Convert .msg into .html when uploaded

**Jira Ticket**: ALCS-2199

**Description**: As ALC staff, I want to drag an email from Outlook into ALCS document uploader, then I'll fill out the required fields, and click save. Then I want ALCS to convert that Outlook email (.msg on Windows) into HTML so that when the document appears in the Documents table, it has the file format .html. I dont want to see .msg once the file has been uploaded. This will make saving 50% of emails to ALCS super quick.

The other 50% of emails will contain attachments and need to be saved to PDF and manipulated within a PDF editor. Improvement, not perfection.

*Acceptance Criteria*
- [ ] user can drag and drop an outlook email item to the upload section, then ALCS will detect the format of .msg and then automatically convert it to .html file on save
- [ ] the first field (document upload) has .msg extension but the second field (document name) shows .html extension
- [ ] show user a message that the file will be converted into .html
- [ ] if the user clears the document (clicks X REMOVE) then disappear the message
- [ ] put loading spinner while the conversion occurs, after virus scan
- [ ] if the .msg file contains an attachment, show a warning that attachments will not be saved and require the user to 'Dismiss' the warning before saving
- [ ] Key info that needs to be retained
- [ ] Date/Time Sent
- [ ] Sender
- [ ] Recipient
- [ ] Subject Line
- [ ] Body
- [ ] File name of any attachments

Figma: https://www.figma.com/design/5qXmbgyOL5M8PO7m7oqCeR/ALCS?node-id=5232-132315&t=pDkYFdDqFabYAzJI-4

**Last Updated**: 2025-02-19 10:17

**Additional Context**
- look into using a an external library, try to avoid making something brand new

Start at the document upload and save flow, then trace how virus scanning, file names, and upload status are handled. Review the linked Figma and investigate an external library for extracting the required Outlook message fields; done means .msg uploads are converted to .html with the specified warnings, spinner, and retained metadata.

Written by the indexing model from the issue text.