Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit

I'll never remember to update them, people should just look at the last
CI run.

Git stats


Failed to load latest commit information.
Latest commit message
Commit time
December 26, 2022 20:42
September 25, 2015 23:40
December 26, 2022 21:49
September 25, 2015 23:40
September 25, 2015 23:40
December 26, 2022 20:42


Down is a utility tool for streaming, flexible and safe downloading of remote files. It can use open-uri + Net::HTTP, http.rb, HTTPX, or wget as the backend HTTP library.


gem "down", "~> 5.0"


The primary method is, which downloads the remote file into a Tempfile:

require "down"

tempfile ="")
tempfile #=> #<Tempfile:/var/folders/k7/6zx6dx6x7ys3rv3srh0nyfj00000gn/T/20150925-55456-z7vxqz.jpg>


The returned Tempfile has some additional attributes extracted from the response data:

tempfile.content_type      #=> "text/plain"
tempfile.original_filename #=> "document.txt"
tempfile.charset           #=> "utf-8"

Maximum size

When you're accepting URLs from an outside source, it's a good idea to limit the filesize (because attackers want to give a lot of work to your servers). Down allows you to pass a :max_size option:"", max_size: 5 * 1024 * 1024) # 5 MB
# Down::TooLarge: file is too large (max is 5MB)

What is the advantage over simply checking size after downloading? Well, Down terminates the download very early, as soon as it gets the Content-Length header. And if the Content-Length header is missing, Down will terminate the download as soon as the downloaded content surpasses the maximum size.


By default the remote file will be downloaded into a temporary location and returned as a Tempfile. If you would like the file to be downloaded to a specific location on disk, you can specify the :destination option:"", destination: "/path/to/destination")
#=> nil

In this case won't have any return value, so if you need a File object you'll have to create it manually.

You can also keep the tempfile, but override the extension:

tempfile ="", extension: "txt")
File.extname(tempfile.path) #=> ".txt"

Basic authentication and will automatically detect and apply HTTP basic authentication from the URL:"")"")

Progress supports :content_length_proc, which gets called with the value of the Content-Length header as soon as it's received, and :progress_proc, which gets called with current filesize whenever a new chunk is downloaded. "",
  content_length_proc: -> (content_length) { ... },
  progress_proc:       -> (progress)       { ... }


Down has the ability to retrieve content of the remote file as it is being downloaded. The method returns a Down::ChunkedIO object which represents the remote file on the given URL. When you read from it, Down internally downloads chunks of the remote file, but only how much is needed.

remote_file ="")
remote_file.size # read from the "Content-Length" header # downloads and returns first 1 KB # downloads and returns next 1 KB

remote_file.eof? #=> false # downloads and returns the rest of the file content
remote_file.eof? #=> true

remote_file.close # closes the HTTP connection and deletes the internal Tempfile

The following IO methods are implemented:

  • #read & #readpartial
  • #gets
  • #seek
  • #pos & #tell
  • #eof?
  • #rewind
  • #close


By default the downloaded content is internally cached into a Tempfile, so that when you rewind the Down::ChunkedIO, it continues reading the cached content that it had already retrieved.

remote_file ="")*1024*1024) # downloads, caches, and returns first 1MB
remote_file.rewind*1024*1024) # reads the cached content*1024*1024) # downloads the next 1MB

If you want to save on IO calls and on disk usage, and don't need to be able to rewind the Down::ChunkedIO, you can disable caching downloaded content:"", rewindable: false)

Yielding chunks

You can also yield chunks directly as they're downloaded via #each_chunk, in which case the downloaded content is not cached into a file regardless of the :rewindable option.

remote_file ="")
remote_file.each_chunk { |chunk| ... }


You can access the response status and headers of the HTTP request that was made:

remote_file ="")[:status]   #=> 200[:headers]  #=> { "Content-Type" => "image/jpeg", ... } (header names are normalized)[:response] # returns the response object

Note that a Down::ResponseError exception will automatically be raised if response status was 4xx or 5xx.


The performs HTTP logic and returns an instance of Down::ChunkedIO. However, Down::ChunkedIO is a generic class that can wrap any kind of streaming. It accepts an Enumerator that yields chunks of content, and provides IO-like interface over that enumerator, calling it whenever more content is needed.

require "down/chunked_io"
  • :chunksEnumerator that yields chunks of content
  • :size – size of the file if it's known (returned by #size)
  • :on_close – called when streaming finishes or IO is closed
  • :data - custom data that you want to store (returned by #data)
  • :rewindable - whether to cache retrieved data into a file (defaults to true)
  • :encoding - force content to be returned in specified encoding (defaults to Encoding::BINARY)

Here is an example of creating a streaming IO of a MongoDB GridFS file:

require "down/chunked_io"

mongo =
bucket = mongo.database.fs

content_length = bucket.find(_id: id).first[:length]
stream = bucket.open_download_stream(id)

io =
  size: content_length,
  chunks: stream.enum_for(:each),
  on_close: -> { stream.close },


Down tries to recognize various types of exceptions and re-raise them as one of the Down::Error subclasses. This is Down's exception hierarchy:

  • Down::Error
    • Down::TooLarge
    • Down::InvalidUrl
    • Down::TooManyRedirects
    • Down::NotModified
    • Down::ResponseError
      • Down::ClientError
        • Down::NotFound
      • Down::ServerError
    • Down::ConnectionError
    • Down::TimeoutError
    • Down::SSLError


The following backends are available:

You can use the backend directly:

require "down/net_http""...")"...")

Or you can set the backend globally (default is :net_http):

require "down"

Down.backend :http # use the Down::Http backend"...")"...")


