<?xml version="1.0" ?>
<head>
<title>Random-Lists</title>
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
<link rev="made" href="mailto:rurban@x-ray.at" />
</head>
<body style="background-color: white">
<p><a name="__index__"></a></p>
<!-- INDEX BEGIN -->
<ul>
<li><a href="#venue">VENUE</a></li>
<li><a href="#synopsis">SYNOPSIS</a></li>
<li><a href="#description">DESCRIPTION</a></li>
<ul>
<li><a href="#venue">Venue</a></li>
<li><a href="#character_encoding">Character Encoding</a></li>
<li><a href="#values_and_default_values">Values and Default Values</a></li>
<li><a href="#numbers__strings_and_heredocuments">Numbers, Strings and Here-Documents</a></li>
<li><a href="#list_values">List Values</a></li>
<li><a href="#binary_data">Binary Data</a></li>
<li><a href="#embedded_perl_code__nanoscripts_">Embedded Perl Code (Nanoscripts)</a></li>
<li><a href="#comments">Comments</a></li>
</ul>
<li><a href="#package_interface">PACKAGE INTERFACE</a></li>
<ul>
<li><a href="#construction">Construction</a></li>
<li><a href="#attribute_access">Attribute Access</a></li>
<li><a href="#public_functions">Public Functions</a></li>
<li><a href="#static_functions">Static Functions</a></li>
<li><a href="#implementation_functions">Implementation Functions</a></li>
<li><a href="#auxiliary_functions">Auxiliary Functions</a></li>
<li><a href="#predefined_options">Predefined Options</a></li>
<li><a href="#exports">Exports</a></li>
<ul>
<li><a href="#exporter_tags">Exporter Tags</a></li>
<li><a href="#autoexported_functions">Auto-Exported Functions</a></li>
</ul>
</ul>
<li><a href="#examples">EXAMPLES</a></li>
<li><a href="#notes">NOTES</a></li>
<li><a href="#data__dumper"><em>Data::Dumper</em></a></li>
<ul>
<li><a href="#rlist_vs__perl_syntax">Rlist vs. Perl Syntax</a></li>
<li><a href="#debugging_data">Debugging Data</a></li>
<li><a href="#speeding_up_compilation__explicit_quoting_">Speeding up Compilation (Explicit Quoting)</a></li>
<li><a href="#quoting_strings_that_look_like_numbers">Quoting strings that look like numbers</a></li>
<li><a href="#installing_rlist_pm_locally">Installing <em>Rlist.pm</em> locally</a></li>
<li><a href="#an_rlistmode_for_emacs">An Rlist-Mode for Emacs</a></li>
<li><a href="#implementation_details">Implementation Details</a></li>
<ul>
<li><a href="#perl">Perl</a></li>
<ul>
<li><a href="#package_dependencies">Package Dependencies</a></li>
<li><a href="#a_short_story_of_typeglobs">A Short Story of Typeglobs</a></li>
</ul>
<li><a href="#c__">C++</a></li>
</ul>
</ul>
<li><a href="#bugs">BUGS</a></li>
<li><a href="#copyright_license">COPYRIGHT/LICENSE</a></li>
</ul>
<!-- INDEX END -->
<hr />
<p>
</p>
<h1><a name="venue">VENUE</a></h1>
<p>Data::Rlist - A lightweight data language for Perl and C++</p>
<p>
</p>
<hr />
<h1><a name="synopsis">SYNOPSIS</a></h1>
<pre>
use Data::Rlist;</pre>
<p>File and string I/O for any Perl data <em>$thing</em>:</p>
<pre>
### Compile data as text.</pre>
<pre>
WriteData $thing, $filename; # compile data into file
WriteData $thing, \$string; # compile data into buffer
$string_ref = WriteData $thing; # dto.</pre>
<pre>
$string = OutlineData $thing; # compile printable text
$string = StringizeData $thing; # compile text in a compact form (no newlines)
$string = SqueezeData $thing; # compile text in a super-compact form (no whitespace)</pre>
<pre>
### Parse data from text.</pre>
<pre>
$thing = ReadData $filename; # parse data from file
$thing = ReadData \$string; # parse data from string buffer</pre>
<p><em><a href="#item_readdata">ReadData</a></em>, <em><a href="#item_writedata">WriteData</a></em> etc. are <a href="#exports">auto-exported functions</a>. Alternately we
use:</p>
<pre>
### Qualified functions to parse text.</pre>
<pre>
$thing = Data::Rlist::read($filename);
$thing = Data::Rlist::read($string_ref);
$thing = Data::Rlist::read_string($string_or_string_ref);</pre>
<pre>
### Qualified functions to compile data into text.</pre>
<pre>
Data::Rlist::write($thing, $filename);
$string_ref = Data::Rlist::write_string($thing);
$string = Data::Rlist::write_string_value($thing);</pre>
<pre>
### Print data to STDOUT.</pre>
<pre>
PrintData $thing;</pre>
<p>The object-oriented interface:</p>
<pre>
### For objects the '-output' attribute refers to a string buffer or is a filename.
### The '-data' attribute defines the value or reference to be compiled into text.</pre>
<pre>
$object = new Data::Rlist(-data => $thing, -output => \$target)</pre>
<pre>
$string_ref = $object->write; # compile into $target, return \$target
$string_ref = $object->write_string; # compile into new string ($target not touched)
$string = $object->write_string_value; # dto. but return string value</pre>
<pre>
### Print data to STDOUT.</pre>
<pre>
print $object->write_string_value;
print ${$object->write}; # returns \$target</pre>
<pre>
### Set output file and write $thing to disk.</pre>
<pre>
$object->set(-output => ".foorc");</pre>
<pre>
$object->write; # write "./.foorc", return 1
$object->write(".barrc"); # write "./.barrc" (the filename overrides -output)</pre>
<pre>
### The '-input' attribute defines the text to be compiled, either as
### string reference or filename.</pre>
<pre>
$object->set(-input => \$input_string); # assign some text</pre>
<pre>
$thing = $object->read; # parse $input_string into Perl data
$thing = $object->read($other_string); # parse $other_string (the argument overrides -input)</pre>
<pre>
$object->set(-input => ".foorc"); # assign some input file</pre>
<pre>
$foorc = $object->read; # parse ".foorc"
$barrc = $object->read(".barrc"); # parse some other file
$thing = $object->read(\$string); # parse some string buffer
$thing = $object->read_string($string_or_ref); # dto.</pre>
<p>Create deep-copies of any Perl data. The metaphor ``keelhaul'' vividly connotes that <em>$thing</em> is
stringified, then compiled back:</p>
<pre>
### Compile a value or ref $thing into text, then parse back into data.</pre>
<pre>
$reloaded = KeelhaulData $thing;
$reloaded = Data::Rlist::keelhaul($thing);</pre>
<pre>
$object = new Data::Rlist(-data => $thing);
$reloaded = $object->keelhaul;</pre>
<p>Do deep-comparisons of any Perl data:</p>
<pre>
### Deep-compare $a and $b and get a description of all type/value differences.</pre>
<pre>
@diffs = CompareData($a, $b);</pre>
<p>For more information see <em><a href="#item_compile">compile</a></em>, <em><a href="#item_keelhaul">keelhaul</a></em>, and <em><a href="#item_deep_compare">deep_compare</a></em>.</p>
<p>
</p>
<hr />
<h1><a name="description">DESCRIPTION</a></h1>
<p>
</p>
<h2><a name="venue">Venue</a></h2>
<p><em>Random-Lists</em> (Rlist) is a tag/value text format, which can ``stringify'' any data structure in
7-bit ASCII text. The basic types are lists and scalars. The syntax is similar, but not equal to
Perl's. For example,</p>
<pre>
( "hello", "world" )
{ "hello" = "world"; }</pre>
<p>designates two lists, the first of which is sequential, the second associative. The format...</p>
<p>- allows the definition of hierachical and constant data,</p>
<p>- has no user-defined types, no keywords, no variables,</p>
<p>- has no arithmetic expressions,</p>
<p>- uses 7-bit-ASCII character encoding and escape sequences,</p>
<p>- uses C-style numbers and strings,</p>
<p>- has an extremely minimal syntax implementable in any programming language and system.</p>
<p>You can write any Perl data structure into files as legible text. Like with CSV the lexical
overhead of Rlist is minimal: files are merely data.</p>
<p>You can read compiled texts back in Perl and C++ programs. No information will be lost between
different program languages, and floating-point numbers keep their precision.</p>
<p>You can also compile structured CSV text from Perl data, using special functions from this package
that will keep numbers precise and properly quote strings.</p>
<p>Since Rlist has no user-defined types the data is structured out of simple scalars and lists. It
is conceivable, however, to develop a simple type system and store type information along with the
actual data. Otherwise the data structures are tacit consents between the users of the data. See
also the implemenation notes for <a href="#perl">Perl</a> and <a href="#c__">C++</a>.</p>
<p>
</p>
<h2><a name="character_encoding">Character Encoding</a></h2>
<p>Rlist text uses the 7-bit-ASCII character set. The 95 printable character codes 32 to 126 occupy
one character. Codes 0 to 31 and 127 to 255 require four characters each: the <em>\</em> escape
character followed by the octal code number. For example, the German Umlaut character <em>ü</em>
(252) is translated into <em>\374</em>. An exception are the following codes:</p>
<pre>
ASCII ESCAPED AS
----- ----------
9 tab \t
10 linefeed \n
13 return \r
34 quote " \"
39 quote ' \'
92 backslash \ \\</pre>
<p>
</p>
<h2><a name="values_and_default_values">Values and Default Values</a></h2>
<p><em>Values</em> are either scalars, array elements or the value of a pair. Each value is constant.</p>
<p>The default scalar value is the empty string <code>""</code>. So in Perl <em>undef</em> is compiled into <code>""</code>.</p>
<p>
</p>
<h2><a name="numbers__strings_and_heredocuments">Numbers, Strings and Here-Documents</a></h2>
<p>Numbers constants adhere to the IEEE 754 syntax for integer- and floating-point numbers (i.e., the
same lexical conventions as in C and C++ apply).</p>
<p>Strings constants consisting only of <code>[a-zA-Z_0-9-/~:.@]</code> characters ``look like identifiers'' (aka
symbols) need not to be quoted. Otherwise string constants follow the C language lexicography.
They strings must be placed in double-quotes (single-quotes are not allowed). Quoted strings are
also escaped (i.e., characters are converted to the input character set of 7-bit ASCII).</p>
<p>You can define a string using a line-oriented form of quoting based on the UNIX shell
<em>here-document</em> syntax and RFC 111. Multiline quoted strings can be expressed with</p>
<pre>
<<DELIMITER</pre>
<p>Following the sigil <em> << </em> an identifier specifies how to terminate the string scalar. The value
of the scalar will be all lines following the current line down to the line starting with the
delimiter (i.e., the delimiter must be at column 1). There must be no space between the sigil and
the identifier.</p>
<p><strong>EXAMPLES</strong></p>
<p>Quoted strings:</p>
<pre>
"Hello, World!"</pre>
<p>Unquoted strings (symbols, identifiers):</p>
<pre>
foobar cogito.ergo.sum Memento::mori</pre>
<p>Here-document strings:</p>
<pre>
<<hamlet
"This above all: to thine own self be true". - (Act I, Scene III).
hamlet</pre>
<p>Integegers and floats:</p>
<pre>
38 10e-6 -.7 3.141592653589793</pre>
<p>For more information see <em><a href="#item_is_symbol">is_symbol</a></em>, <em><a href="#item_is_number">is_number</a></em> and <em><a href="#item_escape7">escape7</a></em>.</p>
<p>
</p>
<h2><a name="list_values">List Values</a></h2>
<p>We have two types of lists: sequential (aka array) and associative (aka map, hash, dictionary).</p>
<p><strong>EXAMPLES</strong></p>
<p>Arrays:</p>
<pre>
( 1, 2, ( 3, "Audiatur et altera pars!" ) )</pre>
<p>Maps:</p>
<pre>
{
key = value;
standalone-key;
Pi = 3.14159;</pre>
<pre>
"meta-syntactic names" = (foo, bar, "lorem ipsum", Acme, ___);</pre>
<pre>
var = {
log = {
messages = <<LOG;
Nov 27 21:55:04 localhost kernel: TSC appears to be running slowly. Marking it as unstable
Nov 27 22:34:27 localhost kernel: Uniform CD-ROM driver Revision: 3.20
Nov 27 22:34:27 localhost kernel: Loading iSCSI transport class v2.0-724.<6>PNP: No PS/2 controller found. Probing ports directly.
