gpt_chrome_autodrafter_v3/email_parser.js

/**
 * Parse a MIME email and extract its components
 * @param {string} mimeContent - The raw MIME content of the email
 * @returns {Object} Parsed email with headers, text and HTML content
 */
function parseMimeEmail(mimeContent) {
  console.log('Starting MIME email parsing...');
  
  // Extract headers and body
  const parts = mimeContent.split(/\r?\n\r?\n/, 2);
  const headers = parseHeaders(parts[0]);
  const body = parts.length > 1 ? parts.slice(1).join('\n\n') : '';
  
  console.log('Headers parsed:', Object.keys(headers).length, 'headers found');
  
  // Find boundary if multipart
  const contentType = headers['Content-Type'] || '';
  const boundaryMatch = contentType.match(/boundary="?([^";\r\n]+)"?/i);
  const boundary = boundaryMatch ? boundaryMatch[1] : null;
  
  console.log('Content-Type:', contentType);
  console.log('Boundary detected:', boundary || 'None');
  
  // Parse body parts if multipart
  let textContent = '';
  let htmlContent = '';
  
  if (boundary) {
    console.log('Processing multipart email with boundary:', boundary);
    const bodyParts = body.split(new RegExp(`--${boundary}(?:--)?\r?\n?`));
    console.log('Found', bodyParts.length, 'body parts');
    
    // Process each part
    bodyParts.forEach((part, index) => {
      if (!part.trim()) {
        console.log(`Part ${index}: Empty part, skipping`);
        return;
      }
      
      console.log(`Processing part ${index}...`);
      const partParts = part.split(/\r?\n\r?\n/, 2);
      const partHeaders = parseHeaders(partParts[0]);
      const partContent = partParts.length > 1 ? partParts.slice(1).join('\n\n') : '';
      
      const partContentType = partHeaders['Content-Type'] || '';
      console.log(`Part ${index} Content-Type:`, partContentType);
      
      // Handle nested multipart
      if (partContentType.includes('multipart/')) {
        console.log(`Part ${index}: Found nested multipart content`);
        const nestedBoundaryMatch = partContentType.match(/boundary="?([^";\r\n]+)"?/i);
        if (nestedBoundaryMatch) {
          const nestedBoundary = nestedBoundaryMatch[1];
          console.log(`Part ${index}: Nested boundary:`, nestedBoundary);
          const nestedParts = partContent.split(new RegExp(`--${nestedBoundary}(?:--)?\r?\n?`));
          console.log(`Part ${index}: Found`, nestedParts.length, 'nested parts');
          
          nestedParts.forEach((nestedPart, nestedIndex) => {
            if (!nestedPart.trim()) {
              console.log(`Part ${index}, Nested part ${nestedIndex}: Empty, skipping`);
              return;
            }
            
            console.log(`Processing nested part ${nestedIndex} of part ${index}...`);
            const nestedPartParts = nestedPart.split(/\r?\n\r?\n/, 2);
            const nestedPartHeaders = parseHeaders(nestedPartParts[0]);
            const nestedPartContent = nestedPartParts.length > 1 ? nestedPartParts.slice(1).join('\n\n') : '';
            
            const nestedPartContentType = nestedPartHeaders['Content-Type'] || '';
            console.log(`Nested part ${nestedIndex} Content-Type:`, nestedPartContentType);
            
            if (nestedPartContentType.includes('text/plain')) {
              console.log(`Nested part ${nestedIndex}: Found plain text content`);
              textContent = decodeContent(nestedPartContent, nestedPartHeaders);
              console.log('Plain text content length:', textContent.length);
            } else if (nestedPartContentType.includes('text/html')) {
              console.log(`Nested part ${nestedIndex}: Found HTML content`);
              htmlContent = decodeContent(nestedPartContent, nestedPartHeaders);
              console.log('HTML content length:', htmlContent.length);
            } else {
              console.log(`Nested part ${nestedIndex}: Unknown content type, skipping`);
            }
          });
        } else {
          console.log(`Part ${index}: No nested boundary found in multipart content`);
        }
      } 
      // Handle regular content types
      else if (partContentType.includes('text/plain')) {
        console.log(`Part ${index}: Found plain text content`);
        textContent = decodeContent(partContent, partHeaders);
        console.log('Plain text content length:', textContent.length);
      } else if (partContentType.includes('text/html')) {
        console.log(`Part ${index}: Found HTML content`);
        htmlContent = decodeContent(partContent, partHeaders);
        console.log('HTML content length:', htmlContent.length);
      } else {
        console.log(`Part ${index}: Unknown content type, skipping`);
      }
    });
  } else {
    // Not multipart, treat as plain text
    console.log('Email is not multipart, treating body as plain text');
    textContent = body;
    console.log('Plain text content length:', textContent.length);
  }
  
  console.log('MIME parsing complete');
  return {
    headers,
    textContent,
    htmlContent,
    raw: mimeContent
  };
}

/**
 * Parse email headers into an object
 * @param {string} headerText - Raw header text
 * @returns {Object} Parsed headers
 */
function parseHeaders(headerText) {
  console.log('Parsing headers...');
  const headers = {};
  let currentHeader = '';
  let currentValue = '';
  
