3 stable releases

1.0.2	Apr 28, 2024
1.0.1	Apr 27, 2024

#321 in Text processing

86 downloads per month

MIT license

1.5MB
1.5K SLoC

Santoka

arukeba kimpouge suwareba kimpouge
if i walk the buttercups if i sit the buttercups

Translations of 668 of Taneda Santōka's free-verse haiku, including excellent translations by Hiroaki Satō, Scott Watson, and Cid Corman.

Available as JSON for easy parsing, or in the Leaflet.md format for viewing in Markdown-friendly apps like Obsidian.

You can explore the poems in the dataset at lucaaurelia.com/santoka.

Source

Many of these poems were originally compiled and digitized by Gábor Terebess for Terebess Asia Online. I've added metadata like publication URLs and converted to a structured format for easy parsing.

Poems

See ./poems.json. Poems are in this format:

{
  "id": 68,
  "publicationId": 3,
  "englishText": "Absolutely no cloud I take off my hat",
  "japaneseText": "Mattaku kumo ga nai kasa o nugi"
}

The japaneseText field is missing for some poems, and can contain either romaji or kana/kanji.

Publications

See ./publications.json. Publications are in this format:

{
  "id": 12,
  "name": "Santôka",
  "translatorIds": [11, 16],
  "year": 2006,
  "description": "Santôka: A Translation with Photographic Images. Photographs by Hakudô Inoue; book and cover design by Kazuya Takaoka; English text by Emiko Miyashita and Paul Watsky. (PIE Books, Tokyo, 2006). 400 pages",
  "url": "https://thehaikufoundation.org/omeka/items/show/2643",
  "lucaRanking": 5
}

Field	Description
`name`	This is my best attempt to identify a single name for the publication, but it's occasionally a judgment call since the data includes personal web pages and other informal sources.
`translatorIds`	This is an array since sometimes publications have multiple people working on the English text, like in the example above. Publications with one translator (most of them) use a one-element array: `translatorIds: [8]`.
`year`	This is either a number, or `null` if I couldn't determine a publication year.
`description`	This is a free-form text description of the publication.
`url`	A somewhat authoritative URL for the publication, when I could find it. `null` if I couldn't.
`lucaRanking`	This is a subjective ranking based on where I want the publication to show up on lucaaurelia.com/santoka.

Translators

See ./translators.json. Translators are in this format:

{
  "id": 10,
  "name": "Scott Watson"
}

Installation

If you're using Rust, you can install this dataset from crates.io.

cargo add santoka

Example usage

Parsing JSON is straightforward in most languages. Here's a JavaScript example:

import fs from "fs";

const poemsJson = fs.readFileSync("./santoka/poems.json");
const poems = JSON.parse(poemsJson);
console.log(poems);

If you're using Rust, the santoka crate takes care of parsing for you:

fn main() {
    let dataset = santoka::Dataset::new();

    for poem in &dataset.poems {
        dbg!(&poem);

        let publication = dataset.publication(poem.publication_id);
        dbg!(&publication);

        let translators = dataset.translators(publication.translator_ids);
        dbg!(&translators);
    }
}