Nov 27 22:34:27 localhost kernel: wifi0: Atheros 5212: mem=0x26000000, irq=11
LOG
};
};
}</pre>
<p>
</p>
<h2><a name="binary_data">Binary Data</a></h2>
<p>Binary data can be represented as base64-encoded string, or <a href="#numbers__strings_and_heredocuments">here-document</a> string. For example,</p>
<pre>
use MIME::Base64;</pre>
<pre>
$str = encode_base64($binary_buf);</pre>
<p>The result <em>$str</em> will be a string broken into lines of no more than 76 characters each; the 76th
character will be a newline <code>"\n"</code>. Here is a complete Perl program that creates a file
<em>random.rls</em>:</p>
<pre>
use MIME::Base64;
use Data::Rlist;</pre>
<pre>
our $binary_data = join('', map { chr(int rand 256) } 1..300);
our $sample = { random_string => encode_base64($binary_data) };</pre>
<pre>
WriteData $sample, 'random.rls';</pre>
<p>These few lines create a file <em>random.rls</em> containing text like the following:</p>
<pre>
{
random_string = <<___
w5BFJIB3UxX/NVQkpKkCxEulDJ0ZR3ku1dBw9iPu2UVNIr71Y0qsL4WxvR/rN8VgswNDygI0xelb
aK3FytOrFg6c1EgaOtEudmUdCfGamjsRNHE2s5RiY0ZiaC5E5XCm9H087dAjUHPtOiZEpZVt3wAc
KfoV97kETH3BU8/bFGOqscCIVLUwD9NIIBWtAw6m4evm42kNhDdQKA3dNXvhbI260pUzwXiLYg8q
MDO8rSdcpL4Lm+tYikKrgCih9UxpWbfus+yHWIoKo/6tW4KFoufGFf3zcgnurYSSG2KRLKkmyEa+
s19vvUNmjOH0j1Ph0ZTi2pFucIhok4krJi0B5yNbQStQaq23v7sTqNom/xdRgAITROUIoel5sQIn
CqxenNM/M4uiUBV9OhyP
___
;
}</pre>
<p>Note that <em><a href="#item_writedata">WriteData</a></em> uses the predefined <code>"default"</code> configuration, which enables here-doc
strings. See also <a href="/MIME/Base64.html">the MIME::Base64 manpage</a>.</p>
<p>
</p>
<h2><a name="embedded_perl_code__nanoscripts_">Embedded Perl Code (Nanoscripts)</a></h2>
<p>Rlist text can define embedded Perl programs, called <em>nanonscripts</em>. The embedded program text
has the form of a <a href="#numbers__strings_and_heredocuments">here-document</a> with the special delimiter
<code>"perl"</code>. After the Rlist text has been parsed you call <em><a href="#item_evaluate_nanoscripts">evaluate_nanoscripts</a></em> to <em>eval</em>
all embedded Perl in the order of definiton. The function arranges it that within the <em>eval</em>...</p>
<ul>
<li>
<p>the <em>$root</em> variable refers to the root of the input, as unblessed array- or hash-reference;</p>
</li>
<li>
<p>the <em>$this</em> variable refers to the array or hash that stores the currently <em>eval</em>'d nanoscript;</p>
</li>
<li>
<p>the <em>$where</em> variable stores the name of the key, or the index, within <em>$this</em>.</p>
</li>
</ul>
<p>The nanoscript can use this information to oriented itself within the parsed data, or even to
modify the data in-place. The result of <em>eval</em>'ing will replace the nanoscript text. You can
also <em>eval</em> the embedded Perl codes programmatically, using the <em><a href="#item_nanoscripts">nanoscripts</a></em> and
<em><a href="#item_result">result</a></em> functions.</p>
<p><strong>EXAMPLES</strong></p>
<p>Simple example of an Rlist text that hosts Perl code:</p>
<pre>
(<<perl)
print "Hello, World!";
perl</pre>
<p>Here is a more complex example that defines a list of nanoscripts, and evaluates them:</p>
<pre>
use Data::Rlist;</pre>
<pre>
$data = join('', <DATA>);
$data = EvaluateData \$data;</pre>
<pre>
__END__
( <<perl, <<perl, <<perl, <<perl )
print "Hello World!\n" # english
perl
print "Hallo Welt!\n" # german
perl
print "Bonjour le monde!\n" # french
perl
print "Olá mundo!\n" # spanish
perl</pre>
<p>When we execute the above script the following output is printed before the script exits:</p>
<pre>
Hello World!
Hallo Welt!
Bonjour le monde!
Olá mundo!</pre>
<p>Note that when the Rlist text after <em>__END__</em> is placed in <em>some_file</em>, we can call
<em><a href="#item_evaluatedata">EvaluateData(<code>"some_file"</code>)</a></em> for the same effect. The next example modifies the parsed data
in place. Imagine a file <em>this_file_modifies_itself</em> with the following content:</p>
<pre>
( <<perl )
ReadData(\\'{ foo = bar; }');
perl</pre>
<p>When we parse this file using</p>
<pre>
$data = ReadData("this_file_modifies_itself");</pre>
<p>to <em>$data</em> will be assigned the following Perl value</p>
<pre>
[ "ReadData(\\'{ foo = bar; }');\n" ]</pre>
<p>Next we call <em>Data::Rlist::<a href="#item_evaluate_nanoscripts">evaluate_nanoscripts</a>()</em> to ``morph'' this value into</p>
<pre>
[ { 'foo' => 'bar' } ]</pre>
<p>The same effect can be achieved in just one call</p>
<pre>
$data = EvaluateData("this_file_modifies_itself");</pre>
<p>
</p>
<h2><a name="comments">Comments</a></h2>
<p>Rlist supports multiple forms of comments: <em>//</em> or <em>#</em> single-line-comments, and <em>/* */</em>
multi-line-comments. You may use all three forms at will.</p>
<p>
</p>
<hr />
<h1><a name="package_interface">PACKAGE INTERFACE</a></h1>
<p>The core functions to cultivate package objects are <em><a href="#item_new">new</a></em>, <em><a href="#item_dock">dock</a></em>, <em><a href="#item_set">set</a></em> and
<em><a href="#item_get">get</a></em>. When a regular package function is called in object context some omitted arguments are
read from object attributes. This is true for the following functions: <em><a href="#item_read">read</a></em>, <em><a href="#item_write">write</a></em>,
<em><a href="#item_read_string">read_string</a></em>, <em><a href="#item_write_string">write_string</a></em>, <em><a href="#item_read_csv">read_csv</a></em>, <em><a href="#item_write_csv">write_csv</a></em>, <em><a href="#item_read_conf">read_conf</a></em>,
<em><a href="#item_write_conf">write_conf</a></em> and <em><a href="#item_keelhaul">keelhaul</a></em>.</p>
<p>Unless called in object context the first argument has an indifferent meaning (i.e., it is no
<em>Data::Rlist</em> reference). Then <em><a href="#item_read">read</a></em> expects an input file or string, <em><a href="#item_write">write</a></em> the data
to compile etc.</p>
<p>
</p>
<h2><a name="construction">Construction</a></h2>
<dl>
<dt><strong><a name="item_new"><em>new([ATTRIBUTES])</em></a></strong>
<dd>
<p>Create a <em>Data::Rlist</em> object from the hash ATTRIBUTES. For example,</p>
</dd>
<dd>
<pre>
$self = Data::Rlist->new(-input => 'this.dat',
-data => $thing,
-output => 'that.dat');</pre>
</dd>
<dd>
<p>For this object the call <em><a href="#item_read">$self->read()</a></em> reads from <em>this.dat</em>, and
<em><a href="#item_write">$self->write()</a></em> writes any Perl data <em>$thing</em> to <em>that.dat</em>.</p>
</dd>
<dd>
<p><strong>REGULAR OBJECT ATTRIBUTES</strong></p>
</dd>
<dl>
<dt><strong><a name="item__2dinput__3d_3e_input"><code>-input => INPUT</code></a></strong>
<dt><strong><a name="item__2dfilter__3d_3e_filter"><code>-filter => FILTER</code></a></strong>
<dt><strong><a name="item__2dfilter_args__3d_3e_filter_2dargs"><code>-filter_args => FILTER-ARGS</code></a></strong>
<dd>
<p>Defines what Rlist text to parse and how to preprocess an input file. INPUT is a filename or
string reference. FILTER can be 1 to select the standard C preprocessor <em>cpp</em>. These attributes
are applied by <em><a href="#item_read">read</a></em>, <em><a href="#item_read_string">read_string</a></em>, <em><a href="#item_read_conf">read_conf</a></em> and <em><a href="#item_read_csv">read_csv</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__2ddata__3d_3e_data"><code>-data => DATA</code></a></strong>
<dt><strong><a name="item__2doptions__3d_3e_options"><code>-options => OPTIONS</code></a></strong>
<dt><strong><a name="item__2doutput__3d_3e_output"><code>-output => OUTPUT</code></a></strong>
<dd>
<p>Defines the Perl data to be <a href="#item_compile">compiled</a> into text (DATA), how it shall be compiled
(OPTIONS) and where to store the compiled text (OUTPUT). When OUTPUT is string reference the
compiled text will be stored in that string. When OUTPUT is <em>undef</em> a new string is created.
When OUTPUT is a string value it is a filename. These attributes are applied by <em><a href="#item_write">write</a></em>,
<em><a href="#item_write_string">write_string</a></em>, <em><a href="#item_write_conf">write_conf</a></em>, <em><a href="#item_write_csv">write_csv</a></em> and <em><a href="#item_keelhaul">keelhaul</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__2dheader__3d_3e_header"><code>-header => HEADER</code></a></strong>
<dd>
<p>Defines an array of text lines, each of which will by prefixed by a <em>#</em> and then written at the
top of the output file.</p>
</dd>
</li>
<dt><strong><a name="item__2ddelimiter__3d_3e_delimiter"><code>-delimiter => DELIMITER</code></a></strong>
<dd>
<p>Defines the field delimiter for <em>.csv</em>-files. Applied by <em><a href="#item_read_csv">read_csv</a></em> and <em><a href="#item_read_conf">read_conf</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__2dcolumns__3d_3e_strings"><code>-columns => STRINGS</code></a></strong>
<dd>
<p>Defines the column names for <em>.csv</em>-files to be written into the first line.</p>
</dd>
</li>
</dl>
<p><strong>ATTRIBUTES THAT MASQUERADE PACKAGE GLOBALS</strong></p>
<p>The attributes listed below raise new values for package globals for the time an object method
runs.</p>
<dl>
<dt><strong><a name="item__2dinputrecordseparator__3d_3e_flag"><code>-InputRecordSeparator => FLAG</code></a></strong>
<dd>
<p>Masquerades <em>$/</em>, which affects how lines are read and written to and from Rlist- and CSV-files.
You may also set <em>$/</em> by yourself. See <em>perlport</em> and <em>perlvar</em>.</p>
</dd>
</li>
<dt><strong><a name="item__2dmaxdepth__3d_3e_integer"><code>-MaxDepth => INTEGER</code></a></strong>
<dt><strong><a name="item__2dsafecppmode__3d_3e_flag"><code>-SafeCppMode => FLAG</code></a></strong>
<dt><strong><a name="item__2droundscientific__3d_3e_flag"><code>-RoundScientific => FLAG</code></a></strong>
<dd>
<p>Masquerade <em><a href="#debugging_data">$Data::Rlist::MaxDepth</a></em>, <em><a href="#item_open_input">$Data::Rlist::SafeCppMode</a></em>
and <em><a href="#item_round">$Data::Rlist::RoundScientific</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__2dechostderr__3d_3e_flag"><code>-EchoStderr => FLAG</code></a></strong>
<dd>
<p>Print read errors and warnings message on STDERR (default: off).</p>
</dd>
</li>
<dt><strong><a name="item__2ddefaultcsvdelimiter__3d_3e_regex"><code>-DefaultCsvDelimiter => REGEX</code></a></strong>
<dt><strong><a name="item__2ddefaultconfdelimiter__3d_3e_regex"><code>-DefaultConfDelimiter => REGEX</code></a></strong>
<dd>
<p>Masquerades <em>$Data::Rlist::DefaultCsvDelimiter</em> and <em>$Data::Rlist::DefaultConfDelimiter</em>. These
globals define the default regexes to use when the <em>-options</em> attribute does not specifiy the
<a href="#compile_options"><code>"delimiter"</code></a> regex. Applied by <em><a href="#item_read_csv">read_csv</a></em> and <em><a href="#item_read_conf">read_conf</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__2ddefaultconfseparator__3d_3e_string"><code>-DefaultConfSeparator => STRING</code></a></strong>
<dd>
<p>Masquerades <em>$Data::Rlist::DefaultConfSeparator</em>, the default string to use when the <em>-options</em>
attribute does not specifiy the <a href="#compile_options"><code>"separator"</code></a> string. Applied by
<em><a href="#item_write_conf">write_conf</a></em>.</p>
</dd>
</li>
</dl>
<dt><strong><a name="item_dock"><em>dock(SELF, SUB)</em></a></strong>
<dd>
<p>Localize object SELF within the package and run SUB. This means that some of SELF's attribute
masqquerade few package globals for the time SUB runs. SELF then locks the package, and
<em>$Data::Rlist::Locked</em> is greater than 0.</p>
</dd>
</li>
</dl>
<p>
</p>
<h2><a name="attribute_access">Attribute Access</a></h2>
<dl>
<dt><strong><a name="item_set"><em>set(SELF[, ATTRIBUTE]...)</em></a></strong>
<dd>
<p>Reset or initialize object attributes, then return SELF. Each ATTRIBUTE is a name/value-pair. See
<em><a href="#item_new">new</a></em> for a list of valid names. For example,</p>
</dd>
<dd>
<pre>
$obj->set(-input => \$str, -output => 'temp.rls', -options => 'squeezed');</pre>
</dd>
</li>
<dt><strong><a name="item_get"><em>get(SELF, NAME[, DEFAULT])</em></a></strong>
<dt><strong><a name="item_require"><em>require(SELF[, NAME])</em></a></strong>
<dt><strong><a name="item_has"><em>has(SELF[, NAME])</em></a></strong>
<dd>
<p>Get some attribute NAME from object SELF. Unless NAME exists returns DEFAULT. The <em>require</em>
method has no default value, hence it dies unless NAME exists. <em>has</em> returns true when NAME
exists, false otherwise. For NAME the leading hyphen is optional. For example,</p>
</dd>
<dd>
<pre>
$self->get('foo'); # returns $self->{-foo} or undef
$self->get(-foo=>); # dto.
$self->get('foo', 42); # returns $self->{-foo} or 42</pre>
</dd>
</li>
</dl>
<p>
</p>
<h2><a name="public_functions">Public Functions</a></h2>
<dl>
<dt><strong><a name="item_read"><em>read(INPUT[, FILTER, FILTER-ARGS])</em></a></strong>
<dd>
<p>Parse data from INPUT, which specifies some Rlist-text. See also <em><a href="#item_errors">errors</a></em>, <em><a href="#item_write">write</a></em>.</p>
</dd>
<dd>
<p><strong>PARAMETERS</strong></p>
</dd>
<dd>
<p>INPUT shall be either</p>
</dd>
<dd>
<p>- some Rlist object created by <em><a href="#item_new">new</a></em>,</p>
</dd>
<dd>
<p>- a string reference, in which case <em>read</em> and <em><a href="#item_read_string">read_string</a></em> parse Rlist text from it,</p>
</dd>
<dd>
<p>- a string scalar, in which case <em>read</em> assumes a file to parse.</p>
</dd>
<dd>
<p>See <em><a href="#item_open_input">open_input</a></em> for the FILTER and FILTER-ARGS parameters, which are used to preprocess an
input file. When an input file cannot be <em>open</em>'d and <em>flock</em>'d this function dies. When INPUT
is an object, arguments for FILTER and FILTER-ARGS eventually override the <em>-filter</em> and
<em>-filter_args</em> attributes.</p>
</dd>
<dd>
<p><strong>RESULT</strong></p>
</dd>
<dd>
<p>The parsed data as array- or hash-reference, or <em>undef</em> if there was no data. The latter may also
be the case when file consist only of comments/whitespace.</p>
</dd>
<dd>
<p><strong>NOTES</strong></p>
</dd>
<dd>
<p>This function may die. Dying is Perl's mechanism to raise exceptions, which eventually can be
catched with <em>eval</em>. For example,</p>
</dd>
<dd>
<pre>
my $host = eval { use Sys::Hostname; hostname; } || 'some unknown machine';</pre>
</dd>
<dd>
<p>This code fragment traps the <em>die</em> exception, so that <em>eval</em> returns <em>undef</em> or the result of
calling <em>hostname</em>. The following example uses <em>eval</em> to trap exceptions thrown by <em>read</em>:</p>
</dd>
<dd>
<pre>
$object = new Data::Rlist(-input => $thingfile);
$thing = eval { $object->read };</pre>
</dd>
<dd>
<pre>
unless (defined $thing) {
if ($object->errors) {
print STDERR "$thingfile has syntax errors"
} else {
print STDERR "$thingfile not found, is locked or empty"
}
} else {
# Can use $thing
.
