five

divinetribe/es-en-offline-dictionary

收藏
Hugging Face2026-04-09 更新2026-04-12 收录
下载链接:
https://hf-mirror.com/datasets/divinetribe/es-en-offline-dictionary
下载链接
链接失效反馈
官方服务:
资源简介:
--- license: gpl-3.0 tags: - translation - spanish - english - offline - on-device - dictionary - idioms - phrases - rule-based - ios - mobile - no-cloud language: - es - en task_categories: - translation size_categories: - 100K<n<1M --- # Spanish-English Offline Dictionary ### 268,622 translations. 6,056 idioms. 4 MB. No internet. No excuses. --- ## The problem with translating Spanish word-by-word Try translating these literally. I dare you. | Spanish | What it literally says | What it actually means | |---|---|---| | *donde el diablo perdió el poncho* | where the devil lost his poncho | **in the middle of nowhere** | | *tener pájaros en la cabeza* | to have birds in your head | **to have bats in the belfry** | | *buscar tres pies al gato* | to look for three legs on a cat | **to look for trouble** | | *tener mente de pez* | to have a fish brain | **to have a memory like a sieve** | | *un ojo de la cara* | an eye from your face | **an arm and a leg** | | *no tener donde caerse muerto* | to not have a place to drop dead | **to not have a penny** | | *en boca cerrada no entran moscas* | flies don't enter a closed mouth | **silence is golden** | | *tener el mono* | to have the monkey | **to have withdrawal symptoms** | | *coser y cantar* | to sew and sing | **piece of cake** | | *más sabe el diablo por viejo que por diablo* | the devil knows more from being old than from being the devil | **experience is the best teacher** | Google Translate will give you the middle column. This dictionary gives you the right one. That's why it exists: **6,056 idiomatic phrases** where the whole is not the sum of the parts, backed by **268,622 word-level translations** for everything else. All in a 4 MB gzip that runs anywhere — phone, laptop, Raspberry Pi, airplane seat 23C with no WiFi. --- ## What's in the file One file: `es_final_with_rules.json.gz` (4.2 MB compressed, ~25 MB decompressed) ```json { "version": "3.1-merged", "dictionary": { ... }, // 268,622 word/sentence translations "phrases": { ... }, // 6,056 idiomatic phrases "rules": { ... } // translation guidelines + glossary } ``` ### The dictionary — 268,622 entries Everything from single words to full contextual sentences, each with an English translation and part-of-speech tag: ```json "abandonar": { "translation": "to abandon", "pos": "VERB" }, "abogado": { "translation": "lawyer", "pos": "NOUN" }, "sangre fría": { "translation": "cold blood", "pos": "NOUN" }, "loco de atar": { "translation": "stark raving mad", "pos": "ADJ" } ``` Many entries go beyond single words — conjugated verbs, prepositional phrases, regional variations, and full sentences with natural context: ```json "a decir verdad, esto es un poco demasiado picante para mi.": { "translation": "to tell you the truth, this is a little too spicy for me.", "pos": "UNKNOWN" } ``` **By the numbers:** | Part of speech | Entries | Flavor | |---|---|---| | Sentences & expressions | 213,954 | The bulk — contextual translations, multi-word | | Verbs | 22,299 | Including conjugations and verb phrases | | Nouns | 21,262 | People, places, things, concepts | | Adjectives | 9,133 | Descriptive, emotional, physical | | Adverbs | 703 | How, when, where | | Everything else | 1,471 | Pronouns, interjections, determiners, conjunctions | ### The idioms — 6,056 phrases that'll make you smile This is the part Google gets wrong. A curated collection of Spanish expressions, proverbs, and slang — the stuff that makes a language a *language* and not just vocabulary with grammar. **Animals say the darndest things:** | Spanish | English | Literally... | |---|---|---| | *estar como el perro y el gato* | to fight like cats and dogs | to be like the dog and the cat | | *gato con guantes no caza ratones* | you can't be too careful | a cat with gloves doesn't catch mice | | *perro ladrador, poco mordedor* | his bark is worse than his bite | barking dog, little biter | | *hablar como un loro* | to talk nonstop | to talk like a parrot | | *trabajar como un burro* | to work like a dog | to work like a donkey | | *tener más vidas que un gato* | to have nine lives | to have more lives than a cat | | *hacer el mono* | to monkey around | to do the monkey | | *tener un hambre de lobo* | to be starving | to have a wolf's hunger | | *travieso como un mono* | mischievous as a monkey | — | | *ser un gallina* | to be a chicken | to be a hen | **The devil gets around:** | Spanish | English | |---|---| | *ir como alma que lleva el diablo* | to go like a bat out of hell | | *donde el diablo perdió el poncho* | in the back of beyond | | *tener el diablo en el cuerpo* | to be full of mischief | | *sin encomendarse a dios ni al diablo* | without thinking twice | | *el diablo las carga* | speak of the devil | | *más sabe el diablo