Work around for files getting uploaded as application/octet-stream

app-controllers-application_controller.rb

class ApplicationController < ActionController::Base
# omitted stuff
private
  # Used to correct content type for files uploaded with bulkupload or from PicMonkey
  def correct_content_type_for_octet_stream(filedata)
    return nil if filedata.blank?
    # see what the unix file command thinks this is
    if filedata.content_type == 'application/octet-stream'
      filedata.content_type = type_from_file_command(filedata.path)
    end
    # 'file' thinks new Office docs are zip files, so we may need to correct what we just did above
    if filedata.content_type =~ /application\/(x-)?zip/
      filedata.content_type = type_from_file_extension(filedata.original_filename)
    end
    filedata
  end

  def type_from_file_command(file)
    Paperclip::FileCommandContentTypeDetector.new(file).detect
  end

  def type_from_file_extension(filename)
    Paperclip::ContentTypeDetector.new(filename).send(:possible_types).first
  end
end

app-controllers-documents_controller.rb

class DocumentsController < ApplicationController
  before_action :set_document, only: [:edit, :update, :destroy]
# .... omitted

  # POST /documents
  def create
    @document = Document.new(document_params)
    @document.file = correct_content_type_for_octet_stream(params[:document][:file])
    if @document.save
      redirect_to document_category_path(id: @document.document_category_id), notice: 'Document was successfully created.'
    else
      render :new 
    end
  end
  
  # PUT /documents/1
  def update
    @document.file = correct_content_type_for_octet_stream(params[:document][:file]) unless params[:document][:file].blank?
    if @document.update_attributes(document_params)
      redirect_to document_path(@document), notice: 'Document was successfully updated.' 
    else
      render :edit
    end
  end
  
  private
  # Use callbacks to share common setup or constraints between actions.
  def set_document
    @document = Document.find(params[:id])
  end

  # Only allow a trusted parameter "white list" through.
  def document_params
    params.require(:document).permit(:document_category_id, :campus_only, :login_only)
  end
end

app-models-document.rb

class Document < ActiveRecord::Base
  belongs_to :category, :class_name => 'DocumentCategory', :foreign_key => 'document_category_id'
  
  # Store documents in S3 - one bucket for all CMS instances - subfolders for each site
  has_attached_file :file,
                    :storage => :s3, 
                    :s3_credentials => "#{Rails.root}/config/amazon_s3.yml",
                    :s3_permissions => "authenticated-read",
                    :path => "#{SiteConfig.hostname}/:class/:id_partition/:filename",
                    :filename_cleaner => PaperclipModifications::FilenameCleaner.new

  validates_attachment :file, :presence => true, :size => { :less_than => 100.megabytes }, 
      :content_type => { :content_type => [/\Aimage\/.*\Z/, /\Aapplication\/.*\Z/, /\Atext\/.*\Z/]}
end

config-initializers-paperclip.rb

# Allow type mismatch for Office documents - our version of `file` calls all of them zip files
if Object::RUBY_PLATFORM =~ /darwin/
  Paperclip.options[:content_type_mappings] = {
    :docx => "application/zip",
    :xlsx => "application/zip",
    :pptx => "application/zip",
    :pages => "application/zip",
    :numbers => "application/zip",
    :key => "application/zip"
  }
else
  Paperclip.options[:content_type_mappings] = {
    :docx => "application/x-zip",
    :xlsx => "application/x-zip",
    :pptx => "application/x-zip",
    :pages => "application/x-zip",
    :numbers => "application/x-zip",
    :key => "application/x-zip"
  }
end

module Paperclip
  class MediaTypeSpoofDetector
    # Make spoof detection stricter - original only compared type so all 'application' types matched all others
    def media_type_mismatch?
      ! supplied_file_content_types.include?(calculated_content_type)
    end
  end
end