Add option to ActiveStorage::Blob to set extract_content_type_from_io #32833

ryanwhocodes · May 6, 2018

activestorage CHANGELOG.md addition

Use extract_content_type_from_io: false as an argument in
ActiveStorage::Blob.create_after_upload to manually override the
automatic extraction of content type from the io.

This address the issue where the user defines a content type as an argument
but the automatic MimeType inference from the io is not what is expected.
It means a user avoids needing force the correct content type by
patching ActiveStorage::Blob.extract_content_type.

Ryan Davidson

Summary

This adds a boolean extract_content_type_from_io to ActiveStorage::Blob
methods #create_after_upload, #build_after_upload and #upload. It
allows a user to manually override the automatic MimeType extraction from
the io. This provides a patch for when Marcel::MimeType.for returns an
incorrect content type when #extract_content_type is called from the #upload method.

This patch aims to address:

#32632 Provide a way to deactivate ActiveStorage MIME type inference

It fixes bugs where a content_type is automatically extracted from the input and gives an inaccurate result. This is due to a bug in the Marcel::MimeType gem that is called in the extract_content_type method (which was called by default in the #upload method, which is in turn called by #build_after_upload and #create_after_upload)

2.5.0 Marcel::MimeType.for(StringIO.new("1895 1"), name: 'text.csv', declared_type: 'text/csv')
=> "text/csv"
2.5.0 Marcel::MimeType.for(StringIO.new("Article dates analysis"), name: 'text.csv', declared_type: 'text/csv')
=> "message/news"
2.5.0 Marcel::MimeType.for(StringIO.new("dates analysis"), name: 'text.csv', declared_type: 'text/csv')
=> "text/csv"
2.5.0 Marcel::MimeType.for(StringIO.new("Dates analysis"), name: 'text.csv', declared_type: 'text/csv')
=> "text/csv"
2.5.0 Marcel::MimeType.for(StringIO.new("article"), name: 'text.csv', declared_type: 'text/csv')
=> "text/csv"
2.5.0 Marcel::MimeType.for(StringIO.new("Article"), name: 'text.csv', declared_type: 'text/csv')
=> "message/news"

Other Information

Added document comments to activestorage/app/models/active_storage/blob.rb

Finally, if your pull request affects documentation or any non-code
changes, guidelines for those changes are available
here

Thanks for contributing to Rails!

rails-bot · May 6, 2018

Thanks for the pull request, and welcome! The Rails team is excited to review your changes, and you should hear from @schneems (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

This repository is being automatically checked for code quality issues using Code Climate. You can see results for this analysis in the PR status below. Newly introduced issues should be fixed before a Pull Request is considered ready to review.

Please see the contribution instructions for more information.

georgeclaghorn · May 8, 2018

georgeclaghorn requested changes May 8, 2018

View changes

georgeclaghorn · May 8, 2018

Sign in to view

-    def build_after_upload(io:, filename:, content_type: nil, metadata: nil)
+    # Set the +content_type+ and +extract_content_type_from_io+ to +false+ to manually
+    # specify the content type rather than using automatic content type inference.
+    def build_after_upload(io:, filename:, content_type: nil, metadata: nil, extract_content_type_from_io: true)


I’d like to name this argument identify here and in create_after_upload for consistency with the identify method.

Renamed to identify in this commit 4f9defb42cf7c543906a7de72c7b6f91fa5058d4

georgeclaghorn · May 8, 2018

Sign in to view

  #
  # 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)
+  def upload(io, extract_content_type_from_io = true)


Let’s use a keyword argument here. Name it identify to match build_after_upload and create_after_upload.

Making it a keyword argument makes it much clearer. Amended here e77f3db21842d20a489853b41eed4233c5394666

georgeclaghorn · May 8, 2018

Sign in to view

+  test "create after upload extracts content_type when extract_content_type_from_io: true" do
+    blob = create_csv_blob data: "Article,dates,analysis\n1, 2, 3", content_type: "text/csv", extract_content_type_from_io: true
+    assert_equal "message/news", blob.content_type
+  end


The new option defaults to true, so this test is superfluous. The preexisting tests above cover this behavior.

Removed this test in this commit 539e880fe4731ef430a857a35ec9efc167edf263

georgeclaghorn · May 8, 2018

Sign in to view

@@ -54,6 +54,10 @@ def create_blob(data: "Hello world!", filename: "hello.txt", content_type: "text
      ActiveStorage::Blob.create_after_upload! io: StringIO.new(data), filename: filename, content_type: content_type
    end

+    def create_csv_blob(data: "article,dates,analysis\n1, 2, 3", filename: "my_csv.csv", content_type: "text/csv", extract_content_type_from_io: true)
+      ActiveStorage::Blob.create_after_upload! io: StringIO.new(data), filename: filename, content_type: content_type, extract_content_type_from_io: extract_content_type_from_io
+    end