.
}</pre>
</dd>
</li>
<dt><strong><a name="item_read_csv"><em>read_csv(INPUT[, OPTIONS, FILTER, FILTER-ARGS])</em></a></strong>
<dt><strong><a name="item_read_conf"><em>read_conf(INPUT[, OPTIONS, FILTER, FILTER-ARGS])</em></a></strong>
<dd>
<p>Parse data from INPUT, which specifies some comma-separated-values (CSV) text. Both functions</p>
</dd>
<dd>
<p>- read data from strings or files,</p>
</dd>
<dd>
<p>- use an optional delimiter,</p>
</dd>
<dd>
<p>- ignore delimiters in quoted strings,</p>
</dd>
<dd>
<p>- ignore empty lines,</p>
</dd>
<dd>
<p>- ignore lines begun with <em>#</em>.</p>
</dd>
<dd>
<p><em>read_conf</em> is a variant of <em>read_csv</em> dedicated to configuration files. Such files consist
of lines of the form</p>
</dd>
<dd>
<pre>
key = value</pre>
</dd>
<dd>
<p><strong>PARAMETERS</strong></p>
</dd>
<dd>
<p>For INPUT see <em><a href="#item_read">read</a></em>. For FILTER, FILTER-ARGS see <em><a href="#item_open_input">open_input</a></em>.</p>
</dd>
<dd>
<p>OPTIONS can be used to override the <a href="#compile_options"><code>"delimiter"</code></a> regex. For example, a
delimiter of <code>'\s+'</code> splits the line at horizontal whitespace into multiple values (with respect
of quoted strings). For <em>read_csv</em> the delimiter defaults to <code>'\s*,\s*'</code>, and for <em>read_conf</em>
to <code>'\s*=\s*'</code>. See also <em><a href="#item_write_csv">write_csv</a></em> and <em><a href="#item_write_conf">write_conf</a></em>.</p>
</dd>
<dd>
<p><strong>RESULT</strong></p>
</dd>
<dd>
<p>Both functions return a list of lists. Each embedded array defines the fields in a line.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<p>Un/quoting of values happens implicitly. Given a file <em>db.conf</em></p>
</dd>
<dd>
<pre>
# Comment
SERVER = hostname
DATABASE = database_name
LOGIN = "user,password"</pre>
</dd>
<dd>
<p>the call <em>$opts=ReadConf(<code>"db.conf"</code>)</em> assigns</p>
</dd>
<dd>
<pre>
[ [ 'SERVER', 'hostname' ],
[ 'DATABASE', 'database_name' ],
[ 'LOGIN', 'user,password' ]
]</pre>
</dd>
<dd>
<p>The <em><a href="#item_writeconf">WriteConf</a></em> function can be used to create or update the configuration:</p>
</dd>
<dd>
<pre>
push @$opts, [ 'MAGIC VALUE' => 3.14_15 ];</pre>
</dd>
<dd>
<pre>
WriteConf('db.conf', { precision => 2 });</pre>
</dd>
<dd>
<p>This writes to <em>db.conf</em>:</p>
</dd>
<dd>
<pre>
SERVER = hostname
DATABASE = database_name
LOGIN = "user,password"
"MAGIC VALUE" = 3.14</pre>
</dd>
</li>
<dt><strong><a name="item_read_string"><em>read_string(INPUT)</em></a></strong>
<dd>
<p>Calls <em><a href="#item_read">read</a></em> to parse Rlist language productions from the string or string-reference INPUT.
When INPUT is an object do this for its <em>-input</em> attribute.</p>
</dd>
</li>
<dt><strong><a name="item_result"><em>result([SELF])</em></a></strong>
<dd>
<p>Return the last result of calling <em><a href="#item_read">read</a></em>, which is either <em>undef</em> or some array- or
hash-reference. When SELF is passed as object reference, returns the result that occured the last
time SELF had called <em><a href="#item_read">read</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_nanoscripts"><em>nanoscripts([SELF])</em></a></strong>
<dd>
<p>In list context return an array of nanoscripts defined by the last call to <em><a href="#item_read">read</a></em>. When SELF
is passed return this information for the last time SELF had called <em><a href="#item_read">read</a></em>. The result has the
form:</p>
</dd>
<dd>
<pre>
( [ $hash_or_array_ref, $key_or_index ], # 1st nanoscript
[ $hash_or_array_ref, $key_or_index ], # 2nd nanoscript
.
.
.
)</pre>
</dd>
<dd>
<p>In scalar context return a reference to the above. This information defines the location of all
embedded Perl scripts within the result, and can be used to <em>eval</em> them programmatically. See
also <em><a href="#item_result">result</a></em>, <em><a href="#item_evaluate_nanoscripts">evaluate_nanoscripts</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_evaluate_nanoscripts"><em>evaluate_nanoscripts([SELF])</em></a></strong>
<dd>
<p>Evaluates all nanoscripts defined by the last call to <em><a href="#item_read">read</a></em>. When called as method evaluates
the nanoscripts defined by the last time SELF had called <em><a href="#item_read">read</a></em>. Returns the number of
scripts or 0 if none were available. Each script is replaced by the result of <em>eval</em>'ing it.
(For details and examples see <a href="#embedded_perl_code__nanoscripts_">Embedded Perl Code (Nanoscripts)</a>.)</p>
</dd>
</li>
<dt><strong><a name="item_messages"><em>messages([SELF])</em></a></strong>
<dd>
<p>In list context returns a list of compile-time messages that occurred in the last call to
<em><a href="#item_read">read</a></em>. In scalar context returns an array reference. When an package object SELF is passed
returns the information for the last time SELF had called <em><a href="#item_read">read</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_errors"><em>errors([SELF])</em></a></strong>
<dt><strong><a name="item_warnings"><em>warnings([SELF])</em></a></strong>
<dd>
<p>Returns the number of syntax errors and warnings that occurred in the last call to <em><a href="#item_read">read</a></em>.
When called as method returns the number that occured the last time SELF had called <em><a href="#item_read">read</a></em>.</p>
</dd>
<dd>
<p>Example:</p>
</dd>
<dd>
<pre>
use Data::Rlist;</pre>
</dd>
<dd>
<pre>
our $data = ReadData 'things.rls';</pre>
</dd>
<dd>
<pre>
if (Data::Rlist::errors() || Data::Rlist::warnings()) {
print join("\n", Data::Rlist::messages())
} else {
# Ok, $data is an array- or hash-reference.
die unless $data;</pre>
</dd>
<dd>
<pre>
}</pre>
</dd>
</li>
<dt><strong><a name="item_broken"><em>broken([SELF])</em></a></strong>
<dd>
<p>Returns the number of times the last <em><a href="#item_compile">compile</a></em> violated <em><a href="#debugging_data">$Data::Rlist::MaxDepth</a></em>. When called as method returns the information for the last time SELF had called
<em><a href="#item_compile">compile</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_missing_input"><em>missing_input([SELF])</em></a></strong>
<dd>
<p>Returns true when the last call to <em><a href="#item_parse">parse</a></em> yielded <em>undef</em>, because there was nothing to
parse. When called as method returns the information for the last time SELF had called
<em><a href="#item_parse">parse</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_write"><em>write(DATA[, OUTPUT, OPTIONS, HEADER])</em></a></strong>
<dd>
<p>Transliterates Perl data into Rlist text and write the text to a file or string buffer. <em>write</em>
is auto-exported as <em><a href="#item_writedata">WriteData</a></em>.</p>
</dd>
<dd>
<p><strong>PARAMETERS</strong></p>
</dd>
<dd>
<p>DATA is either an object generated by <em><a href="#item_new">new</a></em>, or any Perl data including <em>undef</em>. In case of
an object the actual DATA value is defined by its <em>-data</em> attribute. (When <em>-data</em> refers to
another Rlist object, this other object is invoked.)</p>
</dd>
<dd>
<p>OUTPUT defines the output location, as filename, string-reference or <em>undef</em>. When <em>undef</em> the
function allocates a string and returns a reference to it. OUTPUT defaults to the <em>-output</em>
attribute when DATA defines an object.</p>
</dd>
<dd>
<p>OPTIONS define how to compile DATA: when <em>undef</em> or <code>"fast"</code> uses <em><a href="#item_compile_fast">compile_fast</a></em>, when
<code>"perl"</code> uses <em><a href="#item_compile_perl">compile_Perl</a></em>, otherwise <em><a href="#item_compile">compile</a></em>. Defaults to the <em>-options</em>
attribute when DATA is an object.</p>
</dd>
<dd>
<p>HEADER is a reference to an array of strings that shall be printed literally at the top of an
output file. Defaults to the <em>-header</em> attribute when DATA is an object.</p>
</dd>
<dd>
<p><strong>RESULT</strong></p>
</dd>
<dd>
<p>When <em>write</em> creates a file it returns 0 for failure or 1 for success. Otherwise it returns a
string reference.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<pre>
$self = new Data::Rlist(-data => $thing, -output => $output);</pre>
</dd>
<dd>
<pre>
$self->write; # Compile $thing into a file ($output is a filename)
# or string ($output is a string reference).</pre>
</dd>
<dd>
<pre>
Data::Rlist::write($thing, $output); # dto., but using the functional interface.</pre>
</dd>
</li>
<dt><strong><a name="item_write_csv"><em>write_csv(DATA[, OUTPUT, OPTIONS, COLUMNS])</em></a></strong>
<dt><strong><a name="item_write_conf"><em>write_conf(DATA[, OUTPUT, OPTIONS, HEADER])</em></a></strong>
<dd>
<p>Write DATA as comma-separated-values (CSV) to file or string OUTPUT. <em>write_conf</em> writes
configuration files where each line contains a tagname, a separator and a value.</p>
</dd>
<dd>
<p><strong>PARAMETERS</strong></p>
</dd>
<dd>
<p>DATA is either an object, or defines the data to be compiled as reference to an array of arrays.
<em>write_conf</em> uses only the first and second fields. For example,</p>
</dd>
<dd>
<pre>
[ [ a, b, c ], # fields of line 1
[ d, e, f, g ], # fields line 2
.
.
]</pre>
</dd>
<dd>
<p></p>
</dd>
<dd>
<p>OPTIONS specifies the comma-separator (<code>"separator"</code>), how to quote (<code>"auto_quote"</code>), the
linefeed (<code>"eol_space"</code>) and the numeric precision (<code>"precision"</code>). COLUMNS specifies the column
names to be written to the first line. Likewise the text from the HEADER array is written in form
of <em>#</em>-comments at the top of an output file.</p>
</dd>
<dd>
<p><strong>RESULT</strong></p>
</dd>
<dd>
<p>When a file was created both function return 0 for failure, or 1 for success. Otherwise they
return a reference to the compiled text.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<p>Functional interface:</p>
</dd>
<dd>
<pre>
use Data::Rlist; # imports WriteCSV</pre>
</dd>
<dd>
<pre>
WriteCSV($thing, "foo.dat");</pre>
</dd>
<dd>
<pre>
WriteCSV($thing, "foo.dat", { separator => '; ' }, [qw/GBKNR VBKNR EL LaD/]);</pre>
</dd>
<dd>
<pre>
WriteCSV($thing, \$target_string);</pre>
</dd>
<dd>
<pre>
$string_ref = WriteCSV($thing);</pre>
</dd>
<dd>
<p>Object-oriented interface:</p>
</dd>
<dd>
<pre>
$object = new Data::Rlist(-data => $thing, -output => "foo.dat",
-options => { separator => '; ' },
-columns => [qw/GBKNR VBKNR EL LaD LaD_V/]);</pre>
</dd>
<dd>
<pre>
$object->write_csv; # write $thing as CSV to foo.dat
$object->write; # write $thing as Rlist to foo.dat</pre>
</dd>
<dd>
<pre>
$object->set(-output => \$target_string);</pre>
</dd>
<dd>
<pre>
$object->write_csv; # write $thing as CSV to $target_string</pre>
</dd>
<dd>
<p>See also <em><a href="#item_write">write</a></em> and <em><a href="#item_read_csv">read_csv</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_write_string"><em>write_string(DATA[, OPTIONS])</em></a></strong>
<dd>
<p>Stringify any Perl data and return a reference to the string. Works like <em><a href="#item_write">write</a></em> but always
compiles to a new string to which it returns a reference. The default for OPTIONS will be
<a href="#predefined_options"><code>"string"</code></a>.</p>
</dd>
</li>
<dt><strong><a name="item_write_string_value"><em>write_string_value(DATA[, OPTIONS])</em></a></strong>
<dd>
<p>Stringify any Perl dats and return the compiled text string value. OPTIONS default to
<a href="#predefined_options"><code>"default"</code></a>. For example,</p>
</dd>
<dd>
<pre>
print "\n\$thing dumped: ", Data::Rlist::write_string_value($thing);</pre>
</dd>
<dd>
<pre>
$self = new Data::Rlist(-data => $thing);</pre>
</dd>
<dd>
<pre>
print "\nsame \$thing dumped: ", $self->write_string_value;</pre>
</dd>
</li>
<dt><strong><a name="item_keelhaul"><em>keelhaul(DATA[, OPTIONS])</em></a></strong>
<dd>
<p>Do a deep copy of DATA according to <a href="#compile_options">OPTIONS</a>. First the function compiles DATA
to Rlist text, then restores the data from exactly this text. This process is called ``keelhauling
data'', and allows us to</p>
</dd>
<dd>
<p>- adjust the accuracy of numbers,</p>
</dd>
<dd>
<p>- break circular-references,</p>
</dd>
<dd>
<p>- drop <em>\*foo{THING}</em>s,</p>
</dd>
<dd>
<p>- bring multiple data sets to the same, common basis.</p>
</dd>
<dd>
<p>It is useful (e.g.) when DATA had been hatched by some other code, and you don't know whether it
is hierachical, or if typeglob-refs nist inside. Then keelhaul it to clean it from its past. For
example, to bring all numbers in</p>
</dd>
<dd>
<pre>
$thing = { foo => [ [ .00057260 ], -1.6804e-4 ] };</pre>
</dd>
<dd>
<p>to a certain accuracy, use</p>
</dd>
<dd>
<pre>
$deep_copy_of_thing = Data::Rlist::keelhaul($thing, { precision => 4 });</pre>
</dd>
<dd>
<p>All number scalars in <em>$thing</em> are rounded to 4 decimal places, so they're finally comparable as
floating-point numbers. To <em>$deep_copy_of_thing</em> is assigned the hash-reference</p>
</dd>
<dd>
<pre>
{ foo => [ [ 0.0006 ], -0.0002 ] }</pre>
</dd>
<dd>
<p>Likewise one can convert all floats to integers:</p>
</dd>
<dd>
<pre>
$make_integers = new Data::Rlist(-data => $thing, -options => { precision => 0 });</pre>
</dd>
<dd>
<pre>
$thing_without_floats = $make_integers->keelhaul;</pre>
</dd>
<dd>
<p>When <em><a href="#item_keelhaul">keelhaul</a></em> is called in an array context it also returns the text from which the copy had
been built. For example,</p>
</dd>
<dd>
<pre>
$deep_copy = Data::Rlist::keelhaul($thing);</pre>
</dd>
<dd>
<pre>
($deep_copy, $rlist_text) = Data::Rlist::keelhaul($thing);</pre>
</dd>
<dd>
<pre>
$deep_copy = new Data::Rlist(-data => $thing)->keelhaul;</pre>
</dd>
<dd>
<p><strong>DETAILS</strong></p>
</dd>
<dd>
<p><em><a href="#item_keelhaul">keelhaul</a></em> won't throw <em>die</em> nor return an error, but be prepared for the following effects:</p>
</dd>
<ul>
<li>
<p><em>ARRAY</em>, <em>HASH</em>, <em>SCALAR</em> and <em>REF</em> references were compiled, whether blessed or not. (Since
compiling does not store type information, <em>keelhaul</em> will turn blessed references into barbars
again.)</p>
</li>
<li>
<p><em>IO</em>, <em>GLOB</em> and <em>FORMAT</em> references have been converted into strings.</p>
</li>
<li>
<p>Depending on the compile options, <em>CODE</em> references are invoked, deparsed back into their function
bodies, or dropped.</p>
</li>
<li>
<p>Depending on the compile options floats are rounded, or are converted to integers.</p>
</li>
<li>
<p><em>undef</em>'d array elements are converted into the default scalar value <code>""</code>.</p>
</li>
<li>
<p>Unless <em>$Data::Rlist::MaxDepth</em> is 0, anything deeper than <em>$Data::Rlist::MaxDepth</em> will be
thrown away.</p>
</li>
<li>
<p>When the data contains objects, no special methods are triggered to ``freeze'' and ``thaw'' the
objects.</p>
</li>
</ul>
<p>See also <em><a href="#item_compile">compile</a></em> and <em><a href="#item_deep_compare">deep_compare</a></em></p>
</dl>
<p>
</p>
<h2><a name="static_functions">Static Functions</a></h2>
<dl>
<dt><strong><a name="item_predefined_options"><em>predefined_options([PREDEF-NAME])</em></a></strong>
<dd>
<p>Return are predefined hash-reference of compile otppns. PREDEF-NAME defaults to
<a href="#predefined_options"><code>"default"</code></a>.</p>
</dd>
</li>
<dt><strong><a name="item_complete_options"><em>complete_options([OPTIONS[, BASICS]])</em></a></strong>
<dd>
<p>Completes OPTIONS with BASICS, so that all pairs not already in OPTIONS are copied from BASICS.