The Down::NetHttp backend implements downloads using open-uri and Net::HTTP standard libraries.

gem "down", "~> 5.0"
require "down/net_http"

tempfile ="")
tempfile #=> #<Tempfile:/var/folders/k7/6zx6dx6x7ys3rv3srh0nyfj00000gn/T/20150925-55456-z7vxqz.jpg>

io ="")
io #=> #<Down::ChunkedIO ...> is implemented as a wrapper around open-uri, and fixes some of open-uri's undesired behaviours:

  • uses URI::HTTP#open or URI::HTTPS#open directly for security
  • always returns a Tempfile object, whereas open-uri returns StringIO when file is smaller than 10KB
  • gives the extension to the Tempfile object from the URL
  • allows you to limit maximum number of redirects

On the other hand is implemented using Net::HTTP directly, as open-uri doesn't support downloading on-demand.


Down::NetHttp#download turns off open-uri's following redirects, as open-uri doesn't have a way to limit the maximum number of hops, and implements its own. By default maximum of 2 redirects will be followed, but you can change it via the :max_redirects option:"")                   # 2 redirects allowed"", max_redirects: 5) # 5 redirects allowed"", max_redirects: 0) # 0 redirects allowed"")                       # 2 redirects allowed"", max_redirects: 5)     # 5 redirects allowed"", max_redirects: 0)     # 0 redirects allowed


An HTTP proxy can be specified via the :proxy option:"", proxy: "")"", proxy: "")


Timeouts can be configured via the :open_timeout and :read_timeout options:"", open_timeout: 5)"", read_timeout: 10)


Request headers can be added via the :headers option:"", headers: { "Header" => "Value" })"", headers: { "Header" => "Value" })

SSL options

The :ssl_ca_cert and :ssl_verify_mode options are supported, and they have the same semantics as in open-uri:"",
  ssl_ca_cert:     "/path/to/cert",
  ssl_verify_mode: OpenSSL::SSL::VERIFY_PEER)

URI normalization

If the URL isn't parseable by URI.parse, Down::NetHttp will attempt to normalize the URL using Addressable::URI, URI-escaping any potentially unescaped characters. You can change the normalizer via the :uri_normalizer option:

# this skips URL normalization"", uri_normalizer: -> (url) { url })

Additional options

Any additional options passed to will be forwarded to open-uri, so you can for example add basic authentication or a timeout: "",
  http_basic_authentication: ['john', 'secret'],
  read_timeout: 5

You can also initialize the backend with default options:

net_http = 3)"")"")


The Down::Http backend implements downloads using the http.rb gem.

gem "down", "~> 5.0"
gem "http", "~> 5.0"
require "down/http"

tempfile ="")
tempfile #=> #<Tempfile:/var/folders/k7/6zx6dx6x7ys3rv3srh0nyfj00000gn/T/20150925-55456-z7vxqz.jpg>

io ="")
io #=> #<Down::ChunkedIO ...>

Some features that give the http.rb backend an advantage over open-uri and Net::HTTP include:

  • Low memory usage (10x less than open-uri/Net::HTTP)
  • Proper SSL support
  • Support for persistent connections
  • Global timeouts (limiting how long the whole request can take)
  • Chainable builder API for setting default options

Additional options

All additional options will be forwarded to HTTP::Client#request:"", headers: { "Foo" => "Bar" })"", follow: { max_hops: 0 })

However, it's recommended to configure request options using http.rb's chainable API, as it's more convenient than passing raw options."") do |client|
  client.timeout(connect: 3, read: 3)

You can also initialize the backend with default options:

http = { "Foo" => "Bar" })
# or
http = { |client| client.timeout(connect: 3) }"")"")

Request method

By default Down::Http makes a GET request to the specified endpoint, but you can specify a different request method using the :method option:"", method: :post)"", method: :post)

down = :post)"")


The Down::Httpx backend implements downloads using the HTTPX gem, which supports the HTTP/2 protocol, in addition to many other features.

gem "down", "~> 5.0"
gem "httpx", "~> 0.22"
require "down/httpx"

tempfile ="")
tempfile #=> #<Tempfile:/var/folders/k7/6zx6dx6x7ys3rv3srh0nyfj00000gn/T/20150925-55456-z7vxqz.jpg>

io ="")
io #=> #<Down::ChunkedIO ...>

It's implemented in much of the same way as Down::Http, so be sure to check its docs for ways to pass additional options.

Down::Wget (experimental)

The Down::Wget backend implements downloads using the wget command line utility.

gem "down", "~> 5.0"
gem "posix-spawn" # omit if on JRuby
gem "http_parser.rb"
require "down/wget"

tempfile ="")
tempfile #=> #<Tempfile:/var/folders/k7/6zx6dx6x7ys3rv3srh0nyfj00000gn/T/20150925-55456-z7vxqz.jpg>

io ="")
io #=> #<Down::ChunkedIO ...>

One major advantage of wget is that it automatically resumes downloads that were interrupted due to network failures, which is very useful when you're downloading large files.

However, the Wget backend should still be considered experimental, as it wasn't easy to implement a CLI wrapper that streams output, so it's possible that I've made mistakes. Let me know how it's working out for you 😉.

Additional arguments

You can pass additional arguments to the underlying wget commmand via symbols:"", :no_proxy, connect_timeout: 3)"", user: "janko", password: "secret")

You can also initialize the backend with default arguments:

wget =, connect_timeout: 3)"")"")


Tests require that a httpbin server is running locally, which you can do via Docker:

$ docker pull kennethreitz/httpbin
$ docker run -p 80:80 kennethreitz/httpbin

Then you can run tests:

$ bundle exec rake test