We don’t need a special helper for creating a CSV blob. Modify create_blob to take an identify argument, then call it from the new test:

test "create after upload uses content_type when extract_content_type_from_io: false" do blob = create_blob data: "Article,dates,analysis\n1,2,3", filename: "table.csv", content_type: "text/csv", identify: false assert_equal "text/csv", blob.content_type end

Done in this commit 539e880fe4731ef430a857a35ec9efc167edf263

Just remembered to include the filename arg for this updated test d6e1197edd9e626172de6dbf4c0380088c6e28f7

georgeclaghorn · May 8, 2018

Sign in to view

+  test "create after upload extracts content_type from io when no content_type given" do
+    blob = create_csv_blob content_type: nil
+    assert_equal "text/csv", blob.content_type
+  end


This test doesn’t seem terribly valuable. I’d omit it or modify it to cover the case where content_type is nil and identify is false.

Modified for content_type nil and identify: false e34eec2c2c2133edf4512b1370455fe4c669c0ca

ryanwhocodes · May 8, 2018

@georgeclaghorn This PR is now updated with the requests in your review. I felt like I learnt a bit more about Rails architecture and how to write better Ruby from this activity!

georgeclaghorn · May 8, 2018

georgeclaghorn requested changes May 8, 2018

View changes

georgeclaghorn · May 8, 2018

Sign in to view

+    This address the issue where the user defines a content type as an argument
+    but the automatic MimeType inference from the io is not what is expected.
+    It means a user avoids needing force the correct content type by
+    patching `ActiveStorage::Blob.extract_content_type`.


It’s uncommon to call ActiveStorage::Blob#create_after_upload or #build_after_upload directly. Can we demonstrate passing identify: false to ActiveStorage::Attached::{One,Many}#attach instead?

I've added the update but not sure if the content is what you had in mind. It has a similar wording to the comments in blob.rb and I've added in an example based on what is in the RailsGuides.

commit 4152b834102b8dbd9f6cae9965d4d00bc1b3f512

georgeclaghorn · May 8, 2018

Sign in to view

    self.checksum     = compute_checksum_in_chunks(io)
-    self.content_type = extract_content_type(io)
+    self.content_type = extract_content_type(io) if self.content_type.nil? || identify


content_type doesn’t need an explicit receiver in the conditional. Replace self.content_type.nil? with content_type.nil?.

Removed reference to content_type explicit receiver here 606f42457796258f9f94622678f5d4aac10b6e29

georgeclaghorn · May 8, 2018

Sign in to view

@@ -44,21 +44,25 @@ def find_signed(id)
    end

    # Returns a new, unsaved blob instance after the +io+ has been uploaded to the service.
-    def build_after_upload(io:, filename:, content_type: nil, metadata: nil)
+    # Set the +content_type+ and +identify+ to +false+ to manually


The first part of this sentence makes it sound like both content_type and identify should be nil. Here’s a clarifying edit:

When providing a content type, pass identify: false to bypass automatic content type inference.

Clarified this sentence with the edit bad8fa0

Added in the formatting of identify: false here 231ad55

georgeclaghorn · May 8, 2018

Sign in to view

@@ -142,13 +146,14 @@ def service_headers_for_direct_upload
  #
  # Prior to uploading, we compute the checksum, which is sent to the service for transit integrity validation. If the
  # checksum does not match what the service receives, an exception will be raised. We also measure the size of the +io+
-  # and store that in +byte_size+ on the blob record.
+  # and store that in +byte_size+ on the blob record. The content type is automatically extracted from the +io+ unless
+  # the user has specified a +content_type+ and set the +identify+ as false.


Remove the extra “the” between “set” and “identify.”

Removed extra 'the' here
c79c9e1

georgeclaghorn · May 8, 2018

georgeclaghorn reviewed May 8, 2018

View changes

georgeclaghorn · May 8, 2018

Sign in to view

@@ -44,21 +44,24 @@ def find_signed(id)
    end

    # Returns a new, unsaved blob instance after the +io+ has been uploaded to the service.
-    def build_after_upload(io:, filename:, content_type: nil, metadata: nil)
+    # When providing a content type, pass +identify: false+ to bypass automatic content type inference.


+ only works around individual words. Use <tt></tt> instead:

pass <tt>identify: false</tt>

Ah right, updated 209f021c15373832726496c157b86223ceaa291b

georgeclaghorn · May 8, 2018

georgeclaghorn reviewed May 8, 2018

View changes

georgeclaghorn · May 8, 2018

Sign in to view

-    def create_after_upload!(io:, filename:, content_type: nil, metadata: nil)
-      build_after_upload(io: io, filename: filename, content_type: content_type, metadata: metadata).tap(&:save!)
+    # Set the +content_type+ and +identify+ to +false+ to manually
+    # specify the content type rather than using automatic content type inference.


