237 lines
14 KiB
TypeScript
237 lines
14 KiB
TypeScript
import { test } from '@japa/runner'
|
|
import db from '@adonisjs/lucid/services/db'
|
|
import testUtils from '@adonisjs/core/services/test_utils'
|
|
import { Par2Merger } from '#services/par2_merger'
|
|
import { encodeChunk } from '#lib/chunks/index'
|
|
import ChunkPartitions from '#services/chunk_partitions'
|
|
|
|
const hash = (c: string) => c.repeat(32)
|
|
|
|
test.group('Par2Merger', (group) => {
|
|
group.each.setup(() => testUtils.db().withGlobalTransaction())
|
|
|
|
test('merges the one-file releases of an obfuscated upload by its par2 set, named by the par2', async ({ assert }) => {
|
|
const at = new Date('2022-07-08T03:46:00Z')
|
|
const release = async (key: string, posted: Date, groupId: number, files = 1) => {
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({
|
|
release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: key, source: 'stem', groups: '[]', group_id: groupId,
|
|
files, segments: files * 10, bytes: files * 1000, first_article: 100, last_article: 200, posted_at: posted, readable: false, file_total: 20,
|
|
settled_at: posted,
|
|
})
|
|
.returning('id')
|
|
return Number(row.id)
|
|
}
|
|
// The par2 index, two data files of the upload, the same file reposted a week later, and one in another group.
|
|
const par2 = await release('14f5999a', at, 1)
|
|
const a = await release('2b7c00e1', new Date(at.getTime() + 60_000), 1)
|
|
const b = await release('9d01aa3f', new Date(at.getTime() + 120_000), 1, 2)
|
|
const later = await release('5e5e5e5e', new Date(at.getTime() + 7 * 86400_000), 1)
|
|
const elsewhere = await release('6f6f6f6f', at, 2)
|
|
let candidate = 1000
|
|
const hashed = async (releaseId: number, h: string, size: number) => {
|
|
candidate++
|
|
await db.table('probe_candidates').insert({ id: candidate, release_id: releaseId, kind: 'lone', message_id: `<${candidate}@x>`, file: 'x', status: 'done' })
|
|
await db.table('file_hashes').insert({ candidate_id: candidate, release_id: releaseId, hash16k: h, size })
|
|
}
|
|
await hashed(a, hash('a'), 104857600)
|
|
await hashed(b, hash('b'), 104857600)
|
|
await hashed(later, hash('a'), 104857600)
|
|
await hashed(elsewhere, hash('b'), 104857600)
|
|
// A different size: not the same file.
|
|
await hashed(await release('7a7a7a7a', at, 1), hash('a'), 5)
|
|
const setId = 'c'.repeat(32)
|
|
await db.table('par2_files').multiInsert([
|
|
{ set_id: setId, file_id: 'd'.repeat(32), hash16k: hash('a'), length: 104857600, name: 'Some.Movie.2022.1080p.part01.rar' },
|
|
{ set_id: setId, file_id: 'e'.repeat(32), hash16k: hash('b'), length: 104857600, name: 'Some.Movie.2022.1080p.part02.rar' },
|
|
])
|
|
await db.table('par2_sets').insert({ set_id: setId, release_id: par2, files: 2 })
|
|
|
|
const round = await new Par2Merger().round()
|
|
assert.deepEqual(round, { sets: 1, merges: 1, releases: 3, stems: 0, stemMerges: 0, stemReleases: 0, folded: { looked: 0, sidecars: 0, folded: 0 } })
|
|
const target = await db.from('releases').where('id', b).firstOrFail()
|
|
assert.equal(target.files, 4)
|
|
assert.equal(target.segments, 40)
|
|
assert.equal(target.name, 'Some.Movie.2022.1080p')
|
|
assert.isTrue(target.complete === false)
|
|
assert.deepEqual((await db.from('releases').whereIn('id', [par2, a]).select('id')).length, 0)
|
|
assert.equal((await db.from('releases').whereIn('id', [later, elsewhere]).select('id')).length, 2)
|
|
assert.deepEqual((await db.from('probe_candidates').where('release_id', b).orderBy('id').select('id')).map((r) => Number(r.id)), [1001, 1002])
|
|
assert.equal((await db.from('par2_sets').where('set_id', setId).firstOrFail()).release_id, String(b))
|
|
assert.equal((await db.from('release_merges').where('into_id', b).count('* as n'))[0].n, '2')
|
|
|
|
// Again: nothing new.
