C<max_depth> method.</td></tr>
+<tr><td class="h"><a name="1422"></a>1422</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1423"></a>1423</td><td></td><td></td><td></td><td></td><td class="s">Something else could bomb you, too, that I forgot to think of. In that</td></tr>
+<tr><td class="h"><a name="1424"></a>1424</td><td></td><td></td><td></td><td></td><td class="s">case, you get to keep the pieces. I am always open for hints, though...</td></tr>
+<tr><td class="h"><a name="1425"></a>1425</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1426"></a>1426</td><td></td><td></td><td></td><td></td><td class="s">Also keep in mind that JSON::XS might leak contents of your Perl data</td></tr>
+<tr><td class="h"><a name="1427"></a>1427</td><td></td><td></td><td></td><td></td><td class="s">structures in its error messages, so when you serialise sensitive</td></tr>
+<tr><td class="h"><a name="1428"></a>1428</td><td></td><td></td><td></td><td></td><td class="s">information you might want to make sure that exceptions thrown by JSON::XS</td></tr>
+<tr><td class="h"><a name="1429"></a>1429</td><td></td><td></td><td></td><td></td><td class="s">will not end up in front of untrusted eyes.</td></tr>
+<tr><td class="h"><a name="1430"></a>1430</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1431"></a>1431</td><td></td><td></td><td></td><td></td><td class="s">If you are using JSON::XS to return packets to consumption</td></tr>
+<tr><td class="h"><a name="1432"></a>1432</td><td></td><td></td><td></td><td></td><td class="s">by JavaScript scripts in a browser you should have a look at</td></tr>
+<tr><td class="h"><a name="1433"></a>1433</td><td></td><td></td><td></td><td></td><td class="s">L<http://blog.archive.jpsykes.com/47/practical-csrf-and-json-security/> to</td></tr>
+<tr><td class="h"><a name="1434"></a>1434</td><td></td><td></td><td></td><td></td><td class="s">see whether you are vulnerable to some common attack vectors (which really</td></tr>
+<tr><td class="h"><a name="1435"></a>1435</td><td></td><td></td><td></td><td></td><td class="s">are browser design bugs, but it is still you who will have to deal with</td></tr>
+<tr><td class="h"><a name="1436"></a>1436</td><td></td><td></td><td></td><td></td><td class="s">it, as major browser developers care only for features, not about getting</td></tr>
+<tr><td class="h"><a name="1437"></a>1437</td><td></td><td></td><td></td><td></td><td class="s">security right).</td></tr>
+<tr><td class="h"><a name="1438"></a>1438</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1439"></a>1439</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1440"></a>1440</td><td></td><td></td><td></td><td></td><td class="s">=head1 THREADS</td></tr>
+<tr><td class="h"><a name="1441"></a>1441</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1442"></a>1442</td><td></td><td></td><td></td><td></td><td class="s">This module is I<not> guaranteed to be thread safe and there are no</td></tr>
+<tr><td class="h"><a name="1443"></a>1443</td><td></td><td></td><td></td><td></td><td class="s">plans to change this until Perl gets thread support (as opposed to the</td></tr>
+<tr><td class="h"><a name="1444"></a>1444</td><td></td><td></td><td></td><td></td><td class="s">horribly slow so-called "threads" which are simply slow and bloated</td></tr>
+<tr><td class="h"><a name="1445"></a>1445</td><td></td><td></td><td></td><td></td><td class="s">process simulations - use fork, it's I<much> faster, cheaper, better).</td></tr>
+<tr><td class="h"><a name="1446"></a>1446</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1447"></a>1447</td><td></td><td></td><td></td><td></td><td class="s">(It might actually work, but you have been warned).</td></tr>
+<tr><td class="h"><a name="1448"></a>1448</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1449"></a>1449</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1450"></a>1450</td><td></td><td></td><td></td><td></td><td class="s">=head1 BUGS</td></tr>
+<tr><td class="h"><a name="1451"></a>1451</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1452"></a>1452</td><td></td><td></td><td></td><td></td><td class="s">While the goal of this module is to be correct, that unfortunately does</td></tr>