This needs the same treatment as build_after_upload.

create_after_upload! now has the same comment as build_after_upload
e5f7519

georgeclaghorn · May 8, 2018

georgeclaghorn reviewed May 8, 2018

View changes

georgeclaghorn · May 8, 2018

Sign in to view

@@ -142,13 +145,14 @@ def service_headers_for_direct_upload
  #
  # Prior to uploading, we compute the checksum, which is sent to the service for transit integrity validation. If the
  # checksum does not match what the service receives, an exception will be raised. We also measure the size of the +io+
-  # and store that in +byte_size+ on the blob record.
+  # and store that in +byte_size+ on the blob record. The content type is automatically extracted from the +io+ unless
+  # the user has specified a +content_type+ and set +identify+ as false.


Use “you specify” instead of “the user has specified” to match the adjacent paragraphs, which use the second person.

“Set identify as false” doesn’t read quite right. How about “pass identify: false”?

Yes, that read more fluently and consistently with the rest of the test
190693c

ryanwhocodes · May 8, 2018

@georgeclaghorn

Looks like this change would affect the Rails Guides too:

http://edgeguides.rubyonrails.org/active_storage_overview.html#attaching-file-io-objects in the section 3.3 Attaching File/IO Objects

I've added this to the commit and squashed it in.

georgeclaghorn · May 8, 2018

Can you please squash your commits into one?

ryanwhocodes · May 8, 2018

Commits are now squashed and pushed up

georgeclaghorn · May 8, 2018

georgeclaghorn approved these changes May 8, 2018

View changes

rails-bot assigned schneems May 6, 2018

This was referenced May 6, 2018

Closed

Add option to ActiveStorage::Blob to set extract_content_type_from_io #32832

Closed

Provide a way to deactivate ActiveStorage MIME type inference #32632

georgeclaghorn added the activestorage label May 7, 2018

georgeclaghorn assigned georgeclaghorn and unassigned schneems May 8, 2018

georgeclaghorn merged commit 38ea7bb into rails:master May 8, 2018
2 checks passed

2 checks passed

codeclimate All good!
Details

continuous-integration/travis-ci/pr The Travis CI build passed
Details

ryanwhocodes deleted the ryanwhocodes:activestorage_blob_set_content_type branch May 9, 2018

kevindew referenced this pull request Aug 23, 2018
Closed
[DO NOT MERGE] Basic image flow spike #170

rails/rails

Join GitHub today

Add option to ActiveStorage::Blob to set extract_content_type_from_io #32833

Conversation

ryanwhocodes commented May 6, 2018 • edited

activestorage CHANGELOG.md addition

Summary

Other Information

rails-bot assigned schneems May 6, 2018

This comment has been minimized.

rails-bot commented May 6, 2018

This was referenced May 6, 2018

georgeclaghorn added the activestorage label May 7, 2018

georgeclaghorn requested changes May 8, 2018 View changes

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

ryanwhocodes May 8, 2018 • edited

This comment has been minimized.

georgeclaghorn May 8, 2018 • edited

This comment has been minimized.

georgeclaghorn assigned georgeclaghorn and unassigned schneems May 8, 2018

This comment has been minimized.

ryanwhocodes commented May 8, 2018

georgeclaghorn requested changes May 8, 2018 View changes

This comment has been minimized.

This comment has been minimized.

ryanwhocodes May 8, 2018 • edited

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

georgeclaghorn May 8, 2018 • edited

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

georgeclaghorn reviewed May 8, 2018 View changes

This comment has been minimized.

This comment has been minimized.

ryanwhocodes May 8, 2018 • edited

georgeclaghorn reviewed May 8, 2018 View changes

This comment has been minimized.

This comment has been minimized.

georgeclaghorn reviewed May 8, 2018 View changes

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

ryanwhocodes commented May 8, 2018 • edited

This comment has been minimized.

georgeclaghorn commented May 8, 2018

This comment has been minimized.

ryanwhocodes commented May 8, 2018

georgeclaghorn approved these changes May 8, 2018 View changes

Hide details View details georgeclaghorn merged commit 38ea7bb into rails:master May 8, 2018 2 checks passed

2 checks passed

ryanwhocodes deleted the ryanwhocodes:activestorage_blob_set_content_type branch May 9, 2018

kevindew referenced this pull request Aug 23, 2018

ryanwhocodes commented May 6, 2018 •

edited

georgeclaghorn requested changes May 8, 2018

View changes

ryanwhocodes May 8, 2018 •

edited

georgeclaghorn May 8, 2018 •

edited

georgeclaghorn requested changes May 8, 2018

View changes

ryanwhocodes May 8, 2018 •

edited

georgeclaghorn May 8, 2018 •

edited

georgeclaghorn reviewed May 8, 2018

View changes

ryanwhocodes May 8, 2018 •

edited

georgeclaghorn reviewed May 8, 2018

View changes

georgeclaghorn reviewed May 8, 2018

View changes

ryanwhocodes commented May 8, 2018 •

edited

georgeclaghorn approved these changes May 8, 2018

View changes

georgeclaghorn merged commit `38ea7bb` into rails:master May 8, 2018
2 checks passed