psanford / httpreadat

io.ReadAt wrapper for http range requests in Go

httpreadat: Go http range requests via an io.ReaderAt

httpreadat is a library for making http range requests exposed as a normal io.ReaderAt interface.

This is especially useful for interacting with large files stored in object stores like S3, where you only need to read certain parts of the file.

Example

	rr := New(url)
	rr.ReadAt(buf, offset)

Options

Request manipulation

You may need more control over the request beyond simply setting the url. For those cases you can optionally provide a custom http.RoundTripper. The RoundTripper has full access to the request to make any changes necessary before sending the request out.

Caching

Caching is often useful, especially when used with code that is not optimized for doing reads over the network. httpreadat provides an optional interface CacheHandler that you can implement for your own caching strategy.

The cache handler can make requests that are different than what ReadAt was called with. This allows for things like fetching a larger amount of data if the caller makes lots of small sequential reads.

There is an example CacheHandler in diskcache that fetches pages at a time and caches to a local os.File.

S3

The easiest way to use this with S3 is to make a presigned url for the S3 object and pass that url to New():

	req, _ := s3svc.GetObjectRequest(&s3.GetObjectInput{
		Bucket: aws.String(bucket),
		Key:    aws.String(path),
	})

	url, err := req.Presign(1 * time.Hour)
	if err != nil {
		return err
	}

	r := httpreadat.New(url)

httpreadat's Issues

Would you consider tagging a release?

Hi!

Thanks for this, made porting sqlite3vfshttp to github.com/ncruces/go-sqlite3 a breeze.

Would you be so kind as to tag a release?

Also, did you consider using HEAD here?

httpreadat/httpreadat.go

Lines 82 to 84 in 005e63d

	func (rr *RangeReader) Size() (n int64, err error) {
	req, err := http.NewRequest("GET", rr.url, nil)
	if err != nil {

Thanks again!