pico8parse

pico8parse uses a set of "features" to describe Lua versions. Here is the list of what each feature means and for the parsing of which version of Lua it is used.

the strictP8FileFormat feature can be disable with the option ignoreStrictP8FileFormat

Unsorted Notes

.. and quite surely outdated.

About File Format and Sections

The PICO-8 file format (.p8, text) defines a set of rules that must be followed for a file to be correctly loaded:

• the first line of the file must contain the mention "pico-8 cartridge" (16 characters, case sensitive) any character may be present before and after (no new-line sequence before, obviously)

• the next line is entirely ignored (may it contain one of the sequences specified after, it is not parsed and is discarded)

• there exists (as of 0.2.2) 7 sections in a p8 file each identified by a set sequence of character (similar to Python's dunders); a generic __meta:*__ is introduced in 0.2.4c

• a section is entered as soon as a line containing on of these sequences is passed; the following lines are part of said section until any next one of these or EOF

• these sequences are only valid section opening if they are present at the very beginning of the line; any amount of any character may follow

• outside of section (typically right after the "pico-8 cartridge" header, before any sequence), lines are simply discarded

• as a section is closed by a new one, multiple section under the same sequence can be present within the file; section for a given sequence are concatenated in order of appearance

• only within a lua section may PICO-8 Lua be parsed, starting on the very next line

• a lua section may be closed at any point (eg. in the middle of an assignment) and resume in a next lua section, this is still considered valid

• usually, the first line is as bellow, the second line present a version number (noted VER) and each section is present once in the order of the list of sequences hereafter

pico-8 cartridge // http://www.pico-8.com
version VER

About Other PICO-8 Specific Features

The added syntaxes for singleLine[..] are pretty broken, see this test for some kind of overview.

With singleLineIf, the following becomes valid (note the 'do'):

   if ::= 'if' '(' exp ')' 'do' block {elif} ['else' block] 'end'

This snippet has an error (missing space after "name()" -- fixed above 0.2.4):

function name()if (exp) block
    block
end

This snippet prints "no":

if (1) do local a = "yes"
    print(a or "no")
end

And so does this on:

if (nil) else do local a = "yes"
    print(a or "no")
end

So the actual added syntax is closer to:

   if ::= '\s' 'if' '(' exp ')'
         [ block [ 'else' [ 'do' block_else_do '\n' ]
                   block_else_then
                 ] '\n'
         | [ 'do' block_if_do '\n'
           | 'then'
           ] block_if_then {elif} ['else' block] 'end'
         ]

.. where block_if_do has for parent block_if_then and block_else_do has for parent block_else_then...

singleLineIf and singleLineWhile cannot be empty; if (1) error, if (1) else no error, if (1) end do no error (closing with a proper 'end') while (1) error