Always returns a new hash-reference, i.e., neither OPTIONS nor BASICS are modified. Both arguments
define hashes or some <a href="#predefined_options">predefined options name</a>. BASICS defaults to
<a href="#predefined_options"><code>"default"</code></a>. For example,</p>
</dd>
<dd>
<pre>
$options = complete_options({ precision => 0 }, 'squeezed')</pre>
</dd>
<dd>
<p>merges the predefined options for <a href="#predefined_options"><code>"squeezed"</code> text</a> with a numeric
precision of 0 (converts all floats to integers).</p>
</dd>
</li>
</dl>
<p>
</p>
<h2><a name="implementation_functions">Implementation Functions</a></h2>
<dl>
<dt><strong><a name="item_open_input"><em>open_input(INPUT[, FILTER, FILTER-ARGS])</em></a></strong>
<dt><strong><a name="item_close_input"><em>close_input</em></a></strong>
<dd>
<p>Open/close Rlist text file or string INPUT for parsing. Used internally by <em><a href="#item_read">read</a></em> and
<em><a href="#item_read_csv">read_csv</a></em>.</p>
</dd>
<dd>
<p><strong>PREPROCESSING</strong></p>
</dd>
<dd>
<p>The function can preprocess the INPUT file using FILTER. Use the special value 1 to select the
default C preprocessor (<em>gcc -E -Wp,-C</em>). FILTER-ARGS is an optional string of additional
command-line arguments to be appended to FILTER. For example,</p>
</dd>
<dd>
<pre>
my $foo = Data::Rlist::read("foo", 1, "-DEXTRA")</pre>
</dd>
<dd>
<p>eventually does not parse <em>foo</em>, but the output of the command</p>
</dd>
<dd>
<pre>
gcc -E -Wp,-C -DEXTRA foo</pre>
</dd>
<dd>
<p>Hence within <em>foo</em> now C-preprocessor-statements are allowed. For example,</p>
</dd>
<dd>
<pre>
{
#ifdef EXTRA
#include "extra.rlist"
#endif</pre>
</dd>
<dd>
<pre>
123 = (1, 2, 3);
foobar = {
.
.</pre>
</dd>
<dd>
<p><strong>SAFE CPP MODE</strong></p>
</dd>
<dd>
<p>This mode uses <em>sed</em> and a temporary file. It is enabled by setting <em>$Data::Rlist::SafeCppMode</em>
to 1 (the default is 0). It protects single-line <em>#</em>-comments when FILTER begins with either
<em>gcc</em>, <em>g++</em> or <em>cpp</em>. <em><a href="#item_open_input">open_input</a></em> then additionally runs <em>sed</em> to convert all input
lines beginning with whitespace plus the <em>#</em> character. Only the following <em>cpp</em>-commands are
excluded, and only when they appear in column 1:</p>
</dd>
<dd>
<p>- <em>#include</em> and <em>#pragma</em></p>
</dd>
<dd>
<p>- <em>#define</em> and <em>#undef</em></p>
</dd>
<dd>
<p>- <em>#if</em>, <em>#ifdef</em>, <em>#else</em> and <em>#endif</em>.</p>
</dd>
<dd>
<p>For all other lines <em>sed</em> converts <em>#</em> into <em>##</em>. This prevents the C preprocessor from
evaluating them. Because of Perl's limited <em>open</em> function, which isn't able to dissolve long
pipes, the invocation of <em>sed</em> requires a temporary file. The temporary file is created in the
same directory as the input file. When you only use <em>//</em> and <em>/* */</em> comments, however, this
read mode is not required.</p>
</dd>
</li>
<dt><strong><a name="item_lex"><em>lex()</em></a></strong>
<dd>
<p>Lexical scanner. Called by <em><a href="#item_parse">parse</a></em> to split the current line into tokens. <em>lex</em> reads <em>#</em>
or <em>//</em> single-line-comment and <em>/* */</em> multi-line-comment as regular white-spaces. Otherwise it
returns tokens according to the following table:</p>
</dd>
<dd>
<pre>
RESULT MEANING
------ -------
'{' '}' Punctuation
'(' ')' Punctuation
',' Operator
';' Punctuation
'=' Operator
'v' Constant value as number, string, list or hash
'??' Error
undef EOF</pre>
</dd>
<dd>
<p><em>lex</em> appends all here-doc-lines with a newline character. For example,</p>
</dd>
<dd>
<pre>
<<test1
a
b
test1</pre>
</dd>
<dd>
<p>is effectively read as <code>"a\nb\n"</code>, which is the same value as the equivalent here-doc in Perl has.
So, not all strings can be encoded as a here-doc. For example, it might not be quite obvious to
many programmers that <code>"foo\nbar"</code> cannot be expressed as here-doc.</p>
</dd>
</li>
<dt><strong><a name="item_lexln"><em>lexln()</em></a></strong>
<dd>
<p>Read the next line of text from the current input. Return 0 if <em><a href="#item_at_eof">at_eof</a></em>, otherwise return 1.</p>
</dd>
</li>
<dt><strong><a name="item_at_eof"><em>at_eof()</em></a></strong>
<dd>
<p>Return true if current input file/string is exhausted, false otherwise.</p>
</dd>
</li>
<dt><strong><a name="item_parse"><em>parse()</em></a></strong>
<dd>
<p>Read Rlist language productions from current input. This is a fast, non-recursive parser driven by
the parser map <em>%Data::Rlist::Rules</em>, and fed by <em><a href="#item_lex">lex</a></em>. It is called internally by
<em><a href="#item_read">read</a></em>. <em>parse</em> returns an array- or hash-reference, or <em>undef</em> in case of parsing
<em><a href="#item_errors">errors</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_compile"><em>compile(DATA[, OPTIONS, FH])</em></a></strong>
<dd>
<p>Build Rlist text from DATA:</p>
</dd>
<ul>
<li>
<p>Reference-types <em>SCALAR</em>, <em>HASH</em>, <em>ARRAY</em> and <em>REF</em> are compiled into text, whether blessed or
not.</p>
</li>
<li>
<p>Reference-types <em>CODE</em> are compiled depending on the <a href="#compile_options"><code>"code_refs"</code></a> setting in
OPTIONS.</p>
</li>
<li>
<p>Reference-types <em>GLOB</em> (<a href="#a_short_story_of_typeglobs">typeglob-refs</a>), <em>IO</em> and <em>FORMAT</em> (file-
and directory handles) cannot be dissolved, and are compiled into the strings <code>"?GLOB?"</code>,
<code>"?IO?"</code> and <code>"?FORMAT?"</code>.</p>
</li>
<li>
<p><em>undef</em>'d values in arrays are compiled into the default Rlist <code>""</code>.</p>
</li>
</ul>
<p>When FH is defined compile directly to this file and return 1. Otherwise build a string and return
a reference to it. This is the compilation function called when the OPTIONS argument passed to
<em><a href="#item_write">write</a></em> is not omitted, and is not <code>"fast"</code> or <code>"perl"</code>.</p>
<dt><strong><a name="item_compile_fast"><em>compile_fast(DATA)</em></a></strong>
<dd>
<p>Build Rlist text from DATA, as fast as actually possible with pure Perl:</p>
</dd>
<ul>
<li>
<p>Reference-types <em>SCALAR</em>, <em>HASH</em>, <em>ARRAY</em> and <em>REF</em> are compiled into text, whether blessed or
not.</p>
</li>
<li>
<p><em>CODE</em>, <em>GLOB</em>, <em>IO</em> and <em>FORMAT</em> are compiled into the strings <code>"?CODE?"</code>, <code>"?IO?"</code>,
<code>"?GLOB?"</code> and <code>"?FORMAT?"</code>.</p>
</li>
<li>
<p><em>undef</em>'d values in arrays are compiled into the default Rlist <code>""</code>.</p>
</li>
</ul>
<p><em><a href="#item_compile_fast">compile_fast</a></em> is the default compilation function. It is called when you pass <em>undef</em> or
<code>"fast"</code> in place of the OPTIONS parameter (see <em><a href="#item_write">write</a></em>, <em><a href="#item_write_string">write_string</a></em>). Since
<em><a href="#item_compile_fast">compile_fast</a></em> considers no compile options it will not call code, round numbers, detect
self-referential data etc. Also <em><a href="#item_compile_fast">compile_fast</a></em> always compiles into a unique package variable
to which it returns a reference.</p>
<dt><strong><a name="item_compile_perl"><em>compile_Perl(DATA)</em></a></strong>
<dd>
<p>Like <em><a href="#item_compile_fast">compile_fast</a></em>, but do not compile Rlist text - compile DATA into Perl syntax. It can
then be <em>eval</em>'d. This renders more compact, and more exact output as <a href="/Data/Dumper.html">the Data::Dumper manpage</a>. For
example, only strings are quoted. To enable this compilation function pass <code>"perl"</code> to as the
OPTIONS argument, or set the <em>-options</em> attribute of package objects to this string.</p>
</dd>
</li>
</dl>
<p>
</p>
<h2><a name="auxiliary_functions">Auxiliary Functions</a></h2>
<p>The utility functions in this section are generally useful when handling stringified data.
