Where we start to look at the code for Ordeal::Model::Parser.

For stability of these blog posts, we will take a look at the code as in release 0.003.

Let’s start with PARSE:

 1 sub PARSE ($text) {
 2    state $expression = _expression();
 3    my $ast = $expression->(\$text);
 4    my $pos = pos $text;
 5    my ($blanks, $rest) = substr($text, $pos) =~ m{\A (\s*) (.*) }mxs;
 6    if (length $rest) {
 7       $pos += length($blanks // '');
 8       my $prest = $rest;
 9       $prest = length($rest) > SHOW_CHARS
10          ? (substr($rest, 0, SHOW_CHARS - length ELLIPSIS) . ELLIPSIS)
11          : $rest;
12       ouch 400, "unknown sequence starting at $pos '$prest'", $rest;
13    }
14    return $ast;
15 }

It is actually just a wrapper around _expression(), or better around the function that is provided by _expression(). We just take care to verify that the input string parses correctly and does not contain characters we cannot include in the parsing.

To do this, we first run the parsing itself (line 3), then take a look at where we ended up (line 4) and check that there is nothing left, apart some blank spaces (lines 5 and test in line 6).

The function returns whatever was returned from the call in line 3; as the name says, we’re expecting this to be an Abstract Syntax Tree (or AST), i.e. a data structure modeled after the specific expression that was parsed. This is instrumental to later evaluate the expression, because it will mean that we will traverse the AST to get a value from it.

So… the really interesting stuff seems to lie inside _expression(), which provides us with the real entry point. Stay tuned!

The whole series

Want to look at the other parts of this series? Here’s a list of them: