README.md

metadata

license: cc0-1.0
language:
  - en
tags:
  - screenwriting
  - screenplay
  - filmmaking
  - nlp
  - text-classification
size_categories:
  - n<1K

Screenplay Format Conventions Dataset

Structured reference data describing the conventions used in Fountain and Final Draft (.fdx) screenplay formats. Useful for:

• NLP/ML tasks — classifying screenplay elements (scene heading, action, character cue, dialogue, parenthetical, transition)
• AI pre-production tools — preprocessing scripts for downstream image generation
• Education — teaching screenplay formatting to junior writers
• Format converters — building bidirectional Fountain ⇄ FDX tools

Files

• element_types.json — All screenplay element types with regex patterns and examples
• scene_heading_grammar.json — INT/EXT prefixes, location patterns, time of day vocabulary
• transition_keywords.json — Standard transition markers (CUT TO, DISSOLVE TO, FADE OUT, etc.)
• character_cue_rules.json — Conventions for character names, extensions (V.O., O.S., CONT'D)
• industry_glossary.json — Common screenwriting terms and abbreviations

License

CC0 — public domain.

Maintained by

STORYLINER — AI storyboard generator. We use the same conventions internally to parse user-uploaded screenplays. Open-source library: github.com/mcqx4/screenplay-parser.