Internally <em><a href="#item_quote7">quote7</a></em>, <em><a href="#item_escape7">escape7</a></em>, <em><a href="#item_is_integer">is_integer</a></em> etc. apply precompiled regexes and
precomputed ASCII tables. <em><a href="#item_split_quoted">split_quoted</a></em> and <em><a href="#item_parse_quoted">parse_quoted</a></em> simplify
<a href="#text__parsewords">Text::ParseWords</a>. <em><a href="#item_round">round</a></em> and <em><a href="#item_equal">equal</a></em> are working solutions for floating-point
numbers. <em><a href="#item_deep_compare">deep_compare</a></em> is a smart function to ``diff'' two Perl variables. All these
functions are very fast and mature.</p>
<dl>
<dt><strong><a name="item_is_integer"><em>is_integer(SCALAR-REF)</em></a></strong>
<dd>
<p>Returns true when a scalar looks like a positive or negative integer constant. The function
applies the compiled regex <em>$Data::Rlist::REInteger</em>.</p>
</dd>
</li>
<dt><strong><a name="item_is_number"><em>is_number(SCALAR-REF)</em></a></strong>
<dd>
<p>Test for strings that look like numbers. <em>is_number</em> can be used to test whether a scalar looks
like a integer/float constant (numeric literal). The function applies the compiled regex
<em>$Data::Rlist::REFloat</em>. Note that it doesn't match</p>
</dd>
<dd>
<p>- leading or trailing whitespace,</p>
</dd>
<dd>
<p>- lexical conventions such as the <code>"0b"</code> (binary), <code>"0"</code> (octal), <code>"0x"</code> (hex) prefix to denote
a number-base other than decimal, and</p>
</dd>
<dd>
<p>- Perls' legible numbers, e.g. <em>3.14_15_92</em>,</p>
</dd>
<dd>
<p>- the IEEE 754 notations of Infinite and NaN.</p>
</dd>
<dd>
<p>See also</p>
</dd>
<dd>
<pre>
$ perldoc -q "whether a scalar is a number"</pre>
</dd>
</li>
<dt><strong><a name="item_is_symbol"><em>is_symbol(SCALAR-REF)</em></a></strong>
<dd>
<p>Test for symbolic names. <em>is_symbol</em> can be used to test whether a scalar looks like a symbolic
name. Such strings need not to be quoted. Rlist defines symbolic names as a superset of C
identifier names:</p>
</dd>
<dd>
<pre>
[a-zA-Z_0-9] # C/C++ character set for identifiers
[a-zA-Z_0-9\-/\~:\.@] # Rlist character set for symbolic names</pre>
</dd>
<dd>
<pre>
[a-zA-Z_][a-zA-Z_0-9]* # match C/C++ identifier
[a-zA-Z_\-/\~:@][a-zA-Z_0-9\-/\~:\.@]* # match Rlist symbolic name</pre>
</dd>
<dd>
<p>For example, names such as <em>std::foo</em>, <em>msg.warnings</em>, <em>--verbose</em>, <em>calculation-info</em> need not
be quoted.</p>
</dd>
</li>
<dt><strong><a name="item_is_value"><em>is_value(SCALAR-REF)</em></a></strong>
<dd>
<p>Returns true when a scalar is an integer, a number, a symbolic name or some quoted string.</p>
</dd>
</li>
<dt><strong><a name="item_is_random_text"><em>is_random_text(SCALAR-REF)</em></a></strong>
<dd>
<p>The opposite of <em><a href="#item_is_value">is_value</a></em>. Such scalars will be turned into quoted strings by <em><a href="#item_compile">compile</a></em>
and <em><a href="#item_compile_fast">compile_fast</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_quote7"><em>quote7(TEXT)</em></a></strong>
<dt><strong><a name="item_escape7"><em>escape7(TEXT)</em></a></strong>
<dd>
<p>Converts TEXT into 7-bit-ASCII. All characters not in the set of the 95 printable ASCII characters
are escaped. The following ASCII codes will be converted to escaped octal numbers, i.e. 3 digits
prefixed by a slash:</p>
</dd>
<dd>
<pre>
0x00 to 0x1F
0x80 to 0xFF
" ' \</pre>
</dd>
<dd>
<p>The difference between the two functions is that <em>quote7</em> additionally places TEXT into
double-quotes. For example, <em>quote7(qq'``Früher Mittag\n''')</em> returns <code>"\"Fr\374her
Mittag\n\""</code>, while <em>escape7</em> returns <code>\"Fr\374her Mittag\n\"</code></p>
</dd>
</li>
<dt><strong><a name="item_maybe_quote7"><em>maybe_quote7(TEXT)</em></a></strong>
<dd>
<p>Return <em>quote7(TEXT)</em> if <em><a href="#item_is_random_text">is_random_text</a>(TEXT)</em>; otherwise (TEXT defines a symbolic name or
number) return TEXT.</p>
</dd>
</li>
<dt><strong><a name="item_maybe_unquote7"><em>maybe_unquote7(TEXT)</em></a></strong>
<dd>
<p>Return <em>unquote7(TEXT)</em> when TEXT is enclosed by double-quotes; otherwise returns TEXT.</p>
</dd>
</li>
<dt><strong><a name="item_unquote7"><em>unquote7(TEXT)</em></a></strong>
<dt><strong><a name="item_unescape7"><em>unescape7(TEXT)</em></a></strong>
<dd>
<p>Reverses what <em><a href="#item_quote7">quote7</a></em> and <em><a href="#item_escape7">escape7</a></em> did with TEXT.</p>
</dd>
</li>
<dt><strong><a name="item_unhere"><em>unhere(HERE-DOC-STRING[, COLUMNS, FIRSTTAB, DEFAULTTAB])</em></a></strong>
<dd>
<p>Combines recipes 1.11 and 1.12 from the Perl Cookbook. HERE-DOC-STRING shall be a
<a href="#numbers__strings_and_heredocuments">here-document</a>. The function checks whether each line
begins with a common prefix, and if so, strips that off. If no prefix it takes the amount of
leading whitespace found the first line and removes that much off each subsequent line.</p>
</dd>
<dd>
<p>Unless COLUMNS is defined returns the new here-doc-string. Otherwise, takes the string and
reformats it into a paragraph having no line more than COLUMNS characters long. FIRSTTAB will be
the indent for the first line, DEFAULTTAB the indent for every subsequent line. Unless passed,
FIRSTTAB and DEFAULTTAB default to the empty string <code>""</code>.</p>
</dd>
</li>
<dt><strong><a name="item_split_quoted"><em>split_quoted(INPUT[, DELIMITER])</em></a></strong>
<dt><strong><a name="item_parse_quoted"><em>parse_quoted(INPUT[, DELIMITER])</em></a></strong>
<dd>
<p>Divide the string INPUT into a list of strings. DELIMITER is a regular expression specifying where
to split (default: <code>'\s+'</code>). The functions won't split at DELIMITERs inside quotes, or which are
backslashed.</p>
</dd>
<dd>
<p><em>parse_quoted</em> works like <em>split_quoted</em> but additionally removes all quotes and backslashes
from the splitted fields. Both functions effectively simplify the interface of
<em>Text::ParseWords</em>. In an array context they return a list of substrings, otherwise the count of
substrings. An empty array is returned in case of unbalanced double-quotes, e.g.
<em>split_quoted(<code>'foo,"bar'</code>)</em>.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<pre>
sub split_and_list($) {
print ($i++, " '$_'\n") foreach split_quoted(shift)
}</pre>
</dd>
<dd>
<pre>
split_and_list(q("fee foo" bar))</pre>
</dd>
<dd>
<pre>
0 '"fee foo"'
1 'bar'</pre>
</dd>
<dd>
<pre>
split_and_list(q("fee foo"\ bar))</pre>
</dd>
<dd>
<pre>
0 '"fee foo"\ bar'</pre>
</dd>
<dd>
<p>The default DELIMITER <code>'\s+'</code> handles newlines. <em>split_quoted(<code>"foo\nbar\n"</code>)</em> returns
<em>('foo', 'bar', '')</em> and hence can be used to to split a large string of un<em>chomp</em>'d input
lines into words:</p>
</dd>
<dd>
<pre>
split_and_list("foo \r\n bar\n")</pre>
</dd>
<dd>
<pre>
0 'foo'
1 'bar'
2 ''</pre>
</dd>
<dd>
<p>The DELIMITER matches everywhere outside of quoted constructs, so in case of the default <code>'\s+'</code>
you may want to remove heading/trailing whitespace. Consider</p>
</dd>
<dd>
<pre>
split_and_list("\nfoo")
split_and_list("\tfoo")</pre>
</dd>
<dd>
<pre>
0 ''
1 'foo'</pre>
</dd>
<dd>
<p>and</p>
</dd>
<dd>
<pre>
split_and_list(" foo ")</pre>
</dd>
<dd>
<pre>
0 ''
1 'foo'
2 ''</pre>
</dd>
<dd>
<p><em>parse_quoted</em> additionally removes all quotes and backslashes from the splitted fields:</p>
</dd>
<dd>
<pre>
sub parse_and_list($) {
print ($i++, " '$_'\n") foreach parse_quoted(shift)
}</pre>
</dd>
<dd>
<pre>
parse_and_list(q("fee foo" bar))</pre>
</dd>
<dd>
<pre>
0 'fee foo'
1 'bar'</pre>
</dd>
<dd>
<pre>
parse_and_list(q("fee foo"\ bar))</pre>
</dd>
<dd>
<pre>
0 'fee foo bar'</pre>
</dd>
<dd>
<p><strong>MORE EXAMPLES</strong></p>
</dd>
<dd>
<p>String <code>'field\ one "field\ two"'</code>:</p>
</dd>
<dd>
<pre>
('field\ one', '"field\ two"') # split_quoted
('field one', 'field two') # parse_quoted</pre>
</dd>
<dd>
<p>String <code>'field\,one, field", two"'</code> with a DELIMITER of <code>'\s*,\s*'</code>:</p>
</dd>
<dd>
<pre>
('field\,one', 'field", two"') # split_quoted
('field,one', 'field, two') # parse_quoted</pre>
</dd>
<dd>
<p>Split a large string <em>$soup</em> (mnemonic: slurped from a file) into lines, at LF or CR+LF:</p>
</dd>
<dd>
<pre>
@lines = split_quoted($soup, '\r*\n');</pre>
</dd>
<dd>
<p>Then transform all <em>@lines</em> by correctly splitting each line into ``naked'' values:</p>
</dd>
<dd>
<pre>
@table = map { [ parse_quoted($_, '\s*,\s') ] } @lines</pre>
</dd>
<dd>
<p>Here is some more complete code to parse a <em>.csv</em>-file with quoted fields, escaped commas:</p>
</dd>
<dd>
<pre>
open my $fh, "foo.csv" or die $!;
local $/; # enable localized slurp mode
my $content = <$fh>; # slurp whole file at once
close $fh;
my @lines = split_quoted($content, '\r*\n');
die q(unbalanced " in input) unless @lines;
my @table = map { [ map { parse_quoted($_, '\s*,\s') } ] } @lines</pre>
</dd>
<dd>
<p>In core this is what <em><a href="#item_read_csv">read_csv</a></em> does. <em><a href="#item_deep_compare">deep_compare</a></em> allows you to test what
<em><a href="#item_split_quoted">split_quoted</a></em> and <em><a href="#item_parse_quoted">parse_quoted</a></em> return. For example, the following code shall never
die:</p>
</dd>
<dd>
<pre>
croak if deep_compare([split_quoted("fee fie foo")], ['fee', 'fie', 'foo']);
croak if deep_compare( parse_quoted('"fee fie foo"'), 1);</pre>
</dd>
</li>
<dt><strong><a name="item_equal"><em>equal(NUM1, NUM2[, PRECISION])</em></a></strong>
<dd>
<p><em><a href="#item_equal">equal</a></em> returns true if NUM1 and NUM2 are equal to PRECISION number of decimal places
(default: 6). For details see <em><a href="#item_round">round</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_round"><em>round(NUM1[, PRECISION])</em></a></strong>
<dd>
<p>Compare and round floating-point numbers NUM1 and NUM2 (as string- or number scalars).</p>
</dd>
<dd>
<p>When the <code>"precision"</code> compile option is defined, <em><a href="#item_round">round</a></em> is called during compilation on all
numbers.</p>
</dd>
<dd>
<p>Normally <em>round</em> will return a number in fixed-point notation. When the package-global
<em>$Data::Rlist::RoundScientific</em> is true, however, <em>round</em> formats the number in either normal or
exponential (scientific) notation, whichever is more appropriate for its magnitude. This differs
slightly from fixed-point notation in that insignificant zeroes to the right of the decimal point
are not included. Also, the decimal point is not included on whole numbers. For example,
<em><a href="#item_round">round</a>(42)</em> does not return 42.000000, and <em>round(0.12)</em> returns 0.12, not 0.120000.</p>
</dd>
<dd>
<p><strong>MACHINE ACCURACY</strong></p>
</dd>
<dd>
<p>One needs a function like <em>equal</em> to compare floats, because IEEE 754 single- and double precision
implementations are not absolute - in contrast to the numbers they actually represent. In all
machines non-integer numbers are only an approximation to the numeric truth. In other words,
they're not commutative. For example, given two floats <em>a</em> and <em>b</em>, the result of <em>a+b</em> might
be different than that of <em>b+a</em>. For another example, it is a mathematical truth that <em>a * b = b
* a</em>, but not necessarily in a computer.</p>
</dd>
<dd>
<p>Each machine has its own accuracy, called the <em>machine epsilon</em>, which is the difference between 1
and the smallest exactly representable number greater than one. Most of the time only floats can be
compared that have been carried out to a certain number of decimal places. In general this is the
case when two floats that result from a numeric operation are compared - but not two constants.
(Constants are accurate through to lexical conventions of the language. The Perl and C syntaxes for
numbers simply won't allow you to write down inaccurate numbers.)</p>
</dd>
<dd>
<p>See also recipes 2.2 and 2.3 in the Perl Cookbook.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<pre>
CALL RETURNS NUMBER
---- --------------
round('0.9957', 3) 0.996
round(42, 2) 42
round(0.12) 0.120000
round(0.99, 2) 0.99
round(0.991, 2) 0.99
round(0.99, 1) 1.0
round(1.096, 2) 1.10
round(+.99950678) 0.999510
round(-.00057260) -0.000573
round(-1.6804e-6) -0.000002</pre>
</dd>
</li>
<dt><strong><a name="item_deep_compare"><em>deep_compare(A, B[, PRECISION, TRACE_FLAG])</em></a></strong>
<dd>
<p>Compare and analyze two numbers, strings or references. Generates a list of messages describing
exactly all unequal data. Hence, for any Perl data <em>$a</em> and <em>$b</em> one can assert:</p>
</dd>
<dd>
<pre>
croak "$a differs from $b" if deep_compare($a, $b);</pre>
</dd>
<dd>
<p>When PRECISION is defined all numbers in A and B are <em><a href="#item_round">round</a></em>'d before actually comparing them.
When TRACE_FLAG is true traces progress.</p>
</dd>
<dd>
<p><strong>RESULT</strong></p>
</dd>
<dd>
<p>Returns an array of messages, each describing unequal data, or data that cannot be compared because
of type- or value-mismatching. The array is empty when deep comparison of A and B found no unequal
numbers or strings, and only indifferent types.</p>
</dd>
<dd>
<p><strong>EXAMPLES</strong></p>
</dd>
<dd>
<p>The result is line-oriented, and for each mismatch it returns a single message. For a simple
example,</p>
</dd>
<dd>
<pre>
Data::Rlist::deep_compare(undef, 1)</pre>
</dd>
<dd>
<p>yields</p>
</dd>
<dd>
<pre>
<<undef>> cmp <<1>> stop! 1st undefined, 2nd defined (1)</pre>
</dd>
</li>
<dt><strong><a name="item_fork_and_wait"><em>fork_and_wait(PROGRAM[, ARGS...])</em></a></strong>
<dd>
<p>Forks a process and waits for completion. The function will extract the exit-code, test whether
the process died and prints status messages on <em>STDERR</em>. <em>fork_and_wait</em> hence is a handy
wrapper around the built-in <em>system</em> and <em>exec</em> functions. Returns an array of three values:</p>
</dd>
<dd>
<pre>
($exit_code, $failed, $coredump)</pre>
</dd>
<dd>
<p><em>$exit_code</em> is -1 when the program failed to execute (e.g. it wasn't found or the current user
has insufficient rights). Otherwise <em>$exit_code</em> is between 0 and 255. When the program died on
receipt of a signal (like <em>SIGINT</em> or <em>SIGQUIT</em>) then <em>$signal</em> stores it. When <em>$coredump</em> is
true the program died and a <em>core</em>-file was written.</p>
</dd>
</li>
<dt><strong><a name="item_synthesize_pathname"><em>synthesize_pathname(TEXT...)</em></a></strong>
<dd>
<p>Concatenates and forms all TEXT strings into a symbolic name that can be used as a pathname.
