usenet-indexer/app/services/FileStore.ts

import db from '@adonisjs/lucid/services/db'
import FileKeys from '#services/FileKeys'
import type { QueryClientContract } from '@adonisjs/lucid/types/database'
import { packSegments, unpackSegments, type Segment } from '#lib/nzb/segments'

/**
 * The only code that knows how `files` keeps its segment lists (CONCEPT.md
 * D31): compressed in `files.segments`, with the first segment's message-id
 * in the indexed `first_message_id` for lookups. Read them with
 * `segmentsOf(file)`; write new files with `FileStore.insert`.
 */

/** A new `files` row, with its segments as a list. */
export type NewFile = Record<string, unknown> & { segments: Segment[] }

const unpacked = new WeakMap<object, Segment[]>()

/** The segments of a `files` row (a row from `db.from('files')`), in order. */
export function segmentsOf(file: { segments?: Buffer | Segment[] | null }): Segment[] {
  const { segments } = file
  if (!segments) return []
  if (Array.isArray(segments)) return segments
  let list = unpacked.get(file)
  if (!list) {
    list = unpackSegments(segments)
    unpacked.set(file, list)
  }
  return list
}

/** Stored form of a segment list, for the File model's column. */
export function storedSegments(segments: Segment[]) {
  return packSegments(segments)
}

export function loadedSegments(stored: Buffer | null): Segment[] {
  return stored ? unpackSegments(stored) : []
}

/** What inspection learns about a file, kept beside `files` (LOG.md 2026-10-08 (14)). */
export interface FileDetails {
  content_type?: string | null
  archive?: unknown
}

const DETAIL_COLUMNS = ['content_type', 'archive'] as const

/** Rows per INSERT, well below Postgres' 65,535 bind parameters (tests lower it). */
export const insertLimits = { rows: 2000 }

class FileStore {
  /** Sets some of a file's details, leaving the others as they are. */
  async setDetails(fileId: number, details: FileDetails, client: Pick<QueryClientContract, 'rawQuery'> = db) {
    const columns = DETAIL_COLUMNS.filter((c) => details[c] !== undefined)
    if (columns.length === 0) return
    const values = columns.map((c) => (c === 'archive' && details[c] !== null ? JSON.stringify(details[c]) : details[c]))
    await client.rawQuery(
      `INSERT INTO file_details (file_id, ${columns.join(', ')}) VALUES (?, ${columns.map(() => '?').join(', ')})
       ON CONFLICT (file_id) DO UPDATE SET ${columns.map((c) => `${c} = excluded.${c}`).join(', ')}`,
      [fileId, ...(values as any[])]
    )
  }

  /** A file's details (empty when it has none). */
  async details(fileId: number): Promise<FileDetails> {
    return (await db.from('file_details').where('file_id', fileId).select('content_type', 'archive').first()) ?? {}
  }

  /**
   * Inserts files, skipping those whose `file_key` exists. Returns the rows
   * that were inserted (id, filename, file_key). `files` is partitioned by id
   * (CONCEPT.md D39) and each partition has its own unique file_key index:
   * once file_keys (a 64-bit hash of every key) is complete, it says which keys
   * may exist, and only those are looked up in the partitions; two
   * writers adding the same new key at once meet in the newest partition's index.
   */
  async insert(
    files: NewFile[],
    client: Pick<QueryClientContract, 'knexQuery' | 'rawQuery'> = db
  ): Promise<{ id: number; filename: string; file_key: string | null }[]> {
    if (files.length === 0) return []
    const keys = [...new Set(files.map((file) => file.file_key).filter((key): key is string => typeof key === 'string'))]
    const known = new Set<string>()
    const filtered = await FileKeys.complete()
    for (let i = 0; i < keys.length; i += 5000) {
      const chunk = keys.slice(i, i + 5000)
      if (!filtered) {
        const found = await client.knexQuery().from('files').whereIn('file_key', chunk).select('file_key')
        for (const row of found) known.add(row.file_key)
        continue
      }
      const { rows } = await client.rawQuery(
        `SELECT k.key FROM unnest(?::text[]) AS k(key)
         WHERE EXISTS (SELECT 1 FROM file_keys f WHERE f.key_hash = hashtextextended(k.key, 0))`,
        [chunk]
      )
      const maybe = rows.map((row: { key: string }) => row.key)
      if (maybe.length === 0) continue
      const found = await client.knexQuery().from('files').whereIn('file_key', maybe).select('file_key')
      for (const row of found) known.add(row.file_key)
    }
    const seen = new Set<string>()
    const rows = files
      .filter((file) => {
        const key = file.file_key as string | null | undefined
        if (!key) return true
        if (known.has(key) || seen.has(key)) return false
        seen.add(key)
        return true
      })
      .map(({ segments, ...file }) => {
        const sorted = [...segments].sort((a, b) => a.number - b.number)
        return { ...file, segments: packSegments(sorted), first_message_id: sorted[0]?.id ?? null }
      })
    if (rows.length === 0) return []
    // Details given with a new file (e.g. a known content type) go to file_details.
    const details = rows.map((row: Record<string, unknown>) => {
      const d: FileDetails = {}
      for (const c of DETAIL_COLUMNS) if (c in row) (d as any)[c] = row[c]
      for (const c of DETAIL_COLUMNS) delete row[c]
      return d
    })
    // A statement takes at most 65,535 parameters: ~5,400 rows of 12 columns.
    const inserted: { id: number; filename: string; file_key: string | null }[] = []
    for (let i = 0; i < rows.length; i += insertLimits.rows) {
      const chunk = await client.knexQuery().table('files').insert(rows.slice(i, i + insertLimits.rows)).onConflict().ignore().returning(['id', 'filename', 'file_key'])
      for (const row of chunk) inserted.push({ id: Number(row.id), filename: row.filename, file_key: row.file_key ?? null })
    }
    const detailsByKey = new Map(rows.map((row: Record<string, unknown>, i: number) => [row.file_key ?? row.filename, details[i]]))
    for (const row of inserted) {
      const d = detailsByKey.get(row.file_key ?? row.filename)
      if (d && Object.keys(d).length > 0) await this.setDetails(row.id, d, client)
    }
    return inserted
  }
}

export default new FileStore()