The baskerville from jaynewey

baskerville's Introduction

Infer and validate data-type schemas in Rust and Python.

Installation

cargo add baskerville

Example

# mascots.csv
Name,LOC,Species
Ferris,42,Crab
Corro,7,Urchin

use baskerville::{infer_csv_with_options, CsvInput, InferOptions};

fn main() {
    let fields = infer_csv_with_options(
        CsvInput::Path("mascots.csv"),
        &mut InferOptions {
            has_headers: true,
            ..InferOptions::default()
        },
    )
    .unwrap();
    println!("{fields}");
}

Output:

╭──────┬─────────┬─────────╮
│ Name │ LOC     │ Species │
├──────┼─────────┼─────────┤
│ Text │ Integer │ Text    │
│      │ Float   │         │
│      │ Text    │         │
╰──────┴─────────┴─────────╯

Contributing

Versioning

The repo bases versioning from Conventional Commits

baskerville's People

Contributors

Watchers

baskerville's Issues

RegEx validator

Example usage:

let mut validator = RegEx::new("[a-z]");
assert!(validator.validate("a"));
assert!(!validator.validate("0"));

Document rust interfaces

Add doc comments to rust-side structs and functions

Recognise when a column has a limited set of values. For example, if we only see "Ferris" and "Corro" in a column for a file with many records, then perhaps that field should be described as Literal["Ferris", "Corro"]. We should pick this up by default. I think a default of up to 10 literal values is sensible.

Recommend Projects