<em>synthesize_pathname</em> is a useful function to concatenate strings and nearby converting all
characters that do not qualify as filename-characters, into <code>"_"</code> and <code>"-"</code>. The result cannot
only be used as file- or URL name, but also (coinstantaneously) as hash key, database name etc.</p>
</dd>
</li>
</dl>
<dl>
<dt><strong><a name="item__27precision_27__3d_3e_places">'precision' => PLACES</a></strong>
<dd>
<p>Make <em><a href="#item_compile">compile</a></em> round all numbers to PLACES decimal places, by calling <em><a href="#item_round">round</a></em> on each
scalar that <a href="#item_is_number">looks like a number</a>. By default PLACES is <em>undef</em>, which means floats
are not rounded.</p>
</dd>
</li>
<dt><strong><a name="item__27scientific_27__3d_3e_flag">'scientific' => FLAG</a></strong>
<dd>
<p>Causes <em><a href="#item_compile">compile</a></em> to masquerade <em>$Data::Rlist::RoundScientific</em>. See <em><a href="#item_round">round</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__27code_refs_27__3d_3e_token">'code_refs' => TOKEN</a></strong>
<dd>
<p>Defines how <em><a href="#item_compile">compile</a></em> shall treat <em>CODE</em> reference. Legal values for TOKEN are 0 (the
default), <code>"call"</code> and <code>"deparse"</code>.</p>
</dd>
<dd>
<p>- 0 compiles subroutine references into the string <code>"?CODE?"</code>.</p>
</dd>
<dd>
<p>- <code>"call"</code> calls the code, then compiles the return value.</p>
</dd>
<dd>
<p>- <code>"deparse"</code> serializes the code using <em>B::Deparse</em> (reproducing the Perl source).</p>
</dd>
</li>
<dt><strong><a name="item__27threads_27__3d_3e_count">'threads' => COUNT</a></strong>
<dd>
<p>If enabled <em><a href="#item_compile">compile</a></em> internally use multiple threads. Note that can speedup compilation only
on machines with at least COUNT CPUs.</p>
</dd>
</li>
<dt><strong><a name="item__27here_docs_27__3d_3e_flag">'here_docs' => FLAG</a></strong>
<dd>
<p>If enabled strings with at least two newlines in them are written as
<a href="#heredocuments">here-document</a>, when possible. To qualify as here-document a string has to have
at least two LFs (<code>"\n"</code>), one of which must terminate it.</p>
</dd>
</li>
<dt><strong><a name="item__27auto_quote_27__3d_3e_flag">'auto_quote' => FLAG</a></strong>
<dd>
<p>When true (default) do not quote strings that look like identifiers (see <em><a href="#item_is_symbol">is_symbol</a></em>). When
false quote <em>all</em> strings. Hash keys are not affected.</p>
</dd>
<dd>
<p><em><a href="#item_write_csv">write_csv</a></em> and <em><a href="#item_write_conf">write_conf</a></em> interpret this flag differently: false means not to quote at
all; true quotes only strings that don't look like numbers and that aren't yet quoted.</p>
</dd>
</li>
<dt><strong><a name="item__27outline_data_27__3d_3e_number">'outline_data' => NUMBER</a></strong>
<dd>
<p>When NUMBER is greater than 0 use <code>"eol_space"</code> (linefeed) to split data to many lines. It will
insert a linefeed after every NUMBERth array value.</p>
</dd>
</li>
<dt><strong><a name="item__27outline_hashes_27__3d_3e_flag">'outline_hashes' => FLAG</a></strong>
<dd>
<p>If enabled, and <code>"outline_data"</code> is also enabled, prints <em>{</em> and <em>}</em> on distinct lines when
compiling Perl hashes with at least one pair.</p>
</dd>
</li>
<dt><strong><a name="item__27separator_27__3d_3e_string">'separator' => STRING</a></strong>
<dd>
<p>The comma-separator string to be used by <em><a href="#item_write_csv">write_csv</a></em>. The default is <code>','</code>.</p>
</dd>
</li>
<dt><strong><a name="item__27delimiter_27__3d_3e_regex">'delimiter' => REGEX</a></strong>
<dd>
<p>Field-delimiter for <em><a href="#item_read_csv">read_csv</a></em>. There is no default value. To read configuration files, for
example, you may use <code>'\s*=\s*'</code> or <code>'\s+'</code>. To read CSV-files use e.g. <code>'\s*[,;]\s*'</code>.</p>
</dd>
</li>
</dl>
<p>The following options format the generated Rlist; normally you don't want to modify them:</p>
<dl>
<dt><strong><a name="item__27bol_tabs_27__3d_3e_count">'bol_tabs' => COUNT</a></strong>
<dd>
<p>Count of physical, horizontal TAB characters to use at the begin-of-line per indentation
level. Defaults to 1. Note that we don't use blanks, because they blow up the size of generated
text without measure.</p>
</dd>
</li>
<dt><strong><a name="item__27eol_space_27__3d_3e_string">'eol_space' => STRING</a></strong>
<dd>
<p>End-of-line string to use (the linefeed). For example, legal values are <code>""</code>, <code>" "</code>, <code>"\n"</code>,
<code>"\r\n"</code> etc. The default is <em>undef</em>, which means to use the current value of <em>$/</em>. Note that
this is a compile-option that only affects <em><a href="#item_compile">compile</a></em>. When parsing files the builtin
<em>readline</em> function is called, which uses <em>$/</em>.</p>
</dd>
</li>
<dt><strong><a name="item__27paren_space_27__3d_3e_string">'paren_space' => STRING</a></strong>
<dd>
<p>String to write after <em>(</em> and <em>{</em>, and before <em>}</em> and <em>)</em> when compiling arrays and hashes.</p>
</dd>
</li>
<dt><strong><a name="item__27comma_punct_27__3d_3e_string">'comma_punct' => STRING</a></strong>
<dt><strong><a name="item__27semicolon_punct_27__3d_3e_string">'semicolon_punct' => STRING</a></strong>
<dd>
<p>Comma and semicolon strings, which shall be at least <code>","</code> and <code>";"</code>. No matter what,
<em><a href="#item_compile">compile</a></em> will always print the <code>"eol_space"</code> string after the <code>"semicolon_punct"</code> string.</p>
</dd>
</li>
<dt><strong><a name="item__27assign_punct_27__3d_3e_string">'assign_punct' => STRING</a></strong>
<dd>
<p>String to make up key/value-pairs. Defaults to <code>" = "</code>.</p>
</dd>
</li>
</dl>
<p>
</p>
<h2><a name="predefined_options">Predefined Options</a></h2>
<p>The <a href="#compile_options">OPTIONS</a> parameter accepted by some package functions is either a hash-ref
or the name of a predefined set:</p>
<dl>
<dt><strong><a name="item__27default_27">'default'</a></strong>
<dd>
<p>Default if writing to a file.</p>
</dd>
</li>
<dt><strong><a name="item__27string_27">'string'</a></strong>
<dd>
<p>Compact, no newlines/here-docs. Renders a ``string of data''.</p>
</dd>
</li>
<dt><strong><a name="item__27outlined_27">'outlined'</a></strong>
<dd>
<p>Optimize the compiled Rlist for maximum readability.</p>
</dd>
</li>
<dt><strong><a name="item__27squeezed_27">'squeezed'</a></strong>
<dd>
<p>Very compact, no whitespace at all. For very large Rlists.</p>
</dd>
</li>
<dt><strong><a name="item__27perl_27">'perl'</a></strong>
<dd>
<p>Compile data in Perl syntax, using <em><a href="#item_compile_perl">compile_Perl</a></em>, not <em><a href="#item_compile">compile</a></em>. The output then
can be <em>eval</em>'d, but it cannot be <em><a href="#item_read">read</a></em> back.</p>
</dd>
</li>
<dt><strong><a name="item__27fast_27_or_undef">'fast' or <em>undef</em></a></strong>
<dd>
<p>Compile data as fast as possible, using <em><a href="#item_compile_fast">compile_fast</a></em>, not <em><a href="#item_compile">compile</a></em>.</p>
</dd>
</li>
</dl>
<p>All functions that define an <a href="#compile_options">OPTIONS</a> parameter do implicitly call
<em><a href="#item_complete_options">complete_options</a></em> to complete the argument from one of the predefined sets, and additionally
from <code>"default"</code>. Therefore you can always define nothing, or a ``lazy subset of options''. For
example,</p>
<pre>
my $obj = new Data::Rlist(-data => $thing);</pre>
<pre>
$obj->write('thing.rls', { scientific => 1, precision => 8 });</pre>
<p>
</p>
<h2><a name="exports">Exports</a></h2>
<p>Example:</p>
<pre>
use Data::Rlist qw/:floats :strings/;</pre>
<p>
</p>
<h3><a name="exporter_tags">Exporter Tags</a></h3>
<dl>
<dt><strong><a name="item__3afloats"><em>:floats</em></a></strong>
<dd>
<p>Imports <em><a href="#item_equal">equal</a></em>, <em><a href="#item_round">round</a></em> and <em><a href="#item_is_number">is_number</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__3astrings"><em>:strings</em></a></strong>
<dd>
<p>Imports <em><a href="#item_maybe_quote7">maybe_quote7</a></em>, <em><a href="#item_quote7">quote7</a></em>, <em><a href="#item_escape7">escape7</a></em>, <em><a href="#item_unquote7">unquote7</a></em>, <em><a href="#item_unescape7">unescape7</a></em>,
<em><a href="#item_unhere">unhere</a></em>, <em><a href="#item_is_random_text">is_random_text</a></em>, <em><a href="#item_is_number">is_number</a></em>, <em><a href="#item_is_symbol">is_symbol</a></em>, <em><a href="#item_split_quoted">split_quoted</a></em>, and
<em><a href="#item_parse_quoted">parse_quoted</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__3aoptions"><em>:options</em></a></strong>
<dd>
<p>Imports <em><a href="#item_predefined_options">predefined_options</a></em> and <em><a href="#item_complete_options">complete_options</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item__3aaux"><em>:aux</em></a></strong>
<dd>
<p>Imports <em><a href="#item_deep_compare">deep_compare</a></em>, <em><a href="#item_fork_and_wait">fork_and_wait</a></em> and <em><a href="#item_synthesize_pathname">synthesize_pathname</a></em>.</p>
</dd>
</li>
</dl>
<p>
</p>
<h3><a name="autoexported_functions">Auto-Exported Functions</a></h3>
<p>The following functions are implicitly imported into the callers symbol table. (But you may say
<em>require Data::Rlist</em> instead of <em>use Data::Rlist</em> to prohibit auto-import. See also
<em>perlmod</em>.)</p>
<dl>
<dt><strong><a name="item_readdata"><em>ReadData(INPUT[, FILTER, FILTER-ARGS])</em></a></strong>
<dt><strong><a name="item_readcsv"><em>ReadCSV(INPUT[, OPTIONS, FILTER, FILTER-ARGS])</em></a></strong>
<dt><strong><a name="item_readconf"><em>ReadConf(INPUT[, OPTIONS, FILTER, FILTER-ARGS])</em></a></strong>
<dd>
<p>These are aliases for <em>Data::Rlist::<a href="#item_read">read</a></em>, <em>Data::Rlist::<a href="#item_read_csv">read_csv</a></em> and
<em>Data::Rlist::<a href="#item_read_conf">read_conf</a></em>.</p>
</dd>
</li>
<dt><strong><a name="item_evaluatedata"><em>EvaluateData(INPUT[, FILTER, FILTER-ARGS])</em></a></strong>
<dd>
<p>Like <em><a href="#item_readdata">ReadData</a></em> but implicitly call <em>Data::Rlist::<a href="#item_evaluate_nanoscripts">evaluate_nanoscripts</a></em> in case parsing
was successful.</p>
</dd>
</li>
<dt><strong><a name="item_writedata"><em>WriteData(DATA[, OUTPUT, OPTIONS, HEADER])</em></a></strong>
<dt><strong><a name="item_writecsv"><em>WriteCSV(DATA[, OUTPUT, OPTIONS, COLUMNS, HEADER])</em></a></strong>
<dt><strong><a name="item_writeconf"><em>WriteConf(DATA[, OUTPUT, OPTIONS, HEADER])</em></a></strong>
<dd>
<p>These are aliases for <em>Data::Rlist::<a href="#item_write">write</a></em>, <em>Data::Rlist::<a href="#item_write_string">write_string</a></em>
<em>Data::Rlist::<a href="#item_write_csv">write_csv</a></em> and <em>Data::Rlist::<a href="#item_write_conf">write_conf</a></em>. OPTIONS default to <code>"default"</code>.</p>
</dd>
</li>
<dt><strong><a name="item_outlinedata"><em>OutlineData(DATA[, OPTIONS])</em></a></strong>
<dt><strong><a name="item_stringizedata"><em>StringizeData(DATA[, OPTIONS])</em></a></strong>
<dt><strong><a name="item_squeezedata"><em>SqueezeData(DATA[, OPTIONS])</em></a></strong>
<dd>
<p>These are aliases for <em>Data::Rlist::<a href="#item_write_string_value">write_string_value</a></em>. <em>OutlineData</em> applies the
predefined <a href="#predefined_options"><code>"outlined"</code></a> options, while <em>StringizeData</em> applies
<a href="#predefined_options"><code>"string"</code></a> and <em>SqueezeData</em>() <a href="#predefined_options"><code>"squeezed"</code></a>. When
specified, OPTIONS are merged into the. For example,</p>
</dd>
<dd>
<pre>
print "\n\$thing: ", OutlineData($thing, { precision => 12 });</pre>
</dd>
<dd>
<p><em><a href="#item_round">rounds</a></em> all numbers in <em>$thing</em> to 12 digits.</p>
</dd>
</li>
<dt><strong><a name="item_printdata"><em>PrintData(DATA[, OPTIONS])</em></a></strong>
<dd>
<p>An alias for</p>
</dd>
<dd>
<pre>
print OutlineData(DATA, OPTIONS);</pre>
</dd>
</li>
<dt><strong><a name="item_keelhauldata"><em>KeelhaulData(DATA[, OPTIONS])</em></a></strong>
<dt><strong><a name="item_comparedata"><em>CompareData(A, B[, PRECISION, TRACE_FLAG])</em></a></strong>
<dd>
<p>These are aliases for <em><a href="#item_keelhaul">keelhaul</a></em> and <em><a href="#item_deep_compare">deep_compare</a></em>. For example,</p>
</dd>
<dd>
<pre>
use Data::Rlist;
.
