93 lines
5.1 KiB
JavaScript
93 lines
5.1 KiB
JavaScript
#!/usr/bin/env node
|
|
/**
|
|
* fetch-arclight-index.mjs — cache the JESUS Film Project's Arclight catalogue by ISO 639-3.
|
|
*
|
|
* THE SAME SHAPE THAT PAID OFF WITH EBIBLE. Arclight was ALREADY a checked pool, and only
|
|
* 306 language codes carried a film link. The catalogue lists 2,330 media languages, 2,087
|
|
* of them with a feature film. Being inside a pool is not the same as having been read.
|
|
*
|
|
* ★ featureFilm IS THE COVERAGE SIGNAL, NOT ROSTER MEMBERSHIP. A language appearing in
|
|
* media-languages does not mean the film exists in it; counts.featureFilm >= 1 does. 21 of
|
|
* the 2,330 are listed with no feature film and are dropped here.
|
|
*
|
|
* ★ THE WATCH URL CANNOT BE VERIFIED BY STATUS CODE, and this is the part worth reading.
|
|
* https://www.jesusfilm.org/watch/?languageId=<id> is a single-page app: it answers 200 for
|
|
* EVERY id, including `0` and `abcdef`. MEASURED — all three return 200. So a sample of
|
|
* watch-URLs "resolving 200" is a check that cannot fail and proves nothing at all about
|
|
* whether a film exists.
|
|
*
|
|
* The API is what discriminates: media-languages/<id> returns 200 with a real
|
|
* counts.featureFilm for a genuine language and 404 for a bogus one. So existence is
|
|
* established from the API, and the watch URL is the deep link for a language the API has
|
|
* already confirmed. Verifying it the other way round would have shipped 1,781 links on the
|
|
* strength of a status code that is always 200.
|
|
*
|
|
* ★ WHAT THE WATCH URL DOES AND DOES NOT DO — measured, because a confident wrong answer
|
|
* is available in BOTH directions here.
|
|
* - It cannot be verified by fetching. `?languageId=<id>` is a query parameter on a
|
|
* client-rendered page: the server returns a byte-identical 867,227-byte shell for a
|
|
* real id, a bogus id and every language alike. Grading it by STATUS says ALIVE for
|
|
* everything; grading it by CONTENT says DEAD for everything, because the phrase
|
|
* "WatchNotFound" lives in an i18n bundle embedded in every page. Both checks are
|
|
* vacuous and each looks authoritative.
|
|
* - Rendered in a real browser it does NOT land on the film in that language: id 529,
|
|
* id 1304 and id 999999999 all render the same generic library page. So the link is
|
|
* real and the film's existence is real, but the URL is not a language deep link.
|
|
* - No constructible scheme works either. The legacy /watch/jesus.html/<slug>.html form
|
|
* is retired (english.html 404s, not just rare languages), and /watch/<slug>.html/videos
|
|
* — a shape the live page itself renders — also 404s to a fetch, because it resolves
|
|
* client-side.
|
|
* THE API IS THE GROUND TRUTH, and the only thing here that discriminates: media-languages/<id>
|
|
* returns a real counts.featureFilm for a genuine language and 404 for a bogus one. Existence
|
|
* is asserted from that; the URL is offered as a way in, labelled for what it is.
|
|
*
|
|
* DOCTRINE: the JESUS Film is a dramatisation of Luke's Gospel by the JESUS Film Project.
|
|
* Inside the filter, and named explicitly in the standing instruction. Only featureFilm is
|
|
* taken; Arclight also carries short films and other media that are not screened here.
|
|
*/
|
|
import { writeSnapshot } from './_snapshot-guard.mjs'
|
|
import { resolve, dirname } from 'node:path'
|
|
import { fileURLToPath } from 'node:url'
|
|
|
|
const HERE = dirname(fileURLToPath(import.meta.url))
|
|
const OUT = resolve(HERE, '../../..', 'data/tables/arclight-index.json')
|
|
const API = 'https://api.arclight.org/v2/media-languages'
|
|
const UA = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36'
|
|
const WATCH = (id) => `https://www.jesusfilm.org/watch/?languageId=${id}`
|
|
|
|
const all = []
|
|
let page = 1, pages = 1
|
|
do {
|
|
const res = await fetch(`${API}?limit=1000&page=${page}`, { headers: { 'User-Agent': UA } })
|
|
if (!res.ok) throw new Error(`arclight page ${page}: HTTP ${res.status}`)
|
|
const d = await res.json()
|
|
pages = d.pages && d.limit ? Math.ceil(d.total / d.limit) : 1
|
|
all.push(...(d._embedded?.mediaLanguages || []))
|
|
page++
|
|
} while (page <= pages && page < 20)
|
|
|
|
const stats = { fetched: all.length, noFeatureFilm: 0, noIso: 0, kept: 0 }
|
|
const codes = {}
|
|
for (const r of all) {
|
|
const ff = Number(r?.counts?.featureFilm?.value || 0)
|
|
if (ff < 1) { stats.noFeatureFilm++; continue }
|
|
const iso = (r.iso3 || '').trim()
|
|
if (iso.length !== 3) { stats.noIso++; continue }
|
|
const prev = codes[iso]
|
|
// more speakers wins when one ISO code maps to several Arclight languages
|
|
const spk = Number(r?.counts?.speakerCount?.value || 0)
|
|
if (!prev || spk > prev.spk) {
|
|
codes[iso] = { u: WATCH(r.languageId), id: r.languageId, n: r.name || '', spk }
|
|
}
|
|
}
|
|
stats.kept = Object.keys(codes).length
|
|
|
|
writeSnapshot(OUT, JSON.stringify({
|
|
_note: 'Arclight (JESUS Film Project) catalogue by ISO 639-3. Only languages with counts.featureFilm >= 1. Existence is established from the API: the /watch/ URL answers 200 for ANY id and cannot verify anything.',
|
|
generated: new Date().toISOString().slice(0, 10),
|
|
source: API,
|
|
stats,
|
|
codes,
|
|
}), (j) => Object.keys(j.codes || {}).length, 'Arclight ISO 639-3 codes')
|
|
console.log(JSON.stringify(stats))
|
|
console.log(`${stats.kept} ISO 639-3 codes -> ${OUT}`)
|