usenet-indexer/lib/spotnet/password.ts

/**
 * Archive passwords that spotters put in the title or description instead of
 * the NZB's `<meta type="password">`:
 *   `Title{{secret}}` (the newznab/SABnzbd convention),
 *   `Wachtwoord: KKKK2`, `[b]Password[/b]: x`, `WW: p3IU5Rz9R`, `ww= 0112358`,
 *   a few words between keyword and separator (`Het wachtwoord voor de nzb is: dw0067`,
 *   `ww in nzb: X`, `Wachtwoord inbegrepen: wysiwyg`),
 *   the value on the next line (`WW:` / `tijdloos`),
 *   `wachtwoord is arn2`, `… voorzien van een wachtwoord. Deze is: X`,
 *   `WW FC0401` (abbreviation without separator: with a digit in it, or as the whole line).
 * The value must be the rest of its line and a single token, so sentences
 * such as "Wachtwoord: deze download is zonder wachtwoord geplaatst" or
 * "met een password beschermd" don't count (LOG.md 2026-10-08 (10)).
 */
const KEYWORD = String.raw`(?:wachtwoord|wachtw|password|passwort|passwd|pwd|pass|ww|pw)(?![a-z])`
const WITH_SEPARATOR = new RegExp(String.raw`(?:^|[^a-z0-9])${KEYWORD}(?:\s+[a-zà-ÿ]+){0,3}?\s*(?:is\s*)?[:=]\s*(.*)$`, 'i')
const IS_VALUE = new RegExp(String.raw`(?:^|[^a-z0-9])${KEYWORD}\s+is\s+(\S+)\s*$`, 'i')
/** "Deze is voorzien van een wachtwoord. Deze is: X": the value in the next sentence. */
const NEXT_SENTENCE = /(?:^|[.!?]\s*)(?:deze|dit|het|die|this|it)\s+is\s*[:=]\s*(\S+)\s*$/i
const MENTIONS = new RegExp(String.raw`(?:^|[^a-z0-9])${KEYWORD}`, 'i')
const ABBREVIATION = /(?:^|[^a-z0-9])(?:ww|pw)\s+(\S+)\s*$/i
/** "WW carrie" as a line of its own; "WW II" is the war. */
const WHOLE_LINE_ABBREVIATION = /^(?:ww|pw)\s+(\S+)$/i
const NOT_PASSWORD_TOKEN = /^(?:i{1,3}|iv|v|[12])$/i
const NO_PASSWORD = /\b(zonder|geen|no|without|none|niet|kein|ohne)\b.{0,20}\b(wachtwoord|password|passwort|ww|pw)\b|\b(wachtwoord|password|passwort)\b.{0,20}\b(niet nodig|not needed|none|geen|nvt|n\/a)\b/i
const NOT_A_VALUE = /^(?:-+|geen|none|no|nee|nvt|n\/a|niet|not|leeg|empty|zonder|without|onbekend|unknown|zie|see|in|staat|volgt|nzb|spot)$/i

/** Lines of a spot description: UBB/HTML line breaks split, other tags dropped. */
function lines(description: string): string[] {
  return description
    .replace(/\[br\]|<br\s*\/?>/gi, '\n')
    .replace(/\[\/?[a-z]+(?:=[^\]]*)?\]|<\/?[a-z][^>]*>/gi, '')
    .split(/\r?\n/)
    .map((line) => line.trim())
}

function clean(value: string): string | null {
  const token = value
    .trim()
    // "Wachtwoord : yakowyke !!"
    .replace(/^[!*~#]+\s*|\s*[!*~#]+$/g, '')
    .replace(/^["'“‘«(<]+|["'”’»)>.,;!]+$/g, '')
  if (!token || /\s/.test(token) || token.length < 2 || token.length > 64 || NOT_A_VALUE.test(token)) return null
  return token
}

export function passwordFromSpot(title: string, description?: string | null): string | null {
  const braces = /\{\{([^{}]{1,64})\}\}/.exec(title)
  if (braces && clean(braces[1])) return clean(braces[1])
  const all = lines(description ?? '')
  for (const [i, line] of all.entries()) {
    if (!line || NO_PASSWORD.test(line)) continue
    const separated = WITH_SEPARATOR.exec(line)
    if (separated) {
      const value = separated[1].trim() ? clean(separated[1]) : nextLineValue(all, i)
      if (value) return value
      continue
    }
    const is = IS_VALUE.exec(line)
    if (is) {
      const value = clean(is[1])
      if (value) return value
    }
    const sentence = MENTIONS.test(line) ? NEXT_SENTENCE.exec(line) : null
    if (sentence) {
      const value = clean(sentence[1])
      if (value) return value
    }
    const whole = WHOLE_LINE_ABBREVIATION.exec(line)
    const abbreviated = whole ?? ABBREVIATION.exec(line)
    if (abbreviated && (whole ? !NOT_PASSWORD_TOKEN.test(abbreviated[1]) : /\d/.test(abbreviated[1]))) {
      const value = clean(abbreviated[1])
      if (value) return value
    }
  }
  return null
}

/** `WW:` with the value alone on the next non-empty line. */
function nextLineValue(all: string[], i: number): string | null {
  const next = all.slice(i + 1).find((line) => line !== '')
  return next ? clean(next) : null
}