.
my($copy, $as_text) = KeelhaulData($thing);</pre>
</dd>
</li>
</dl>
<p>
</p>
<hr />
<h1><a name="examples">EXAMPLES</a></h1>
<p>String- and number values:</p>
<pre>
"Hello, World!"
foo # compiles to { 'foo' => undef }
3.1415 # compiles to { 3.1415 => undef }</pre>
<p>Array values:</p>
<pre>
(1, a, 4, "b u z") # list of numbers/strings</pre>
<pre>
((1, 2),
(3, 4)) # list of list (4x4 matrix)</pre>
<pre>
((1, a, 3, "foo bar"),
(7, c, 0, "")) # another list of lists</pre>
<p>Here-document strings:</p>
<pre>
$hello = ReadData(\<<HELLO)
( <<DEUTSCH, <<ENGLISH, <<FRANCAIS, <<CASTELLANO, <<KLINGON, <<BRAINF_CK )
Hallo Welt!
DEUTSCH
Hello World!
ENGLISH
Bonjour le monde!
FRANCAIS
Ola mundo!
CASTELLANO
~ nuqneH { ~ 'u' ~ nuqneH disp disp } name
nuqneH
KLINGON
++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++
..+++.>++.<<+++++++++++++++.>.+++.------.--------.>+.>.
BRAINF_CK
HELLO</pre>
<p>Compiles <em>$hello</em> as</p>
<pre>
[ "Hallo Welt!\n", "Hello World!\n", "Bonjour le monde!\n", "Ola mundo!\n",
"~ nuqneH { ~ 'u' ~ nuqneH disp disp } name\n",
"++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++\n..+++.>++.<<+++++++++++++++.>.+++.------.--------.>+.>.\n" ]</pre>
<p>Configuration object as hash:</p>
<pre>
{
contribution_quantile = 0.99;
default_only_mode = Y;
number_of_runs = 10000;
number_of_threads = 10;
# etc.
}</pre>
<p>Altogether:</p>
<pre>
Metaphysic-terms =
{
Numbers =
{
3.141592653589793 = "The ratio of a circle's circumference to its diameter.";
2.718281828459045 = <<___;
The mathematical constant "e" is the unique real number such that the value of
the derivative (slope of the tangent line) of f(x) = e^x at the point x = 0 is
exactly 1.
___
42 = "The Answer to Life, the Universe, and Everything.";
};</pre>
<pre>
Words =
{
ACME = <<Value;
A fancy-free Company [that] Makes Everything: Wile E. Coyote's supplier of equipment and gadgets.
Value
<<Key = <<Value;
foo bar foobar
Key
[JARGON] A widely used meta-syntactic variable; see foo for etymology. Probably
originally propagated through DECsystem manuals [...] in 1960s and early 1970s;
confirmed sightings go back to 1972. [...]
Value
};
};</pre>
<p>
</p>
<hr />
<h1><a name="notes">NOTES</a></h1>
<p>The <em>Random Lists</em> (Rlist) syntax is inspired by NeXTSTEP's <em>Property Lists</em>. But Rlist is
simpler, more readable and more portable. The Perl and C++ implementations are fast, stable and
free. Markus Felten, with whom I worked a few month in a project at Deutsche Bank, Frankfurt in
summer 1998, arrested my attention on Property lists. He had implemented a Perl variant of it
<p>The term ``Random'' underlines the fact that the language</p>
<ul>
<li>
<p>has four primitive/anonymuous types;</p>
</li>
<li>
<p>the basic building block is a list, which is combined at random with other lists.</p>
</li>
</ul>
<p>Hence the term <em>Random</em> does not mean <em>aimless</em> or <em>accidental</em>. <em>Random Lists</em> are
<em>arbitrary</em> lists.</p>
<p>
</p>
<hr />
<h1><a name="data__dumper"><em>Data::Dumper</em></a></h1>
<p>The main difference between <em>Data::Dumper</em> and <em>Data::Rlist</em> is that scalars will be properly
encoded as number or string. <em>Data::Dumper</em> writes numbers always as quoted strings, for example</p>
<pre>
$VAR1 = {
'configuration' => {
'verbose' => 'Y',
'importance_sampling_loss_quantile' => '0.04',
'distribution_loss_unit' => '100',
'default_only' => 'Y',
'num_threads' => '5',
.
.
}
};</pre>
<p>where <em>Data::Rlist</em> writes</p>
<pre>
{
configuration = {
verbose = Y;
importance_sampling_loss_quantile = 0.04;
distribution_loss_unit = 100;
default_only = Y;
num_threads = 5;
.
.
};
}</pre>
<p>As one can see <em>Data::Dumper</em> writes the data right in Perl syntax, which means the dumped text
can be simply <em>eval</em>'d, and the data can be restored very fast. Rlists are not quite Perl-syntax:
a dedicated parser is required. But therefore Rlist text is portable and can be read from other
programming languages such as <a href="#c__">C++</a>.</p>
<p>With <em>$Data::Dumper::Useqq</em> enabled it was observed that <em>Data::Dumper</em> renders output
significantly slower than <em><a href="#item_compile">compile</a></em>. This is actually suprising, since <em>Data::Rlist</em> tests
for each scalar whether it is numeric, and truely quotes/escapes strings. <em>Data::Dumper</em> quotes
all scalars (including numbers), and it does not escape strings. This may also result in some odd
behaviors. For example,</p>
<pre>
use Data::Dumper;
print Dumper "foo\n";</pre>
<p>yields</p>
<pre>
$VAR1 = 'foo
';</pre>
<p>while</p>
<pre>
use Data::Rlist;
PrintData "foo\n"</pre>
<p>yields</p>
<pre>
{ "foo\n"; }</pre>
<p>Finally, <em>Data::Rlist</em> generates smaller files. With the default <em>$Data::Dumper::Indent</em> of 2
<em>Data::Dumper</em>'s output is 4-5 times that of <em>Data::Rlist</em>'s. This is because <em>Data::Dumper</em>
recklessly uses blanks, instead of horizontal tabulators, which blows up file sizes without
measure.</p>
<p>
</p>
<h2><a name="rlist_vs__perl_syntax">Rlist vs. Perl Syntax</a></h2>
<p>Rlists are not Perl syntax:</p>
<pre>
RLIST PERL
----- ----
5; { 5 => undef }
"5"; { "5" => undef }
5=1; { 5 => 1 }
{5=1;} { 5 => 1 }
(5) [ 5 ]
{} { }
; { }
() [ ]</pre>
<p>
</p>
<h2><a name="debugging_data">Debugging Data</a></h2>
<p>To reduce recursive data structures (into true hierachies) set <em>$Data::Rlist::MaxDepth</em> to an
integer above 0. It then defines the depth under which <em><a href="#item_compile">compile</a></em> shall not venture deeper.
The compilation of Perl data (into Rlist text) then continues, but on <em>STDERR</em> a message like the
following is printed:</p>
<pre>
ERROR: compile2() broken in deep ARRAY(0x101aaeec) (depth = 101, max-depth = 100)</pre>
<p>This message will also be repeated as comment when the compiled Rlist is written to a file.
Furthermore <em>$Data::Rlist::Broken</em> is incremented by one. While the compilation continues,
effectively any attempt to venture deeper as suggested by <em>$Data::Rlist::MaxDepth</em> will be
blocked.</p>
<p>See <em><a href="#item_broken">broken</a></em>.</p>
<p>
</p>
<h2><a name="speeding_up_compilation__explicit_quoting_">Speeding up Compilation (Explicit Quoting)</a></h2>
<p>Much work has been spent to optimize <em>Data::Rlist</em> for speed. Still it is implemented in pure
Perl (no XS). A rough estimation for Perl 5.8 is ``each MB takes one second per GHz''. For example,
when the resulting Rlist file has a size of 13 MB, compiling it from a Perl script on a 3-GHz-PC
requires about 5-7 seconds. Compiling the same data under Solaris, on a sparcv9 processor
operating at 750 MHz, takes about 18-22 seconds.</p>
<p>The process of compiling can be speed up by calling <em><a href="#item_quote7">quote7</a></em> explicitly on scalars. That is,
before calling <em><a href="#item_write">write</a></em> or <em><a href="#item_write_string">write_string</a></em>. Big data sets may compile faster when for
scalars, that certainly not qualify as symbolic name, <em><a href="#item_quote7">quote7</a></em> is called in advance:</p>
<pre>
use Data::Rlist qw/:strings/;</pre>
<pre>
$data{quote7($key)} = $value;
.
.
Data::Rlist::write("data.rlist", \%data);</pre>
<p>instead of</p>
<pre>
$data{$key} = $value;
.
.
Data::Rlist::write("data.rlist", \%data);</pre>
<p>It depends on the case whether the first variant is faster: <em><a href="#item_compile">compile</a></em> and <em><a href="#item_compile_fast">compile_fast</a></em>
both have to call <em><a href="#item_is_random_text">is_random_text</a></em> on each scalar. When the scalar is already quoted, i.e.,
its first character is <code>"</code>, this test ought to run faster.</p>
<p>Internally <em><a href="#item_is_random_text">is_random_text</a></em> applies the precompiled regex <em>$Data::Rlist::REValue</em>. Note that
the expression <em>($s!~$Data::Rlist::REValue)</em> can be up to 20% faster than the equivalent
<em>is_random_text($s)</em>.</p>
<p>
</p>
<h2><a name="quoting_strings_that_look_like_numbers">Quoting strings that look like numbers</a></h2>
<p>Normally you don't have to care about strings, since un/quoting happens as required when
reading/compiling Rlist or CSV text. A common problem, however, occurs when some string uses the
same lexicography than numbers do.</p>
<p>Perl defines the string as the basic building block for all program data, then lets the program
decide <em>what strings mean</em>. Analogical, in a printed book the reader has to decipher the glyphs
and decide what evidence they hide. Printed text uses well-defined glyphs and typographic
conventions, and finally the competence of the reader, to recognize numbers. But computers need to
know the exact number type and format. Integer? Float? Hexadecimal? Scientific? Klingon? The
Perl Cookbook recommends the use of a regular expression to distinguish number from string scalars
(recipe 2.1).</p>
<p>In Rlist, string scalars that look like numbers need to be quoted explicitly. Otherwise, for
example, the string scalar <code>"-3.14"</code> appears as <em>-3.14</em> in the output, <code>"007324"</code> is compiled
into 7324 etc. Such text is lost and read back as a number. Of course, in most cases this is just
what you want. For hash keys, however, it might be a problem. One solution is to prefix the string
with <code>"_"</code>:</p>
<pre>
my $s = '-9'; $s = "_$s";</pre>
<p>Such strings do not qualify as a number anymore. In the C++ implementation it will then become
some <em>std::string</em>, not a <em>double</em>. But the leading <code>"_"</code> has to be removed by the reading
program. Perhaps a better solution is to explicitly call <em><a href="#item_quote7">quote7</a></em>:</p>
<pre>
use Data::Rlist qw/:strings/;</pre>
<pre>
$k = -9;
$k = quote7($k); # returns qq'"-9"'</pre>
<pre>
$k = 3.14_15_92;
$k = quote7($k); # returns qq'"3.141592"'</pre>
<p>Again, the need to quote strings that look like numbers is a problem evident only in the Perl
implementation of Rlist, since Perl is a language with weak types. With the C++ implementation of
Rlist there's no need to quote strings that look like numbers.</p>
<p>See also <em><a href="#item_write">write</a></em>, <em><a href="#item_is_number">is_number</a></em>, <em><a href="#item_is_symbol">is_symbol</a></em>, <em><a href="#item_is_random_text">is_random_text</a></em> and
<p>
</p>
<h2><a name="installing_rlist_pm_locally">Installing <em>Rlist.pm</em> locally</a></h2>
<p>Installing CPAN packages usually requires administrator privileges. Another way is to copy the
<em>Rlist.pm</em> file into a directory of your choice. Instead of <em>use Data::Rlist;</em>, however, you
then use the following code. It will find <em>Rlist.pm</em> also in <em>.</em> and <em>~/bin</em>, and it calls the
<em>Exporter</em> explicitly:</p>
<pre>
BEGIN {
$0 =~ /[^\/]+$/;
push @INC, $`||'.', "$ENV{HOME}/bin";
require Rlist;
Data::Rlist->import();
Data::Rlist->import(qw/:floats :strings/);
}</pre>
<p>
</p>
<h2><a name="an_rlistmode_for_emacs">An Rlist-Mode for Emacs</a></h2>
<pre>
(define-generic-mode 'rlist-generic-mode
(list "//" ?#)
nil
'(;; Punctuators
("\\([(){},;?=]\\)" 1 'cperl-array-face)
;; Numbers
("\\([-+]?[0-9]+\\(\\.[0-9]+\\)?[dDlL]?\\)" 1 'font-lock-constant-face)
;; Identifier names
("\\([-~A-Za-z_][-~A-Za-z0-9_]+\\)" 1 'font-lock-variable-name-face))
(list "\\.[rR][lL][iI]?[sS]$")
;; Extra functions to setup mode.
(list 'generic-bracket-support
'(lambda()
(require 'cperl-mode)
;;(hl-line-mode t) ; highlight cursor-line
(local-set-key [?\t] (lambda()(interactive)(cperl-indent-command)))
(local-set-key [?\M-q] 'fill-paragraph)
(set-fill-column 100)))
"Generic mode for Random Lists (Rlist) files.")</pre>
<p>
</p>
<h2><a name="implementation_details">Implementation Details</a></h2>
<p>
</p>
<h3><a name="perl">Perl</a></h3>
<p>
</p>
<h4><a name="package_dependencies">Package Dependencies</a></h4>
<p><em>Data::Rlist</em> depends only on few other packages:</p>
<pre>
Exporter
Carp
strict
integer
Sys::Hostname
Scalar::Util # deep_compare() only
Text::Wrap # unhere() only
Text::ParseWords # split_quoted(), parse_quoted() only</pre>
<p><em>Data::Rlist</em> is free of <em>$&</em>, <em>$`</em> or <em>$'</em>. Reason: once Perl sees that you need one of these
meta-variables anywhere in the program, it has to provide them for every pattern match. This may
substantially slow your program (see also <em>perlre</em>).</p>
<p>
</p>
<h4><a name="a_short_story_of_typeglobs">A Short Story of Typeglobs</a></h4>
<p>This is supplement information for <em><a href="#item_compile">compile</a></em>, the function internally called by <em><a href="#item_write">write</a></em>
and <em><a href="#item_write_string">write_string</a></em>. We will discuss why <em><a href="#item_compile">compile</a></em>, <em><a href="#item_compile_fast">compile_fast</a></em> and
<em><a href="#item_compile_perl">compile_Perl</a></em> transliterate typeglobs and typeglob-refs into <code>"?GLOB?"</code>. This is an
attempted explanation.</p>
<p><strong>TYPEGLOBS ARE A PERL IDIOSYNCRACY</strong></p>
<p>Perl uses a symbol table per package to map symbolic names like <em>x</em> to Perl values. Typeglob (aka
glob) objects are complete symbol table entries, as hash values. The symbol table hash (<em>stash</em>)
is an ordinary hash, named like the package with two colons appended. In the package stash the
symbol name is mapped to a memory address which holds the actual data of your program. In Perl we
do not have real global values, only package globals. Any Perl code is always running in one
package or another.</p>
<p>The main symbol table's name is <em>%main::</em>, or <em>%::</em>. In the C implementation of the Perl
interpreter, the main symbol is simply a global variable, called the <em>defstash</em> (default stash).