  // Split by lines and process
  const lines = headerText.split(/\r?\n/);
  console.log('Header lines to process:', lines.length);
  
  lines.forEach((line, index) => {
    // If line starts with whitespace, it's a continuation
    if (/^\s+/.test(line)) {
      currentValue += ' ' + line.trim();
      console.log(`Line ${index+1}: Continuation of header "${currentHeader}"`);
    } else {
      // Save previous header if exists
      if (currentHeader) {
        headers[currentHeader] = currentValue.trim();
        console.log(`Saved header: ${currentHeader}`);
      }
      
      // Parse new header
      const match = line.match(/^([^:]+):\s*(.*)/);
      if (match) {
        currentHeader = match[1];
        currentValue = match[2];
        console.log(`Line ${index+1}: New header "${currentHeader}"`);
      } else {
        console.log(`Line ${index+1}: Not a valid header line: "${line}"`);
      }
    }
  });
  
  // Save the last header
  if (currentHeader) {
    headers[currentHeader] = currentValue.trim();
    console.log(`Saved final header: ${currentHeader}`);
  }
  
  console.log('Headers parsing complete, found', Object.keys(headers).length, 'headers');
  return headers;
}

/**
 * Decode content based on Content-Transfer-Encoding
 * @param {string} content - Raw content
 * @param {Object} headers - Headers with encoding information
 * @returns {string} Decoded content
 */
function decodeContent(content, headers) {
  const encoding = headers['Content-Transfer-Encoding'] || '';
  console.log('Decoding content with encoding:', encoding || 'none specified');
  
  if (encoding.toLowerCase() === 'quoted-printable') {
    console.log('Decoding quoted-printable content');
    const decoded = decodeQuotedPrintable(content);
    console.log('Decoded content length:', decoded.length);
    return decoded;
  } else if (encoding.toLowerCase() === 'base64') {
    console.log('Decoding base64 content');
    try {
      const decoded = atob(content.replace(/\s+/g, ''));
      console.log('Decoded content length:', decoded.length);
      return decoded;
    } catch (error) {
      console.error('Base64 decoding failed:', error);
      return content;
    }
  }
  
  console.log('No decoding needed, returning raw content');
  return content;
}

/**
 * Decode quoted-printable content
 * @param {string} content - Quoted-printable encoded content
 * @returns {string} Decoded content
 */
function decodeQuotedPrintable(content) {
  console.log('Starting quoted-printable decoding');
  console.log('Input content length:', content.length);
  
  // Remove soft line breaks
  const withoutLineBreaks = content.replace(/=\r?\n/g, '');
  console.log('After removing soft line breaks:', withoutLineBreaks.length);
  
  // Decode hex characters
  const decoded = withoutLineBreaks.replace(/=([0-9A-F]{2})/gi, (_, hex) => {
    const char = String.fromCharCode(parseInt(hex, 16));
    return char;
  });
  
  console.log('After decoding hex characters:', decoded.length);
  return decoded;
}

/**
 * Extract email thread from MIME content
 * @param {string} mimeContent - Raw MIME content
 * @returns {Object} Structured email data
 */
function extractEmailThread(mimeContent) {
  console.log('Extracting email thread from MIME content...');
  const parsed = parseMimeEmail(mimeContent);
  
  // Get sender information
  const from = parsed.headers['From'] || '';
  console.log('From header:', from);
  const fromMatch = from.match(/(.*?)\s*<([^>]+)>/);
  const senderName = fromMatch ? fromMatch[1].trim() : from;
  const senderEmail = fromMatch ? fromMatch[2] : from;
  console.log('Sender parsed:', senderName, senderEmail);
  
  // Get recipient information
  const to = parsed.headers['To'] || '';
  console.log('To header:', to);
  const toMatch = to.match(/(.*?)\s*<([^>]+)>/);
  const recipientName = toMatch ? toMatch[1].trim() : to;
  const recipientEmail = toMatch ? toMatch[2] : to;
  console.log('Recipient parsed:', recipientName, recipientEmail);
  
  // Get subject
  const subject = parsed.headers['Subject'] || '';
  console.log('Subject:', subject);
  
  // Get date
  const date = parsed.headers['Date'] || '';
  console.log('Date:', date);
  
  // Prefer HTML content if available, otherwise use text
  const content = parsed.htmlContent || parsed.textContent;
  console.log('Using content type:', parsed.htmlContent ? 'HTML' : 'Plain text');
  console.log('Content length:', content.length);
  
  console.log('Email thread extraction complete');
  return {
    subject,
    date,
    sender: {
      name: senderName,
      email: senderEmail
    },
    recipient: {
      name: recipientName,
      email: recipientEmail
    },
    content,
    cc: parsed.headers['Cc'] || '',
    bcc: parsed.headers['Bcc'] || '',
    messageId: parsed.headers['Message-ID'] || '',
    inReplyTo: parsed.headers['In-Reply-To'] || '',
    references: parsed.headers['References'] || '',
    headers: parsed.headers,
    raw: parsed.raw
  };
}

// Export functions
if (typeof module !== 'undefined' && module.exports) {
  module.exports = {
    parseMimeEmail,
    extractEmailThread
  };
}