gunzip and untar passthrough writers (oras-project#191)

Signed-off-by: Avi Deitcher <[email protected]>

pkg/content/consts.go

@@ -26,3 +26,9 @@ const (
 	// Reference: https://github.com/opencontainers/image-spec/blob/master/image-layout.md#indexjson-file
 	OCIImageIndexFile = "index.json"
 )
+
+const (
+	// DefaultBlocksize default size of each slice of bytes read in each write through in gunzipand untar.
+	// Simply uses the same size as io.Copy()
+	DefaultBlocksize = 32768
+)

pkg/content/gunzip.go

@@ -0,0 +1,48 @@
+package content
+
+import (
+	"compress/gzip"
+	"fmt"
+	"io"
+
+	"github.com/containerd/containerd/content"
+)
+
+// NewGunzipWriter wrap a writer with a gunzip, so that the stream is gunzipped
+func NewGunzipWriter(writer content.Writer, blocksize int) content.Writer {
+	if blocksize == 0 {
+		blocksize = DefaultBlocksize
+	}
+	return NewPassthroughWriter(writer, func(r io.Reader, w io.Writer, done chan<- error) {
+		gr, err := gzip.NewReader(r)
+		if err != nil {
+			done <- fmt.Errorf("error creating gzip reader: %v", err)
+			return
+		}
+		// write out the uncompressed data
+		b := make([]byte, blocksize, blocksize)
+		for {
+			var n int
+			n, err = gr.Read(b)
+			if err != nil && err != io.EOF {
+				err = fmt.Errorf("GunzipWriter data read error: %v\n", err)
+				break
+			}
+			l := n
+			if n > len(b) {
+				l = len(b)
+			}
+			if _, err2 := w.Write(b[:l]); err2 != nil {
+				err = fmt.Errorf("GunzipWriter: error writing to underlying writer: %v", err2)
+				break
+			}
+			if err == io.EOF {
+				// clear the error
+				err = nil
+				break
+			}
+		}
+		gr.Close()
+		done <- err
+	})
+}

pkg/content/untar.go

@@ -0,0 +1,62 @@
+package content
+
+import (
+	"archive/tar"
+	"fmt"
+	"io"
+
+	"github.com/containerd/containerd/content"
+)
+
+// NewUntarWriter wrap a writer with an untar, so that the stream is untarred
+func NewUntarWriter(writer content.Writer, blocksize int) content.Writer {
+	if blocksize == 0 {
+		blocksize = DefaultBlocksize
+	}
+	return NewPassthroughWriter(writer, func(r io.Reader, w io.Writer, done chan<- error) {
+		tr := tar.NewReader(r)
+		var err error
+		for {
+			_, err := tr.Next()
+			if err == io.EOF {
+				// clear the error, since we do not pass an io.EOF
+				err = nil
+				break // End of archive
+			}
+			if err != nil {
+				// pass the error on
+				err = fmt.Errorf("UntarWriter tar file header read error: %v", err)
+				break
+			}
+			// write out the untarred data
+			// we can handle io.EOF, just go to the next file
+			// any other errors should stop and get reported
+			b := make([]byte, blocksize, blocksize)
+			for {
+				var n int
+				n, err = tr.Read(b)
+				if err != nil && err != io.EOF {
+					err = fmt.Errorf("UntarWriter file data read error: %v\n", err)
+					break
+				}
+				l := n
+				if n > len(b) {
+					l = len(b)
+				}
+				if _, err2 := w.Write(b[:l]); err2 != nil {
+					err = fmt.Errorf("UntarWriter error writing to underlying writer: %v", err2)
+					break
+				}
+				if err == io.EOF {
+					// go to the next file
+					break
+				}
+			}
+			// did we break with a non-nil and non-EOF error?
+			if err != nil && err != io.EOF {
+				break
+			}
+		}
+		done <- err
+	})
+}