|
|
assert.deepEqual(await new Par2Merger().round().then((r) => r.merges), 0)
|
|
})
|
|
|
|
test('a restart goes on from the last whole round, kept in worker_cursors', async ({ assert }) => {
|
|
const hourAgo = new Date(Date.now() - 3600_000)
|
|
await db.table('par2_sets').insert({ set_id: 'f'.repeat(32), release_id: 1, files: 2, created_at: hourAgo })
|
|
// Never run: everything.
|
|
assert.equal((await new Par2Merger().round()).sets, 1)
|
|
const saved = await db.from('worker_cursors').where('name', 'merger').firstOrFail()
|
|
assert.isAbove(new Date(saved.at).getTime(), hourAgo.getTime())
|
|
// A new process (a deploy) reads where the last round started: the set an hour old is behind it.
|
|
assert.equal((await new Par2Merger().round()).sets, 0)
|
|
await db.from('worker_cursors').where('name', 'merger').update({ at: new Date(hourAgo.getTime() - 60_000) })
|
|
assert.equal((await new Par2Merger().round()).sets, 1)
|
|
})
|
|
|
|
test('a set held by two of the releases merged is held once by the target', async ({ assert }) => {
|
|
const at = new Date('2022-07-08T03:46:00Z')
|
|
const ids: number[] = []
|
|
for (const [key, files] of [['bb000001', 1], ['bb000002', 1], ['bb000003', 3]] as const) {
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: key, source: 'stem', groups: '[]', group_id: 1, files, posted_at: at, readable: false })
|
|
.returning('id')
|
|
ids.push(Number(row.id))
|
|
}
|
|
const setId = hash('7')
|
|
// The par2 index and its repost each hold the set; the third release holds the data file.
|
|
await db.table('par2_sets').multiInsert([
|
|
{ set_id: setId, release_id: ids[0], files: 1 },
|
|
{ set_id: setId, release_id: ids[1], files: 1 },
|
|
])
|
|
await db.table('par2_files').insert({ set_id: setId, file_id: hash('6'), hash16k: hash('5'), length: 1000, name: 'x.rar' })
|
|
await db.table('file_hashes').insert({ candidate_id: 3001, release_id: ids[2], hash16k: hash('5'), size: 1000 })
|
|
assert.equal(await new Par2Merger().mergeSet(setId), 2)
|
|
const held = await db.from('par2_sets').where('set_id', setId).select('release_id')
|
|
assert.deepEqual(held.map((r) => Number(r.release_id)), [ids[2]])
|
|
})
|
|
|
|
test('a set held by many releases nearby is posted again and again: not merged', async ({ assert }) => {
|
|
const at = new Date('2022-07-08T03:46:00Z')
|
|
const setId = hash('4')
|
|
const ids: number[] = []
|
|
for (let i = 1; i <= 6; i++) {
|
|
const key = `cc00000${i}`
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: `kantodo${i}`, source: 'stem', groups: '[]', group_id: 1, files: 11, posted_at: at, readable: true })
|
|
.returning('id')
|
|
ids.push(Number(row.id))
|
|
await db.table('par2_sets').insert({ set_id: setId, release_id: row.id, files: 1 })
|
|
}
|
|
await db.table('par2_files').insert({ set_id: setId, file_id: hash('3'), hash16k: hash('2'), length: 1073741824, name: 'kantodo.bin' })
|
|
assert.equal(await new Par2Merger().mergeSet(setId), 0)
|
|
assert.equal((await db.from('releases').whereIn('id', ids).select('id')).length, 6)
|
|
})
|
|
|
|
test('merges the one-file releases of an obfuscated upload by the random stem of their inner names', async ({ assert }) => {
|
|
const at = new Date('2022-07-07T19:34:00Z')
|
|
let n = 0
|
|
const release = async (poster: string, posted: Date, total: number, inner: string | null, groupId = 1) => {
|
|
n++
|
|
const key = `dd${String(n).padStart(6, '0')}`
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: `hex${n}`, source: 'stem', groups: '[]', group_id: groupId, files: 1, posted_at: posted, readable: false, file_total: total, poster })
|
|
.returning('id')
|
|
if (inner) {
|
|
await db.table('release_inspections').insert({ release_id: row.id, candidate_id: 5000 + n, kind: 'lone', result: JSON.stringify({ yencName: inner, type: '', size: 1 }) })
|
|
}
|
|
return Number(row.id)
|
|
}
|
|
const hep = 'hep <adsjfhsdh@sdljf>'
|
|
const stem = 'nabNGjSlKhOYJCslnEw1MzJIgPIjffyT'
|
|
const a = await release(hep, at, 15, `${stem}.7z.001`)
|
|
const b = await release(hep, new Date(at.getTime() + 60_000), 15, `${stem}.7z.002`)
|
|
const c = await release(hep, new Date(at.getTime() + 120_000), 15, `${stem}.vol07+08.par2`)
|
|
// Another poster, a week later, another total, another group: apart.
