bug: Missing grammar rules

[Evrey]

Did you check existing issues?

• [X] I have read all the tree-sitter docs if it relates to using the parser
• [X] I have searched the existing issues of tree-sitter-PARSER_NAME

Tree-Sitter CLI Version, if relevant (output of tree-sitter --version)

No response

Describe the bug

Heyo, i was investigating whether this parser would be capable of processing the scripts i have written. Here’s a list of grammar rules that would have to be implemented for that to happen. I’m aware this parser is described as »rudimentary«, so view this as a kind of To-Do list towards eventual feature completeness.

• Title page:
  • [ ] Built-in properties like Title:.
  • [ ] Single- and multi-line property grammar. Basically inline or indented multi-line action text, styling and all.
  • [ ] Custom or future properties Prop name:, which are to be ignored by document processors if unknown.
• Scene headings:
  • [ ] INT/EXT.
  • [ ] I/E.
  • [ ] .A CUSTOM. Heading
  • [ ] .Forced heading
  • [ ] Lowercase or mixed-case headings. They must be rendered as uppercase, but can be all lower or mixed case within the document.
  • [ ] No special semantics for INT/EXT. LOCATION - TIME - EVEN MORE METADATA #7# which would be useful for editors.
• Synopses:
  • [ ] = This scene’s purpose.
• Actions:
  • [ ] !Forced action
  • [ ] >Centred Action Text<
• Transitions:
  • [ ] Incorrect grammar.
  • [ ] ANYTHING TO: or TO:, surrounded by empty lines.
  • [ ] > Forced transition text
• Characters:
  • [ ] @McFORCED CHARACTER
  • [ ] CHARACTER (EXTENSION) special semantics, useful for editors.
  • [ ] DUAL DIALOG MARKER ^
  • [ ] Uppercase text rule is not Unicode-compliant.
• Dialogue:
  • [ ] ~Lyrics special semantics.
• Text emphasising:
  • [ ] _underscored_
  • [ ] *italicised*
  • [ ] **bold**
  • [ ] Mixed.
  • [ ] Escaped \* \_ emphasis characters
• Page breaks:
  • [ ] === or =========…

Nice to haves:

• Actions & Dialog:
  • Special grammar for ALL CAPS WORDS or @McWORDS, useful for document processors to link up mentioned characters or locations.
• Scene headings:
  • Special-case well-known time markers DAY, NIGHT, CONTINUOUS.

Steps To Reproduce/Bad Parse Tree

Example structure for the scripts i’m writing:

Title:
    _**MY SCRIPT**_
Credit: Written by
Author: Me
Source: Based on That Thing
        by These Guys
Draft date: 1997-01-31

.PAGE. Location, Sublocation - NIGHT #1#

= Introduces this or that concept.

!**PANEL 1 - FULL ROW, NO BORDER**

Abusing action text as panel headings, scene headings as page headings, turning this into a comic book script. The `!` above should not appear as action text, for it’s a forced-action marker character.

SFX
poof

CHARACTER (V.O.)
Stuff happened.

> TO DOUBLE:

===

.PAGES①. Location, Sublocation - CONTINUOUS #2#

Custom scene transition with `>`, explicit page break.

> CONT:

===

>(CONTINUE)<
.PAGES②. Location, Sublocation - CONTINUOUS #3#

TO:

===

Expected Behavior/Parse Tree

Semantics of title page metadata, custom transitions, forced actions, etc. should’ve been recognised and preserved.

Repro

No response

[emalcxe]

3 meets Most of these tasks, I wrote a tree-sitter test based on the example. Text emphasis left out, and a few rules here and there

[Evrey]

Hi, awesome, thanks! I’ll take a look.