Extract content types from blob data

4 files changed, 64 insertions, 5 deletions

diff --git a/activestorage/app/models/active_storage/attachment.rb b/activestorage/app/models/active_storage/attachment.rb
index 9f61a5dbf3..19f48c57d6 100644
--- a/activestorage/app/models/active_storage/attachment.rb
+++ b/activestorage/app/models/active_storage/attachment.rb
@@ -14,7 +14,7 @@ class ActiveStorage::Attachment < ActiveRecord::Base
 
   delegate_missing_to :blob
 
-  after_create_commit :analyze_blob_later
+  after_create_commit :identify_blob, :analyze_blob_later
 
   # Synchronously purges the blob (deletes it from the configured service) and destroys the attachment.
   def purge
@@ -29,6 +29,10 @@ class ActiveStorage::Attachment < ActiveRecord::Base
   end
 
   private
+    def identify_blob
+      blob.identify
+    end
+
     def analyze_blob_later
       blob.analyze_later unless blob.analyzed?
     end
diff --git a/activestorage/app/models/active_storage/blob.rb b/activestorage/app/models/active_storage/blob.rb
index 2ac09d3062..7067c58259 100644
--- a/activestorage/app/models/active_storage/blob.rb
+++ b/activestorage/app/models/active_storage/blob.rb
@@ -14,12 +14,12 @@
 # update a blob's metadata on a subsequent pass, but you should not update the key or change the uploaded file.
 # If you need to create a derivative or otherwise change the blob, simply create a new blob and purge the old one.
 class ActiveStorage::Blob < ActiveRecord::Base
-  include Analyzable, Representable
+  include Analyzable, Identifiable, Representable
 
   self.table_name = "active_storage_blobs"
 
   has_secure_token :key
-  store :metadata, accessors: [ :analyzed ], coder: JSON
+  store :metadata, accessors: [ :analyzed, :identified ], coder: JSON
 
   class_attribute :service
 
@@ -136,8 +136,10 @@ class ActiveStorage::Blob < ActiveRecord::Base
   # Normally, you do not have to call this method directly at all. Use the factory class methods of +build_after_upload+
   # and +create_after_upload!+.
   def upload(io)
-    self.checksum  = compute_checksum_in_chunks(io)
-    self.byte_size = io.size
+    self.checksum     = compute_checksum_in_chunks(io)
+    self.content_type = extract_content_type(io)
+    self.byte_size    = io.size
+    self.identified   = true
 
     service.upload(key, io, checksum: checksum)
   end
@@ -182,6 +184,10 @@ class ActiveStorage::Blob < ActiveRecord::Base
       end.base64digest
     end
 
+    def extract_content_type(io)
+      Marcel::MimeType.for io, name: filename.to_s, declared_type: content_type
+    end
+
     def forcibly_serve_as_binary?
       ActiveStorage.content_types_to_serve_as_binary.include?(content_type)
     end
diff --git a/activestorage/app/models/active_storage/blob/identifiable.rb b/activestorage/app/models/active_storage/blob/identifiable.rb
new file mode 100644
index 0000000000..40ca84ac70
--- /dev/null
+++ b/activestorage/app/models/active_storage/blob/identifiable.rb
@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module ActiveStorage::Blob::Identifiable
+  def identify
+    ActiveStorage::Identification.new(self).apply
+  end
+
+  def identified?
+    identified
+  end
+end
diff --git a/activestorage/app/models/active_storage/identification.rb b/activestorage/app/models/active_storage/identification.rb
new file mode 100644
index 0000000000..4f295257ae
--- /dev/null
+++ b/activestorage/app/models/active_storage/identification.rb
@@ -0,0 +1,38 @@
+# frozen_string_literal: true
+
+class ActiveStorage::Identification
+  attr_reader :blob
+
+  def initialize(blob)
+    @blob = blob
+  end
+
+  def apply
+    blob.update!(content_type: content_type, identified: true) unless blob.identified?
+  end
+
+  private
+    def content_type
+      Marcel::MimeType.for(identifiable_chunk, name: filename, declared_type: declared_content_type)
+    end
+
+
+    def identifiable_chunk
+      Net::HTTP.start(uri.host, uri.port, use_ssl: uri.scheme == "https") do |client|
+        client.get(uri, "Range" => "0-4096").body
+      end
+    end
+
+    def uri
+      @uri ||= URI.parse(blob.service_url)
+    end
+
+
+    def filename
+      blob.filename.to_s
+    end
+
+    def declared_content_type
+      blob.content_type
+    end
+end