{
  "_comment": "Build manifest for the Dictionary Genie content pack, published under the CC BY-SA 4.0 share-alike obligation described at https://dictionarygenie.app/sources#pack. Regenerated by pipeline/install_pack.sh on every pack build; do not edit by hand.",
  "pack_id": "lexicon-en",
  "file": "lexicon-en.sqlite",
  "pack_version": "1.0.0",
  "content_version": "2026-09-02",
  "schema_version": 2,
  "built_at": "2026-09-07T01:51:45Z",
  "sha256": "747e5ec52c0b3a028c0067093b814bbae7d51d6b6cf23d56b1a9e3503f72bfd0",
  "size_bytes": 432463872,
  "row_counts": {
    "daily_phrase": 1830,
    "daily_word": 366,
    "etymology": 66864,
    "example": 204006,
    "example_translation": 265723,
    "language": 23,
    "lexeme": 223768,
    "pronunciation": 126271,
    "relation": 733614,
    "sense": 437666,
    "source": 6,
    "suggest": 223768,
    "translation": 751353,
    "word_form": 349167
  },
  "languages": [
    "en",
    "es",
    "fr",
    "de",
    "it",
    "pt",
    "ru",
    "ja",
    "ko",
    "zh",
    "nl",
    "pl",
    "uk",
    "tr",
    "ar",
    "hi",
    "id",
    "vi",
    "th",
    "sv",
    "el",
    "he",
    "la"
  ],
  "selection": {
    "min_zipf": 1.8,
    "strict_zipf": false,
    "tier_a_zipf": 3.6,
    "tier_b_zipf": 2.3,
    "tier_counts": {
      "A": 57008,
      "B": 111156,
      "C": 55604
    }
  },
  "pronunciation": {
    "engine": "pronounce",
    "style_version": "en-respell-1"
  },
  "sources": [
    {
      "id": "cmudict",
      "name": "CMU Pronouncing Dictionary",
      "dataset": "cmusphinx/cmudict cmudict.dict",
      "version": "git commit 0f8072f814306c5ee4fbf992ed853601b12c01f9 (2024-12-17; verified: file at this commit has the recorded SHA-256)",
      "retrieved_at": "2026-09-02T21:37:12Z",
      "license_id": "BSD-2-Clause",
      "license_url": "https://github.com/cmusphinx/cmudict/blob/master/LICENSE",
      "source_url": "https://raw.githubusercontent.com/cmusphinx/cmudict/master/cmudict.dict",
      "attribution_text": "ARPAbet pronunciations from the CMU Pronouncing Dictionary. Copyright (C) 1993-2015 Carnegie Mellon University. All rights reserved. Redistributed under the BSD 2-Clause License.",
      "share_alike": false
    },
    {
      "id": "editorial",
      "name": "Dictionary Genie editorial",
      "dataset": "data/editorial (Word of the Day notes, Latin phrases, pronunciation overrides)",
      "version": null,
      "retrieved_at": null,
      "license_id": "proprietary",
      "license_url": null,
      "source_url": null,
      "attribution_text": "Editorial content © 2026 Omnia Data Analytics LLC.",
      "share_alike": false
    },
    {
      "id": "oewn",
      "name": "Open English WordNet",
      "dataset": "Open English WordNet 2024, WN-LMF XML (english-wordnet-2024.xml.gz)",
      "version": "2024",
      "retrieved_at": "2026-09-02T21:37:12Z",
      "license_id": "CC-BY-4.0",
      "license_url": "https://creativecommons.org/licenses/by/4.0/",
      "source_url": "https://en-word.net/static/english-wordnet-2024.xml.gz",
      "attribution_text": "Thesaurus relations from Open English WordNet 2024 (https://en-word.net), licensed under CC BY 4.0. Based on Princeton WordNet 3.0 (Princeton University).",
      "share_alike": false
    },
    {
      "id": "tatoeba",
      "name": "Tatoeba",
      "dataset": "Tatoeba per-language exports: eng_sentences_detailed.tsv.bz2, eng-<lang>_links.tsv.bz2 and <lang>_sentences_detailed.tsv.bz2 for 22 languages",
      "version": "per-language export downloaded 2026-09-02",
      "retrieved_at": "2026-09-02T21:37:12Z",
      "license_id": "CC-BY-2.0-FR",
      "license_url": "https://creativecommons.org/licenses/by/2.0/fr/",
      "source_url": "https://downloads.tatoeba.org/exports/per_language/",
      "attribution_text": "Example sentences and translations from Tatoeba (https://tatoeba.org), licensed under CC BY 2.0 FR. Each sentence is attributed to its Tatoeba contributor.",
      "share_alike": false
    },
    {
      "id": "wiktionary",
      "name": "Wiktionary (English)",
      "dataset": "kaikki.org English Wiktionary extract (wiktextract), kaikki.org-dictionary-English.jsonl.gz",
      "version": "kaikki dump 2026-08-28 (gzip header mtime 2026-08-28T13:29:09Z)",
      "retrieved_at": "2026-09-02T21:37:12Z",
      "license_id": "CC-BY-SA-4.0",
      "license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
      "source_url": "https://kaikki.org/dictionary/English/kaikki.org-dictionary-English.jsonl.gz",
      "attribution_text": "Definitions, examples, pronunciations, etymologies, forms, sense relations and translations from Wiktionary (https://en.wiktionary.org), extracted by kaikki.org (wiktextract), licensed under CC BY-SA 4.0.",
      "share_alike": true
    },
    {
      "id": "wordfreq",
      "name": "wordfreq",
      "dataset": "wordfreq Python package, English frequency lists",
      "version": "3.1.1",
      "retrieved_at": "2026-09-02T21:33:12Z",
      "license_id": "MIT",
      "license_url": "https://github.com/rspeer/wordfreq/blob/master/LICENSE.txt",
      "source_url": "https://pypi.org/project/wordfreq/3.1.1/",
      "attribution_text": "Word frequencies from wordfreq by Robyn Speer (MIT License), which combines open corpora including Wikipedia, OpenSubtitles, Google Books Ngrams and Tatoeba.",
      "share_alike": false
    }
  ],
  "derived_data_license": {
    "license_id": "CC-BY-SA-4.0",
    "license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
    "note": "The derived lexical data in this pack is offered under CC BY-SA 4.0. The application, its interface, artwork, name and editorial content are not.",
    "request": "alend@omniadataanalytics.com"
  }
}