Paperclip- validate pdfs with content_type='application/octet-stream'
I was using paperclip
for file upload. with validations as below:
validates_attachment_content_type :upload, :content_type=>['application/pdf'],
:if => Proc.new { |module_file| !module_file.upload_file_name.blank? },
:message => "must be in '.pdf' format"
But, my client complained today that he is not able to upload pdf
. After investigating I come to know from request headers is that the file being submitted had content_type=application/octet-stream
.
Allowing application/octet-stream
will allow many type of files for upload.
Please suggest a开发者_JAVA百科 solution to deal with this.
Seems like paperclip doesn't detect content type correctly. Here is how I was able to fix it using custom content-type detection and validation (code in model):
VALID_CONTENT_TYPES = ["application/zip", "application/x-zip", "application/x-zip-compressed", "application/pdf", "application/x-pdf"]
before_validation(:on => :create) do |file|
if file.media_content_type == 'application/octet-stream'
mime_type = MIME::Types.type_for(file.media_file_name)
file.media_content_type = mime_type.first.content_type if mime_type.first
end
end
validate :attachment_content_type
def attachment_content_type
errors.add(:media, "type is not allowed") unless VALID_CONTENT_TYPES.include?(self.media_content_type)
end
Based on the above, here's what I ended up with which is compatible with PaperClip 4.2 and Rails 4:
before_post_process on: :create do
if media_content_type == 'application/octet-stream'
mime_type = MIME::Types.type_for(media_file_name)
self.media_content_type = mime_type.first.to_s if mime_type.first
end
end
For paperclip 3.3 and Rails 3, I did this a bit differently
before_validation on: :create do
if media_content_type == 'application/octet-stream'
mime_type = MIME::Types.type_for(media_file_name)
self.media_content_type = mime_type.first if mime_type.first
end
end
validates_attachment :media, content_type: { content_type: VALID_CONTENT_TYPES }
By the way, i needed to do this because testing with Capybara and phantom js using attach_file did not generate the correct mime type for some files.
精彩评论