thunderbird2docuware/lib/mail.js

// Extraktion von Mail-Inhalten über die Thunderbird messages.* API.

const Mail = {
  /** Rohe RFC822-Nachricht als .eml-File. */
  async getEmlFile(messageId, subject) {
    const raw = await browser.messages.getRaw(messageId); // String (RFC822)
    const blob = new Blob([raw], { type: "message/rfc822" });
    const name = `${this._safeName(subject) || "email"}.eml`;
    return new File([blob], name, { type: "message/rfc822" });
  },

  /** Kopf + Body als strukturierte Metadaten zum Vorbefüllen. */
  async getMeta(messageId) {
    const header = await browser.messages.get(messageId);
    const full = await browser.messages.getFull(messageId);
    const bodyText = this._extractText(full);

    const sender = this._parseAddress(header.author);
    const account = await this._accountForMessage(messageId);

    return {
      rfcMessageId: this._header(full, "message-id"), // RFC822 Message-ID (Dedup-Schlüssel)
      subject: header.subject || "",
      senderEmail: sender.email,
      senderName: sender.name,
      to: (header.recipients || []).join(", "),
      cc: (header.ccList || []).join(", "),
      bcc: (header.bccList || []).join(", "),
      date: header.date instanceof Date ? header.date : new Date(header.date),
      bodyText,
      account,
      direction: this._direction(header),
      sizeBytes: header.size || 0,
    };
  },

  /** Liste der Anhänge: [{partName, name, contentType, size}]. */
  async listAttachments(messageId) {
    const atts = await browser.messages.listAttachments(messageId);
    return (atts || [])
      .filter((a) => a.partName && a.name) // echte Datei-Anhänge
      .map((a) => ({
        partName: a.partName,
        name: a.name,
        contentType: a.contentType,
        size: a.size || 0,
      }));
  },

  /** Anhang als File holen. */
  async getAttachmentFile(messageId, partName) {
    return browser.messages.getAttachmentFile(messageId, partName); // File
  },

  // --- Helfer ---------------------------------------------------------------

  /** Erster Wert eines Mail-Headers (getFull liefert Header als Arrays, klein geschrieben). */
  _header(full, name) {
    const h = full && full.headers && full.headers[name];
    return Array.isArray(h) && h.length ? String(h[0]).trim() : "";
  },

  _extractText(fullPart) {
    // Rekursiv den ersten text/plain-Teil suchen, sonst text/html entkernen.
    let plain = "";
    let html = "";
    const walk = (part) => {
      if (!part) return;
      if (part.contentType === "text/plain" && part.body) plain += part.body;
      else if (part.contentType === "text/html" && part.body) html += part.body;
      (part.parts || []).forEach(walk);
    };
    walk(fullPart);
    if (plain.trim()) return plain;
    if (html.trim()) return this._htmlToText(html);
    return "";
  },

  /**
   * Macht aus HTML lesbaren Klartext: entfernt style/script/head-INHALTE komplett,
   * leitet aus Block-Elementen Zeilenumbrüche ab, dekodiert Entities und glättet
   * Whitespace – ohne Absatzstruktur platt zu walzen.
   */
  _htmlToText(html) {
    let s = String(html);
    s = s.replace(/<!--[\s\S]*?-->/g, " ");
    // style/script/head/title inkl. Inhalt raus (sonst landet CSS/JS im Text).
    s = s.replace(/<(style|script|head|title)\b[^>]*>[\s\S]*?<\/\1\s*>/gi, " ");
    // Block-Grenzen -> Zeilenumbruch; Listenpunkte -> "- ".
    s = s.replace(/<\s*br\s*\/?>/gi, "\n");
    s = s.replace(/<\s*li\b[^>]*>/gi, "\n- ");
    s = s.replace(/<\/\s*(p|div|li|tr|h[1-6]|ul|ol|table|blockquote)\s*>/gi, "\n");
    s = s.replace(/<\/\s*td\s*>/gi, " ");
    // Restliche Tags entfernen, Entities dekodieren.
    s = s.replace(/<[^>]+>/g, " ");
    s = this._decodeEntities(s);
    // Whitespace glätten: Spaces/Tabs/NBSP zusammenfassen, Leerzeilen begrenzen.
    s = s.replace(/[ \t ]+/g, " ");
    s = s.replace(/ *\n */g, "\n");
    s = s.replace(/\n{3,}/g, "\n\n");
    return s.trim();
  },

  _decodeEntities(s) {
    const named = {
      amp: "&", lt: "<", gt: ">", quot: '"', apos: "'", nbsp: " ",
      auml: "ä", ouml: "ö", uuml: "ü", Auml: "Ä", Ouml: "Ö", Uuml: "Ü", szlig: "ß",
      euro: "€", copy: "©", reg: "®", trade: "™", hellip: "…", middot: "·", bull: "•",
      ndash: "–", mdash: "—", laquo: "«", raquo: "»", deg: "°", eacute: "é", agrave: "à",
      ldquo: "“", rdquo: "”", lsquo: "‘", rsquo: "’",
      bdquo: "„", sbquo: "‚",
    };
    return String(s)
      .replace(/&#x([0-9a-fA-F]+);/g, (_, h) => this._fromCp(parseInt(h, 16)))
      .replace(/&#(\d+);/g, (_, d) => this._fromCp(parseInt(d, 10)))
      .replace(/&([a-zA-Z][a-zA-Z0-9]*);/g, (m, n) => (named[n] != null ? named[n] : m));
  },

  _fromCp(cp) {
    try { return Number.isFinite(cp) ? String.fromCodePoint(cp) : ""; }
    catch (_) { return ""; }
  },

  _parseAddress(str) {
    if (!str) return { name: "", email: "" };
    const m = str.match(/^\s*"?([^"<]*)"?\s*<([^>]+)>\s*$/);
    if (m) return { name: m[1].trim(), email: m[2].trim() };
    return { name: "", email: str.trim() };
  },

  async _accountForMessage(messageId) {
    try {
      const msg = await browser.messages.get(messageId);
      const folder = msg.folder;
      if (folder && folder.accountId) {
        const acc = await browser.accounts.get(folder.accountId);
        return acc ? acc.name : "";
      }
    } catch (_) {}
    return "";
  },

  _direction(header) {
    // Heuristik über den Ordnertyp; robustere Erkennung via Identitäten käme später.
    const folderType = header.folder && header.folder.type;
    if (folderType === "sent" || folderType === "outbox") return "Ausgang";
    return "Eingang";
  },

  _safeName(s) {
    return (s || "").replace(/[\\/:*?"<>|]+/g, "_").slice(0, 80).trim();
  },
};

if (typeof module !== "undefined") module.exports = { Mail };