|
|
const other = await release('someone <x@y>', at, 15, `${stem}.7z.003`)
|
|
const later = await release(hep, new Date(at.getTime() + 7 * 86400_000), 15, `${stem}.7z.004`)
|
|
const total = await release(hep, at, 16, `${stem}.7z.005`)
|
|
const group = await release(hep, at, 15, `${stem}.7z.006`, 2)
|
|
// A readable stem isn't merged on: a repost would share it.
|
|
const r1 = await release(hep, at, 15, 'SomeReadableMovieName.7z.001')
|
|
const r2 = await release(hep, at, 15, 'SomeReadableMovieName.7z.002')
|
|
|
|
const round = await new Par2Merger().round()
|
|
assert.equal(round.stemMerges, 1)
|
|
assert.equal(round.stemReleases, 3)
|
|
const left = (await db.from('releases').whereIn('id', [a, b, c]).select('id', 'files')).map((r) => Number(r.files))
|
|
assert.deepEqual(left, [3])
|
|
assert.equal((await db.from('releases').whereIn('id', [other, later, total, group, r1, r2]).select('id')).length, 6)
|
|
const merges = await db.from('release_merges').where('via', 'inner').select('set_id')
|
|
assert.deepEqual([...new Set(merges.map((m) => m.set_id))], [stem])
|
|
})
|
|
|
|
test('an inner-name merge with more files than the counter is more than one upload', async ({ assert }) => {
|
|
const at = new Date('2022-07-07T19:34:00Z')
|
|
const stem = 'R8aUMXqWQfOaei4u74fch9v0skxUY7ah'
|
|
const ids: number[] = []
|
|
for (let i = 1; i <= 3; i++) {
|
|
const key = `ee00000${i}`
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: key, source: 'stem', groups: '[]', group_id: 1, files: 10, posted_at: at, readable: false, file_total: 5, poster: 'p' })
|
|
.returning('id')
|
|
ids.push(Number(row.id))
|
|
await db.table('release_inspections').insert({ release_id: row.id, candidate_id: 6000 + i, kind: 'lone', result: JSON.stringify({ yencName: `${stem}.7z.00${i}` }) })
|
|
}
|
|
assert.equal(await new Par2Merger().mergeStem(stem), 0)
|
|
assert.equal((await db.from('releases').whereIn('id', ids).select('id')).length, 3)
|
|
})
|
|
|
|
test('a set file that matches several releases nearby is ambiguous: not merged', async ({ assert }) => {
|
|
const at = new Date('2022-07-08T03:46:00Z')
|
|
const ids: number[] = []
|
|
for (const key of ['aa000001', 'aa000002', 'aa000003']) {
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key.padEnd(40, '0'), identity_key: key.padEnd(40, '0'), name: key, source: 'stem', groups: '[]', group_id: 1, files: 1, posted_at: at, readable: false })
|
|
.returning('id')
|
|
ids.push(Number(row.id))
|
|
}
|
|
// Two test uploads whose 1 GiB files start alike, and a set describing one such file, held by the third.
