formualizer-parse

High-performance Excel and OpenFormula tokenizer, parser, and pretty-printer.

formualizer-parse turns raw formula strings into a structured AST that downstream crates use for evaluation, analysis, and transformation. It handles both Excel and OpenFormula dialects with source location tracking.

When to use this crate

Use formualizer-parse when you need formula analysis without evaluation:

• Formula linting and validation
• Static analysis of cell dependencies
• AST transformation and rewriting
• Pretty-printing formulas to canonical form
• Building custom formula tooling

If you also need evaluation, use formualizer-workbook or formualizer-eval instead.

Quick start

use formualizer_parse::{FormulaDialect, Tokenizer, canonical_formula};
use formualizer_parse::parser::Parser;

// Tokenize and parse
let tokenizer = Tokenizer::new_with_dialect("=SUM(A1:B3)", FormulaDialect::Excel)?;
let mut parser = Parser::new(tokenizer.items, false);
let ast = parser.parse()?;

// Canonical form
assert_eq!(canonical_formula(&ast), "=SUM(A1:B3)");

Features

• Tokenization — streaming tokenizer with dialect-aware classification, source location tracking, and operator metadata.
• Pratt parser — precedence-climbing parser producing a stable AST with reference normalization.
• Dialects — Excel (default) and OpenFormula syntax support through a single API.
• Pretty-printing — canonicalize formulas or render diagnostic trees for debugging.
• Source spans — every token and AST node carries byte positions for precise error reporting.
• Fingerprinting — 64-bit structural hashes for formula identity comparison.

License

Dual-licensed under MIT or Apache-2.0, at your option.