+<tr><td class="h"><a name="1453"></a>1453</td><td></td><td></td><td></td><td></td><td class="s">not mean it's bug-free, only that I think its design is bug-free. If you</td></tr>
+<tr><td class="h"><a name="1454"></a>1454</td><td></td><td></td><td></td><td></td><td class="s">keep reporting bugs they will be fixed swiftly, though.</td></tr>
+<tr><td class="h"><a name="1455"></a>1455</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1456"></a>1456</td><td></td><td></td><td></td><td></td><td class="s">Please refrain from using rt.cpan.org or any other bug reporting</td></tr>
+<tr><td class="h"><a name="1457"></a>1457</td><td></td><td></td><td></td><td></td><td class="s">service. I put the contact address into my modules for a reason.</td></tr>
+<tr><td class="h"><a name="1458"></a>1458</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1459"></a>1459</td><td></td><td></td><td></td><td></td><td class="s">=cut</td></tr>
+<tr><td class="h"><a name="1460"></a>1460</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1461"></a>1461</td><td class="c3">1</td><td class="c3"><span title="Avg 38µs">38µs</span></td><td></td><td></td><td class="s">our $true = do { bless \(my $dummy = 1), "JSON::XS::Boolean" };</td></tr>
+<tr><td class="h"><a name="1462"></a>1462</td><td class="c3">1</td><td class="c3"><span title="Avg 1µs">1µs</span></td><td></td><td></td><td class="s">our $false = do { bless \(my $dummy = 0), "JSON::XS::Boolean" };</td></tr>
+<tr><td class="h"><a name="1463"></a>1463</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1464"></a>1464</td><td></td><td></td><td></td><td></td><td class="s">sub true() { $true }</td></tr>
+<tr><td class="h"><a name="1465"></a>1465</td><td></td><td></td><td></td><td></td><td class="s">sub false() { $false }</td></tr>
+<tr><td class="h"><a name="1466"></a>1466</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1467"></a>1467</td><td></td><td></td><td></td><td></td><td class="s">sub is_bool($) {</td></tr>
+<tr><td class="h"><a name="1468"></a>1468</td><td></td><td></td><td></td><td></td><td class="s"> UNIVERSAL::isa $_[0], "JSON::XS::Boolean"</td></tr>
+<tr><td class="h"><a name="1469"></a>1469</td><td></td><td></td><td></td><td></td><td class="s"># or UNIVERSAL::isa $_[0], "JSON::Literal"</td></tr>
+<tr><td class="h"><a name="1470"></a>1470</td><td></td><td></td><td></td><td></td><td class="s">}</td></tr>
+<tr><td class="h"><a name="1471"></a>1471</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1472"></a>1472</td><td class="c3">1</td><td class="c0"><span title="Avg 505µs">505µs</span></td><td class="c3">1</td><td class="c0">623µs</td><td class="s">XSLoader::load "JSON::XS", $VERSION;<div class="calls"><div class="calls_out"># spent 623µs making 1 call to <a href="XSLoader-pm-681-block.html#20">XSLoader::load</a></div></div></td></tr>
+<tr><td class="h"><a name="1473"></a>1473</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1474"></a>1474</td><td></td><td></td><td></td><td></td><td class="s">package JSON::XS::Boolean;</td></tr>
+<tr><td class="h"><a name="1475"></a>1475</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1476"></a>1476</td><td></td><td></td><td></td><td></td><td class="s">use overload</td></tr>
+<tr><td class="h"><a name="1477"></a>1477</td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 76µs (17+59) within JSON::XS::Boolean::BEGIN@1477 which was called:
+# once (17µs+59µs) by JSON::BEGIN@2 at <a href="JSON-XS-pm-679-block.html#1480">line 1480</a></div></div> "0+" => sub { ${$_[0]} },</td></tr>
+<tr><td class="h"><a name="1478"></a>1478</td><td></td><td></td><td></td><td></td><td class="s"> "++" => sub { $_[0] = ${$_[0]} + 1 },</td></tr>
+<tr><td class="h"><a name="1479"></a>1479</td><td></td><td></td><td></td><td></td><td class="s"> "--" => sub { $_[0] = ${$_[0]} - 1 },</td></tr>
+<tr><td class="h"><a name="1480"></a>1480</td><td class="c1">2</td><td class="c3"><span title="Avg 28µs">55µs</span></td><td class="c3">2</td><td class="c3">135µs</td><td class="s"> fallback => 1;<div class="calls"><div class="calls_out"> # spent 76µs making 1 call to <a href="JSON-XS-pm-679-block.html#1477">JSON::XS::Boolean::BEGIN@1477</a>
