package main
import (
- "flag"
"fmt"
"html"
"io"
- "mime"
"net/http"
"net/url"
"os"
+ "path"
+ "strconv"
"strings"
+ "sync"
"time"
"git.curoverse.com/arvados.git/sdk/go/arvadosclient"
"git.curoverse.com/arvados.git/sdk/go/keepclient"
)
-type handler struct{}
-
-var (
- clientPool = arvadosclient.MakeClientPool()
- trustAllContent = false
- anonymousTokens []string
- attachmentOnlyHost = ""
-)
-
-func init() {
- flag.BoolVar(&trustAllContent, "trust-all-content", false,
- "Serve non-public content from a single origin. Dangerous: read docs before using!")
- flag.StringVar(&attachmentOnlyHost, "attachment-only-host", "",
- "Accept credentials, and add \"Content-Disposition: attachment\" response headers, for requests at this hostname:port. Prohibiting inline display makes it possible to serve untrusted and non-public content from a single origin, i.e., without wildcard DNS or SSL.")
+type handler struct {
+ Config *Config
+ clientPool *arvadosclient.ClientPool
+ setupOnce sync.Once
}
-// return a UUID or PDH if s begins with a UUID or URL-encoded PDH;
-// otherwise return "".
-func parseCollectionIdFromDNSName(s string) string {
+// parseCollectionIDFromDNSName returns a UUID or PDH if s begins with
+// a UUID or URL-encoded PDH; otherwise "".
+func parseCollectionIDFromDNSName(s string) string {
// Strip domain.
if i := strings.IndexRune(s, '.'); i >= 0 {
s = s[:i]
var urlPDHDecoder = strings.NewReplacer(" ", "+", "-", "+")
-// return a UUID or PDH if s is a UUID or a PDH (even if it is a PDH
-// with "+" replaced by " " or "-"); otherwise return "".
-func parseCollectionIdFromURL(s string) string {
+// parseCollectionIDFromURL returns a UUID or PDH if s is a UUID or a
+// PDH (even if it is a PDH with "+" replaced by " " or "-");
+// otherwise "".
+func parseCollectionIDFromURL(s string) string {
if arvadosclient.UUIDMatch(s) {
return s
}
return ""
}
+func (h *handler) setup() {
+ h.clientPool = arvadosclient.MakeClientPool()
+}
+
+// ServeHTTP implements http.Handler.
func (h *handler) ServeHTTP(wOrig http.ResponseWriter, r *http.Request) {
+ h.setupOnce.Do(h.setup)
+
var statusCode = 0
var statusText string
httpserver.Log(remoteAddr, statusCode, statusText, w.WroteBodyBytes(), r.Method, r.Host, r.URL.Path, r.URL.RawQuery)
}()
+ if r.Method == "OPTIONS" {
+ method := r.Header.Get("Access-Control-Request-Method")
+ if method != "GET" && method != "POST" {
+ statusCode = http.StatusMethodNotAllowed
+ return
+ }
+ w.Header().Set("Access-Control-Allow-Headers", "Range")
+ w.Header().Set("Access-Control-Allow-Methods", "GET, POST")
+ w.Header().Set("Access-Control-Allow-Origin", "*")
+ w.Header().Set("Access-Control-Max-Age", "86400")
+ statusCode = http.StatusOK
+ return
+ }
+
if r.Method != "GET" && r.Method != "POST" {
statusCode, statusText = http.StatusMethodNotAllowed, r.Method
return
}
- arv := clientPool.Get()
+ if r.Header.Get("Origin") != "" {
+ // Allow simple cross-origin requests without user
+ // credentials ("user credentials" as defined by CORS,
+ // i.e., cookies, HTTP authentication, and client-side
+ // SSL certificates. See
+ // http://www.w3.org/TR/cors/#user-credentials).
+ w.Header().Set("Access-Control-Allow-Origin", "*")
+ }
+
+ arv := h.clientPool.Get()
if arv == nil {
- statusCode, statusText = http.StatusInternalServerError, "Pool failed: "+clientPool.Err().Error()
+ statusCode, statusText = http.StatusInternalServerError, "Pool failed: "+h.clientPool.Err().Error()
return
}
- defer clientPool.Put(arv)
+ defer h.clientPool.Put(arv)
pathParts := strings.Split(r.URL.Path[1:], "/")
- var targetId string
+ var targetID string
var targetPath []string
var tokens []string
var reqTokens []string
var pathToken bool
var attachment bool
- credentialsOK := trustAllContent
+ credentialsOK := h.Config.TrustAllContent
- if r.Host != "" && r.Host == attachmentOnlyHost {
+ if r.Host != "" && r.Host == h.Config.AttachmentOnlyHost {
credentialsOK = true
attachment = true
} else if r.FormValue("disposition") == "attachment" {
attachment = true
}
- if targetId = parseCollectionIdFromDNSName(r.Host); targetId != "" {
+ if targetID = parseCollectionIDFromDNSName(r.Host); targetID != "" {
// http://ID.collections.example/PATH...
credentialsOK = true
targetPath = pathParts
} else if len(pathParts) >= 2 && strings.HasPrefix(pathParts[0], "c=") {
// /c=ID/PATH...