por viejo que por diablo* | experience is the best teacher | | *mandar al diablo* | to tell someone to go to hell | **The heart knows:** | Spanish | English | |---|---| | *ojos que no ven, corazón que no siente* | out of sight, out of mind | | *corazón de oro* | heart of gold | | *corazón de piedra* | heart of stone | | *robar el corazón* | to steal one's heart | | *dar un vuelco el corazón* | heart skips a beat | | *hacer de tripas corazón* | to pluck up courage | | *hablar con el corazón en la mano* | to speak from the heart | | *revista del corazón* | gossip magazine | **And the rest of living:** | Spanish | English | |---|---| | *echar leña al fuego* | to add fuel to the fire | | *ponerse como la grana* | to turn red as a beet | | *hacerse de oro* | to strike it rich | | *dormir como un lirón* | to sleep like a log | | *no pegar ojo* | to not sleep a wink | | *en un abrir y cerrar de ojos* | in the blink of an eye | | *tener un tornillo flojo* | to have a screw loose | | *tirar la toalla* | to throw in the towel | | *pedir el oro y el moro* | to ask for the moon | | *helar la sangre* | to make one's blood run cold | | *empezar la casa por el tejado* | to put the cart before the horse | | *matar dos pájaros de un tiro* | to kill two birds with one stone | | *sacar las castañas del fuego* | to pull someone's chestnuts out of the fire | --- ## Where this came from This dictionary powers the offline translation feature in [**RealTime AI Camera**](https://apps.apple.com/us/app/realtime-ai-cam/id6751230739) — a free iOS app that does 601-class object detection, live OCR, and Spanish→English translation, all on-device, no internet. The dictionary was compiled from multiple open sources, merged, deduplicated, POS-tagged, and expanded with contextual sentences. The idiom collection was curated separately because idiomatic phrases are exactly where word-by-word translation falls apart — and that's exactly when you need a dictionary most. The goal was specific: a dictionary that's **complete enough** for real-world camera text (signs, menus, documents, labels, handwritten notes), **small enough** to ship inside an iOS app (4 MB), and **fast enough** for real-time use (key-value lookup, not neural inference). It's a lookup table. No GPU needed. No model. No beam search. Sub-millisecond on any device made in the last decade. --- ## Use it ### Python ```python import gzip, json with gzip.open("es_final_with_rules.json.gz", "rt", encoding="utf-8") as f: data = json.load(f) dictionary = data["dictionary"] phrases = data["phrases"] # Always check phrases first — they catch idioms that # word-by-word lookup would mangle text = "tirar la toalla" if text in phrases: print(phrases[text]) # "to throw in the towel" elif text in dictionary: print(dictionary[text]["translation"]) # Single word print(dictionary["abogado"]["translation"]) # "lawyer" ``` ### Swift (iOS / macOS) ```swift import Foundation guard let url = Bundle.main.url(forResource: "es_final_with_rules", withExtension: "json.gz"), let compressed = try? Data(contentsOf: url), let json = try? (compressed as NSData).decompressed(using: .zlib), let root = try? JSONSerialization.jsonObject(with: json) as? [String: Any], let dictionary = root["dictionary"] as? [String: [String: String]], let phrases = root["phrases"] as? [String: String] else { return } // "to have a screw loose" print(phrases["tener un tornillo flojo"] ?? "not found") // "lawyer" print(dictionary["abogado"]?["translation"] ?? "not found") ``` For the full production implementation with grammar rules and smart fallbacks, see [`SpanishTranslationEngine.swift`](https://github.com/nicedreamzapp/RealTimeAICam/blob/main/SpanishTranslationEngine.swift). ### JavaScript ```javascript const zlib = require("zlib"); const fs = require("fs"); const data = JSON.parse( zlib.gunzipSync(fs.readFileSync("es_final_with_rules.json.gz")) ); const { dictionary, phrases } = data; // "in the middle of nowhere" console.log(phrases["donde el diablo perdió el poncho"]); // "lawyer" console.log(dictionary["abogado"].translation); ``` --- ## License **GPL-3.0** — same license as the parent project. Commercial licensing available for App Store / Play Store distribution. Contact **Matt Macosko** via [nicedreamzwholesale.com](https://nicedreamzwholesale.com/github-realtime-ai-camera/). ## Links - 📱 [RealTime AI Camera on the App Store (free)](https://apps.apple.com/us/app/realtime-ai-cam/id6751230739) - 💻 [Source code on GitHub](https://github.com/nicedreamzapp/RealTimeAICam) - 🤖 [YOLOv8n OIV7 CoreML model](https://huggingface.co/divinetribe/yolov8n-oiv7-coreml) — the detection model that ships alongside this dictionary - 🌐 [NiceDreamzApps](https://nicedreamzwholesale.com/github-realtime-ai-camera/) --- *Built by Matt Macosko / NiceDreamzApps — because "where the devil lost his poncho" should never be translated literally.*
提供机构:
divinetribe
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作