/ 
PDF-Make-0.12
/ PDF::Make::Markup::Parse

SYNOPSIS

use PDF::Make::Markup::Parse;

my $root = PDF::Make::Markup::Parse->parse(<<'MARKUP');
<doc size="A4" margin="36">
  <h1>Invoice 1042</h1>
  <text>Amount due: <b>1,240.00</b> by 30 September.</text>
</doc>
MARKUP

DESCRIPTION

The markup is a closed tag set, not HTML and not XML. Anything outside the grammar is an error with a position rather than a best guess, because a document that renders slightly wrong is worse than one that refuses to render: the wrong one reaches a customer with a number on it.

Nodes

An element:

{ kind => 'elem', tag => 'row', attrs => { weight => '2' },
  children => [ ... ], line => 4, col => 3 }

Text:

{ kind => 'text', text => 'Amount due: ', line => 5, col => 12 }

What the grammar allows

  • The root element must be <doc>.

  • Tags come from the fixed set. tags() returns it, with the void, container and inline flags the parser itself uses.

  • Attribute values are always quoted, with " or '. A repeated attribute is an error rather than a silent winner.

  • Entities are &amp; &lt; &gt; &quot; &apos; and numeric character references &#NNN; and &#xHH;. There is no wider entity table: &nbsp; and friends are a slope that ends in an HTML specification.

  • <!-- comments --> are skipped.

  • Whitespace between the children of a container element is indentation and is dropped. Inside a text-bearing element it is content and survives, so Total <b>due</b> keeps its space.

  • Nesting deeper than 64 levels is an error. Templates arrive from customers, and unbounded depth is a stack overflow waiting for someone else to find.

SEE ALSO

PDF::Make::Markup::Build, PDF::Make::Builder