The symbol <em>Data::</em> in stash <em>%::</em> addresses the stash of package <em>Data</em>, and the symbol
<em>Rlist::</em> in the stash <em>%::Data::</em> addresses the stash of package <em>Data::Rlist</em>.</p>
<p>Typeglobs are an idiosyncracy of Perl: different types need only one stash entry, so that one
symbol can name all types of Perl data (scalars, arrays, hashes) and nondata (functions, formats,
I/O handles). The symbol <em>x</em> is mapped to the typeglob <em>*x</em>. In the typeglob coexist the scalar
<em>$x</em>, the list <em>@x</em>, the hash <em>%x</em>, the code <em>&x</em> and the I/O-handle or format specifieer <em>x</em>.</p>
<p>Most of the time only one glob slot is used. Do typeglobs waste space then? Probably not.
(Although some authors believe that.) Other script languages like (e.g.) Python is not forcing
decoration characters -- the interpreter already knows the type. In terms of C, symbol table
entries are then struct/union-combinations with a type field, a <em>double</em> field, a <em>char*</em> field
and so forth. Perl symbols follow a contrary design: globs are really pointer sets to low-level
structs that hold numbers, strings etc. Naturally pointers to non-existing values are NULL, and so
no type field is required. Perl interpreters can now implement fine-grained smart-pointers for
reference-counting and copy-on-write, and must not necessarily handle abstract unions. In theory,
the garbage-collector should have ``increased recycling opportunities.'' We do know, for example,
that <em>perl</em> is very greedy with RAM: it almost never returns any memory to the operating system.</p>
<p>Modifying <em>$x</em> in a Perl program won't change <em>%x</em>, because the typeglob <em>*x</em> is interposed
between the stash and the program's actual values for <em>$x</em>, <em>@x</em> etc. The sigil <em>*</em> serves as
wildcard for the other sigils <em>%</em>, <em>@</em>, <em>$</em> and <em>&</em>. (Hint: a <em>sigil</em> is a symbol ``created for
a specific magical purpose''; the name derives from the latin <em>sigilum</em> = seal.)</p>
<p>Typeglobs cannot be dissolved by <em><a href="#item_compile">compile</a></em>, because when (e.g.) <em>$x</em> and <em>%x</em> are in use,
the glob <em>*x</em> does not return some useful value like</p>
<pre>
(SCALAR => \$x, HASH => \@x)</pre>
<p>Typeglobs are also not interpolated in strings. <em>perl</em> always plays the ball back. A
typeglob-value is simply a string:</p>
<pre>
$ perl -e '$x=1; @x=(1); print *x'
*main::x</pre>
<pre>
$ perl -e 'print "*x is not interpolated"'
*x is not interpolated</pre>
<pre>
$ perl -e '$x = "this"; print "although ".*x." could be a string"'
although *main::x could be a string</pre>
<p>As one can see, even when only <em>$x</em> is defined the <em>*x</em> does not return its value. Typeglobs
(stash entries) are arranged by <em>perl</em> on the fly, even with the <em>use strict</em> pragma in effect:</p>
<pre>
$ perl -e 'package nirvana; use strict; print *x'
*nirvana::x</pre>
<p>Each typeglob is a full path into the <em>perl</em> stashes, down from the <em>defstash</em>:</p>
<pre>
$ perl -e 'print "*x is \"*main::x\"" if *x eq "*main::x"'
*x is "*main::x"</pre>
<pre>
$ perl -e 'package nirvana; sub f { local *g=shift; print *g."=$g" }; package main; $x=42; nirvana::f(*x)'
*main::x=42</pre>
<p><strong>GLOB-REFS</strong></p>
<p>In the C implementation of Perl, typeglobs have the struct-type <em>GV</em> for ``Glob value''. Each <em>GV</em>
is merely a set of pointers to sub-objects for scalars, arrays, hashes etc. In Perl the special
syntax <em>*x{ARRAY}</em> accesses the array-sub-object, and is another way to say <em>\@x</em>. But when
applied to a typeglob as <em>\*foo</em> it returns a typeglob-ref, or globref. So the Perl backslash
operator <code>\</code> works like the address-of operator <code>&</code> in C.</p>
<pre>
$ perl -e 'print *::'
*main::main:: # ???</pre>
<pre>
$ perl -e '$x = 42; print $::{x}'
*main::x # typeglob-value 'x' in the stash</pre>
<pre>
$ perl -e 'print \*::'
GLOB(0x10010f08) # some globref</pre>
<p>Little do we know what happens inside <em>perl</em>, when we assign REFs to typeglobs:</p>
<pre>
$ perl -e '$x = 42; *x = \$x; print $x'
42
$ perl -e '$y = 42; *x = \$y; print $x'
42</pre>
<p>In Perl4 you had to pass typeglob-refs to call functions by references (the backslash-operator was
not yet ``invented''). Since Perl5 saw the light of day, typeglob-refs can be considered as
artefacts. Note, however, that these veterans are still faster than true references, because true
references are themselves stored in a typeglob (as REF type) and so need to be dereferenced.
Globrefs can be used directly (as <em>GV*</em>'s) by <em>perl</em>. For example,</p>
<pre>
void f1 { my $bar = shift; ++$$bar }
void f2 { local *bar = shift; ++$bar }</pre>
<pre>
f1(\$x); # increments $x
f1(*x); # dto., but faster</pre>
<p><strong>GLOB-ALIASES</strong></p>
<p>Typeglob-aliases offer another interesting application for typeglobs. For example, <em>*bar=*x</em>
aliases the symbol <em>bar</em> in the current stash, so that <em>x</em> and <em>bar</em> point to the same typeglob.
This means that when you declare <em>sub x {}</em> after casting the alias, <em>bar</em> is <em>x</em>.</p>
<p>This smells like a free lunch. The penalty, however, is that the <em>bar</em> symbol cannot be easily
removed from the stash. One way is to say <em>local *bar</em>, wich temporarily assigns a new typeglob
to <em>bar</em> with all pointers zeroized:</p>
<pre>
package nirvana;</pre>
<pre>
sub f { print $bar; }
sub g { local *bar; $bar = 42; f(); }</pre>
<pre>
package main;</pre>
<pre>
nirvana::g();</pre>
<p>Running this code as Perl script prints the number assigned in <em>g</em>. <em>f</em> acts as a closure. The
<em>local</em>-statement will put the <em>bar</em> symbol temporarily into the package stash <em>%::nirvana</em>,
i.e., the same stash in which <em>f</em> and <em>g</em> exist. It will remove <em>bar</em> when <em>g</em> returns.</p>
<p><strong>*foo{THINGS}s</strong></p>
<p>The <em>*x{NAME}</em> expression family is fondly called ``the <em>*foo{THING}</em> syntax'':</p>
<pre>
$scalarref = *x{SCALAR};
$arrayref = *ARGV{ARRAY};
$hashref = *ENV{HASH};
$coderef = *handlers{CODE};</pre>
<pre>
$ioref = *STDIN{IO};
$ioref = *STDIN{FILEHANDLE}; # same as *STDIN{IO}</pre>
<pre>
$globref = *x{GLOB};
$globref = \*x; # same as *x{GLOB}
$undef = *x{THIS_NAME_IS_NOT_SUPPORTED} # yields undef</pre>
<pre>
die unless defined *x{SCALAR}; # ok -> will not die
die unless defined *x{GLOB}; # ok
die unless defined *x{HASH}; # error -> will die</pre>
<p>When THINGs are accessed this way few rules apply. Firstofall, <em>*foo{THING}s</em> are not hashes. The
syntax is a stopgap:</p>
<pre>
$ perl -e 'print \*x, *x{GLOB}, \*x{GLOB}'
GLOB(0x100110b8)GLOB(0x100110b8)REF(0x1002e944)</pre>
<pre>
$ perl -e '$x=1; exists *x{GLOB}'
exists argument is not a HASH or ARRAY element at -e line 1.</pre>
<p>Some <em>*foo{THING}</em> is <em>undef</em> if the requested THING hasn't been used yet. Only <em>*foo{SCALAR}</em>
returns an anonymous scalar-reference:</p>
<pre>
$ perl -e 'print "nope" unless defined *foo{HASH}'
nope
$ perl -e 'print *foo{SCALAR}'
SCALAR(0x1002e94c)</pre>
<p>In Perl5 it is still not possible to get a reference to an I/O-handle (file-, directory- or socket
handle) using the backslash operator. When a function requires an I/O-handle you must therefore
pass a globref. More precisely, it is possible to pass an <em>IO::Handle</em>-reference, a typeglob or a
typeglob-ref as the filehandle. This is obscure bot only for new Perl programmers.</p>
<pre>
sub logprint($@) {
my $fh = shift;
print $fh map { "$_\n" } @_;
}</pre>
<pre>
logprint(*STDOUT{IO}, 'foo'); # pass IO-handle -> IO::Handle=IO(0x10011b44)
logprint(*STDOUT, 'bar'); # ok, pass typeglob-value -> '*main::STDOUT'
logprint(\*STDOUT, 'bar'); # ok, pass typeglob-ref -> 'GLOB(0x10011b2c)'
logprint(\*STDOUT{IO}, 'nope'); # ERROR -> won't accept 'REF(0x10010fe0)'</pre>
<p>It is very amusing that Perl, although refactoring UNIX in form of a language, does not make clear
what a file- or socket-handle is. The global symbol STDOUT is actually an <em>IO::Handle</em> object,
which <em>perl</em> had silently instantiated. To functions like <em>print</em>, however, you may pass an
<em>IO::Handle</em>, globname or globref.</p>
<p><strong>VIOLATING STASHES</strong></p>
<p>As we saw we can access the Perl guts without using a scalpel. Suprisingly, it is also possible to
touch the stashes themselves:</p>
<pre>
$ perl -e '$x = 42; *x = $x; print *x'
*main::42</pre>
<pre>
$ perl -e '$x = 42; *x = $x; print *42'
*main::42</pre>
<p>By assigning the scalar value <em>$x</em> to <em>*x</em> we have demolished the stash (at least, logically):
neither <em>$42</em> nor <em>$main::42</em> are accessible. Symbols like <em>42</em> are invalid, because 42 is a
numeric literal, not a string literal.</p>
<pre>
$ perl -e '$x = 42; *x = $x; print $main::42'</pre>
<p>Nevertheless it is easy to confuse <em>perl</em> this way:</p>
<pre>
$ perl -e 'print *main::42'
*main::42</pre>
<pre>
$ perl -e 'print 1*9'
9</pre>
<pre>
$ perl -e 'print *9'
*main::9</pre>
<pre>
$ perl -e 'print *42{GLOB}'
GLOB(0x100110b8)</pre>
<pre>
$ perl -e '*x = 42; print $::{42}, *x'
*main::42*main::42</pre>
<pre>
$ perl -v
This is perl, v5.8.8 built for cygwin-thread-multi-64int
(with 8 registered patches, see perl -V for more detail)</pre>
<p>Of course these behaviors are not reliable, and may disappear in future versions of <em>perl</em>. In
German you say ``Schmutzeffekt'' (dirt effect) for certain mechanical effects that occur
non-intendedly, because machines and electrical circuits are not perfect, and so is software.
However, ``Schmutzeffekts'' are neither bugs nor features; these are phenomenons.</p>
<p><strong>LEXICAL VARIABLES</strong></p>
<p>Lexical variables (<em>my</em> variables) are not stored in stashes, and do not require typeglobs. These
variables are stored in a special array, the <em>scratchpad</em>, assigned to each block, subroutine, and
thread. These are really private variables, and they cannot be <em>local</em>ized. Each lexical variable
occupies a slot in the scratchpad; hence is addressed by an integer index, not a symbol. <em>my</em>
variables are like <em>auto</em> variables in C. They're also faster than <em>local</em>s, because they can be
allocated at compile time, not runtime. Therefore you cannot declare <em>*x</em> lexically:</p>
<pre>
$ perl -e 'my(*x)'
Can't declare ref-to-glob cast in "my" at -e line 1, near ");"</pre>
<p>Seel also the Perl man-pages <em>perlguts</em>, <em>perlref</em>, <em>perldsc</em> and <em>perllol</em>.</p>
<p>
</p>
<h3><a name="c__">C++</a></h3>
<p>In C++ we use a <em>flex</em>/<em>bison</em> scanner/parser combination to read Rlist language productions.
The C++ parser generates an <em>Abstract Syntax Tree</em> (AST) of <em>double</em>, <em>std::string</em>,
<em>std::vector</em> and <em>std::map</em> values. Since each value is put into the AST, as separate object,
we use a free store management that allows the allocation of huge amounts of tiny objects.</p>
<p>We also use reference-counted smart-pointers, which allocate themselves on our fast free store. So
RAM will not be fragmented, and the allocation of RAM is significantly faster than with the default
process heap. Like with Perl, Rlist files can have hundreds of megabytes of data (!), and are
processable in constant time, with constant memory requirements. For example, a 300 MB Rlist-file
can be read from a C++ process which will not peak over 400-500 MB of process RAM.</p>
<p>
</p>
<hr />
<h1><a name="bugs">BUGS</a></h1>
<p>There are no known bugs, this package is stable. Deficiencies and TODOs:</p>
<ul>
<li>
<p>The <code>"deparse"</code> functionality for the <code>"code_refs"</code> <a href="#compile_options">compile option</a> has not
yet been implemented.</p>
</li>
<li>
<p>The <code>"threads"</code> <a href="#compile_options">compile option</a> has not yet been implemented.</p>
</li>
<li>
<p>IEEE 754 notations of Infinite and NaN not yet implemented.</p>
</li>
<li>
<p><em><a href="#item_compile_perl">compile_Perl</a></em> is experimental.</p>
</li>
</ul>
<p>
</p>
<hr />
<h1><a name="copyright_license">COPYRIGHT/LICENSE</a></h1>
<p>Copyright 1998-2008 Andreas Spindler</p>
<p>This library is free software; you can redistribute it and/or modify it under the same terms as
Perl itself, either Perl version 5.8.8 or, at your option, any later version of Perl 5 you may have
available.</p>
<p>Contact the author for the C++ library at <em><a href="mailto:rlist@visualco.de">rlist@visualco.de</a></em>.</p>
<p>Thank you for your attention.</p>
</body>
</html>