|
|
await db.table('file_hashes').multiInsert([
|
|
{ candidate_id: 2001, release_id: ids[0], hash16k: hash('f'), size: 1073741824 },
|
|
{ candidate_id: 2002, release_id: ids[1], hash16k: hash('f'), size: 1073741824 },
|
|
])
|
|
await db.table('par2_files').insert({ set_id: hash('9'), file_id: hash('8'), hash16k: hash('f'), length: 1073741824, name: 'kantibanito1.bin' })
|
|
await db.table('par2_sets').insert({ set_id: hash('9'), release_id: ids[2], files: 1 })
|
|
assert.equal((await new Par2Merger().round()).merges, 0)
|
|
assert.equal((await db.from('releases').whereIn('id', ids).select('id')).length, 3)
|
|
})
|
|
|
|
test('an .nzb or .nfo apart from its upload is folded into it', async ({ assert }) => {
|
|
await ChunkPartitions.ensure()
|
|
const at = new Date('2022-07-08T03:46:00Z')
|
|
const poster = 'up <up@load.er>'
|
|
let n = 0
|
|
const release = async (name: string, files: number, total: number | null, posted: Date, chunkFiles: string[]) => {
|
|
n++
|
|
const key = `ff${String(n).padStart(6, '0')}`.padEnd(40, '0')
|
|
const [row] = await db
|
|
.table('releases')
|
|
.insert({ release_key: key, identity_key: key, name, source: 'stem', groups: '[]', group_id: 1, files, file_total: total, poster, posted_at: posted, readable: true, settled_at: posted })
|
|
.returning('id')
|
|
const id = Number(row.id)
|
|
const payload = encodeChunk(chunkFiles.map((f, i) => ({ name: f, poster, subject: `${name} [${String(i).padStart(2, '0')}/${total ?? 0}] - "${f}" yEnc`, postedAt: posted.getTime(), groups: ['a.b.test'], segments: [[1, `${id}-${i}@x`, 1000]] })))
|
|
await db.table('release_chunks').insert({ release_id: id, range_id: 1, batch_from: 1, files: chunkFiles.length, segments: chunkFiles.length, payload })
|
|
await db.table('fold_candidates').insert({ release_id: id }).onConflict('release_id').ignore()
|
|
return id
|
|
}
|
|
const upload = await release('Murdoch Mysteries S8D1', 94, 94, at, ['Murdoch Mysteries S8D1.part01.rar', 'Murdoch Mysteries S8D1.part02.rar', 'Murdoch Mysteries S8D1.part03.rar', 'Murdoch Mysteries S8D1.part04.rar'])
|
|
const other = await release('Another Show S1D1', 50, 50, new Date(at.getTime() + 120_000), ['Another Show S1D1.part01.rar', 'Another Show S1D1.part02.rar', 'Another Show S1D1.part03.rar', 'Another Show S1D1.part04.rar'])
|
|
const nzb = await release('Murdoch Mysteries S8D1', 1, 80, new Date(at.getTime() + 60_000), ['Murdoch Mysteries S8D1.nzb'])
|
|
// A picture isn't a sidecar; an nfo an hour later belongs to nothing here.
|
|
const picture = await release('cover', 1, null, at, ['cover.jpg'])
|
|
const late = await release('Murdoch Mysteries S8D1', 1, null, new Date(at.getTime() + 3600_000), ['Murdoch Mysteries S8D1.nfo'])
|
|
// Queued, its release gone meanwhile (merged): leaves the queue with its batch.
|
|
await db.table('fold_candidates').insert({ release_id: 987654321 })
|
|
|
|
const round = await new Par2Merger().round()
|
|
assert.deepEqual(round.folded, { looked: 5, sidecars: 2, folded: 1 })
|
|
assert.isNull(await db.from('releases').where('id', nzb).first())
|
|
assert.equal((await db.from('releases').where('id', upload).firstOrFail()).files, 95)
|
|
assert.equal((await db.from('releases').where('id', other).firstOrFail()).files, 50)
|
|
assert.isNotNull(await db.from('releases').where('id', picture).first())
|
|
assert.isNotNull(await db.from('releases').where('id', late).first())
|
|
assert.equal((await db.from('release_merges').where({ into_id: upload, via: 'sidecar' }).count('* as c'))[0].c, '1')
|
|
assert.equal((await db.from('fold_candidates').count('* as c'))[0].c, '0')
|
|
})
|
|
})
|