camget: finish --shared fetching support

Change-Id: I32edf63e01068a0e96f3255ba4d1313682cf03c4

pkg/blobref/blobref.go

@@ -26,8 +26,9 @@ import (
 
 // Pattern is the regular expression which matches a blobref.
 // It does not contain ^ or $.
-const Pattern = `\b([a-z0-9]+)-([a-f0-9]+)\b`
+const Pattern = `\b([a-z][a-z0-9]*)-([a-f0-9]+)\b`
 
+// whole blobref pattern
 var kBlobRefPattern = regexp.MustCompile("^" + Pattern + "$")
 
 var supportedDigests = map[string]func() hash.Hash{

pkg/client/get.go

@@ -22,6 +22,7 @@ import (
 	"fmt"
 	"io"
 	"log"
+	"regexp"
 
 	"camlistore.org/pkg/blobref"
 	"camlistore.org/pkg/readerutil"
@@ -40,9 +41,32 @@ func (c *Client) FetchMap(b *blobref.BlobRef) (schema.Map, error) {
 }
 
 func (c *Client) FetchStreaming(b *blobref.BlobRef) (io.ReadCloser, int64, error) {
-	return c.FetchVia(b, nil)
+	return c.FetchVia(b, c.viaPathTo(b))
 }
 
+func (c *Client) viaPathTo(b *blobref.BlobRef) (path []*blobref.BlobRef) {
+	if c.via == nil {
+		return nil
+	}
+	it := b.String()
+	// Append path backwards first,
+	for {
+		v := c.via[it]
+		if v == "" {
+			break
+		}
+		path = append(path, blobref.MustParse(v))
+		it = v
+	}
+	// Then reverse it
+	for i := 0; i < len(path)/2; i++ {
+		path[i], path[len(path)-i-1] = path[len(path)-i-1], path[i]
+	}
+	return
+}
+
+var blobsRx = regexp.MustCompile(blobref.Pattern)
+
 func (c *Client) FetchVia(b *blobref.BlobRef, v []*blobref.BlobRef) (io.ReadCloser, int64, error) {
 	pfx, err := c.prefix()
 	if err != nil {
@@ -77,7 +101,30 @@ func (c *Client) FetchVia(b *blobref.BlobRef, v []*blobref.BlobRef) (io.ReadClos
 		return nil, 0, errors.New("blobserver didn't return a Content-Length for blob")
 	}
 
-	return resp.Body, size, nil
+	if c.via == nil {
+		// Not in sharing mode, so return immediately.
+		return resp.Body, size, nil
+	}
+
+	// Slurp 1 MB to find references to other blobrefs for the via path.
+	const maxSlurp = 1 << 20
+	var buf bytes.Buffer
+	_, err = io.Copy(&buf, io.LimitReader(resp.Body, maxSlurp))
+	if err != nil {
+		return nil, 0, err
+	}
+	// If it looks like a JSON schema blob (starts with '{')
+	if schema.LikelySchemaBlob(buf.Bytes()) {
+		for _, blobstr := range blobsRx.FindAllString(buf.String(), -1) {
+			c.via[blobstr] = b.String()
+		}
+	}
+	// Read from the multireader, but close the HTTP response body.
+	type rc struct {
+		io.Reader
+		io.Closer
+	}
+	return rc{io.MultiReader(&buf, resp.Body), resp.Body}, size, nil
 }
 
 func (c *Client) ReceiveBlob(blob *blobref.BlobRef, source io.Reader) (blobref.SizedBlobRef, error) {

pkg/schema/schema.go

@@ -265,18 +265,18 @@ type Superset struct {
 	Members []string `json:"members"` // for static sets (for directory static-sets: blobrefs to child dirs/files)
 
 	// Target is a "share" blob's target (the thing being shared)
-	Target *blobref.BlobRef
+	Target *blobref.BlobRef `json:"target"`
 	// Transitive is a property of a "share" blob.
-	Transitive bool
+	Transitive bool `json:"transitive"`
 	// AuthType is a "share" blob's authentication type that is required.
 	// Currently (2013-01-02) just "haveref" (if you know the share's blobref,
 	// you get access: the secret URL model)
-	AuthType string
+	AuthType string `json:"authType"`
 }
 
 func ParseSuperset(r io.Reader) (*Superset, error) {
 	var ss Superset
-	return &ss, json.NewDecoder(io.LimitReader(r, 1 << 20)).Decode(&ss)
+	return &ss, json.NewDecoder(io.LimitReader(r, 1<<20)).Decode(&ss)
 }
 
 // BytesPart is the type representing one of the "parts" in a "file"
@@ -667,3 +667,14 @@ func RFC3339FromTime(t time.Time) string {
 	}
 	return t.UTC().Format(time.RFC3339Nano)
 }
+
+var bytesCamliVersion = []byte("camliVersion")
+
+// LikelySchemaBlob returns quickly whether buf likely contains (or is
+// the prefix of) a schema blob.
+func LikelySchemaBlob(buf []byte) bool {
+	if len(buf) == 0 || buf[0] != '{' {
+		return false
+	}
+	return bytes.Contains(buf, bytesCamliVersion)
+}