+ # spent 59µs making 1 call to <a href="overload-pm-29-block.html#30">overload::import</a></div></div></td></tr>
+<tr><td class="h"><a name="1481"></a>1481</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1482"></a>1482</td><td class="c3">1</td><td class="c3"><span title="Avg 23µs">23µs</span></td><td></td><td></td><td class="s">1;</td></tr>
+<tr><td class="h"><a name="1483"></a>1483</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1484"></a>1484</td><td></td><td></td><td></td><td></td><td class="s">=head1 SEE ALSO</td></tr>
+<tr><td class="h"><a name="1485"></a>1485</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1486"></a>1486</td><td></td><td></td><td></td><td></td><td class="s">The F<json_xs> command line utility for quick experiments.</td></tr>
+<tr><td class="h"><a name="1487"></a>1487</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1488"></a>1488</td><td></td><td></td><td></td><td></td><td class="s">=head1 AUTHOR</td></tr>
+<tr><td class="h"><a name="1489"></a>1489</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1490"></a>1490</td><td></td><td></td><td></td><td></td><td class="s"> Marc Lehmann <schmorp@schmorp.de></td></tr>
+<tr><td class="h"><a name="1491"></a>1491</td><td></td><td></td><td></td><td></td><td class="s"> http://home.schmorp.de/</td></tr>
+<tr><td class="h"><a name="1492"></a>1492</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="1493"></a>1493</td><td></td><td></td><td></td><td></td><td class="s">=cut</td></tr>
+<tr><td class="h"><a name="1494"></a>1494</td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="s"><a name=""></a> </td><td></td><td></td><td></td><td></td><td class="s"></td></tr>
+<tr><td class="h"><a name="JSON__XS__canonical"></a></td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 900ns within JSON::XS::canonical which was called:
+# once (900ns+0s) by KiokuDB::Backend::Serialize::JSON::_build_json at <a href="KiokuDB-Backend-Serialize-JSON-pm-2330-block.html#41">line 41 of KiokuDB/Backend/Serialize/JSON.pm</a></div></div>sub JSON::XS::canonical; # xsub<br /> </td></tr>
+<tr><td class="h"><a name="JSON__XS__encode"></a></td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 41.3ms within JSON::XS::encode which was called 2326 times, avg 18µs/call:
+# 2326 times (41.3ms+0s) by JSON::Backend::XS::Supportable::_encode at <a href="(eval 459)[JSON-pm-268]-682-line.html#151">line 151 of (eval 459)[JSON.pm:268]</a>, avg 18µs/call</div></div>sub JSON::XS::encode; # xsub<br /> </td></tr>
+<tr><td class="h"><a name="JSON__XS__new"></a></td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 34µs within JSON::XS::new which was called 4 times, avg 8µs/call:
+# 4 times (34µs+0s) by JSON::Backend::XS::__ANON__[(eval 459)[/Users/edenc/perl5/lib/perl5/JSON.pm:268]:63] at <a href="(eval 459)[JSON-pm-268]-682-line.html#61">line 61 of (eval 459)[JSON.pm:268]</a>, avg 8µs/call</div></div>sub JSON::XS::new; # xsub<br /> </td></tr>
+<tr><td class="h"><a name="JSON__XS__utf8"></a></td><td></td><td></td><td></td><td></td><td class="s"><div class="calls"><div class="calls_in"># spent 4µs within JSON::XS::utf8 which was called 2 times, avg 2µs/call:
+# once (2µs+0s) by JSON::Any::__ANON__[/Users/edenc/perl5/lib/perl5/JSON/Any.pm:108] at <a href="JSON-Any-pm-2146-block.html#102">line 102 of JSON/Any.pm</a>
+# once (1µs+0s) by KiokuDB::Backend::Serialize::JSON::_build_json at <a href="KiokuDB-Backend-Serialize-JSON-pm-2330-block.html#41">line 41 of KiokuDB/Backend/Serialize/JSON.pm</a></div></div>sub JSON::XS::utf8; # xsub<br /> </td></tr>
+</tbody></table></div>
+
+ <script type="text/javascript"> $(document).ready(function() {
+
+ $("#subs_table").tablesorter({
+ sortList: [[3,1]],
+ headers: {
+ 3: { sorter: 'fmt_time' },
+ 4: { sorter: 'fmt_time' }
+ }
+ });
+
+ } ); </script>
+
+ <div class="footer">Report produced by the
+ <a href="http://search.cpan.org/dist/Devel-NYTProf/">NYTProf 4.06</a>
+ Perl profiler, developed by
+ <a href="http://www.linkedin.com/in/timbunce">Tim Bunce</a> and
+ <a href="http://code.nytimes.com">Adam Kaplan</a>.
+ </div>
+ <br /><br /><br /><br /><br /><br /><br /><br /><br /><br />
+ </body></html>
\ No newline at end of file
projects / scpubgit/stemmatology.git / blobdiff