- targetId = parseCollectionIdFromURL(pathParts[0][2:])
+ targetID = parseCollectionIDFromURL(pathParts[0][2:])
targetPath = pathParts[1:]
} else if len(pathParts) >= 3 && pathParts[0] == "collections" {
if len(pathParts) >= 5 && pathParts[1] == "download" {
// /collections/download/ID/TOKEN/PATH...
- targetId = pathParts[2]
+ targetID = parseCollectionIDFromURL(pathParts[2])
tokens = []string{pathParts[3]}
targetPath = pathParts[4:]
pathToken = true
} else {
// /collections/ID/PATH...
- targetId = pathParts[1]
- tokens = anonymousTokens
+ targetID = parseCollectionIDFromURL(pathParts[1])
+ tokens = h.Config.AnonymousTokens
targetPath = pathParts[2:]
}
- } else {
+ }
+
+ if targetID == "" {
statusCode = http.StatusNotFound
return
}
- if t := r.FormValue("api_token"); t != "" {
+
+ formToken := r.FormValue("api_token")
+ if formToken != "" && r.Header.Get("Origin") != "" && attachment && r.URL.Query().Get("api_token") == "" {
+ // The client provided an explicit token in the POST
+ // body. The Origin header indicates this *might* be
+ // an AJAX request, in which case redirect-with-cookie
+ // won't work: we should just serve the content in the
+ // POST response. This is safe because:
+ //
+ // * We're supplying an attachment, not inline
+ // content, so we don't need to convert the POST to
+ // a GET and avoid the "really resubmit form?"
+ // problem.
+ //
+ // * The token isn't embedded in the URL, so we don't
+ // need to worry about bookmarks and copy/paste.
+ tokens = append(tokens, formToken)
+ } else if formToken != "" {
// The client provided an explicit token in the query
// string, or a form in POST body. We must put the
// token in an HttpOnly cookie, and redirect to the
// It is not safe to copy the provided token
// into a cookie unless the current vhost
// (origin) serves only a single collection or
- // we are in trustAllContent mode.
+ // we are in TrustAllContent mode.
statusCode = http.StatusBadRequest
return
}
http.SetCookie(w, &http.Cookie{
Name: "arvados_api_token",
- Value: auth.EncodeTokenCookie([]byte(t)),
+ Value: auth.EncodeTokenCookie([]byte(formToken)),
Path: "/",
- Expires: time.Now().AddDate(10, 0, 0),
HttpOnly: true,
})
- redir := (&url.URL{Host: r.Host, Path: r.URL.Path}).String()
+
+ // Propagate query parameters (except api_token) from
+ // the original request.
+ redirQuery := r.URL.Query()
+ redirQuery.Del("api_token")
+
+ redir := (&url.URL{
+ Host: r.Host,
+ Path: r.URL.Path,
+ RawQuery: redirQuery.Encode(),
+ }).String()
w.Header().Add("Location", redir)
statusCode, statusText = http.StatusSeeOther, redir
if credentialsOK {
reqTokens = auth.NewCredentialsFromHTTPRequest(r).Tokens
}
- tokens = append(reqTokens, anonymousTokens...)
+ tokens = append(reqTokens, h.Config.AnonymousTokens...)
}
if len(targetPath) > 0 && targetPath[0] == "_" {
collection := make(map[string]interface{})
found := false
for _, arv.ApiToken = range tokens {
- err := arv.Get("collections", targetId, nil, &collection)
+ err := arv.Get("collections", targetID, nil, &collection)
if err == nil {
// Success
found = true
statusCode, statusText = http.StatusInternalServerError, err.Error()
return
}
+ if kc.Client != nil && kc.Client.Transport != nil {
+ // Workaround for https://dev.arvados.org/issues/9005
+ if t, ok := kc.Client.Transport.(*http.Transport); ok {
+ defer t.CloseIdleConnections()
+ }
+ }
rdr, err := kc.CollectionFileReader(collection, filename)
if os.IsNotExist(err) {
statusCode = http.StatusNotFound
}
defer rdr.Close()
- // One or both of these can be -1 if not found:
- basenamePos := strings.LastIndex(filename, "/")
- extPos := strings.LastIndex(filename, ".")
- if extPos > basenamePos {
- // Now extPos is safely >= 0.
- if t := mime.TypeByExtension(filename[extPos:]); t != "" {
- w.Header().Set("Content-Type", t)
- }
- }
- w.Header().Set("Content-Length", fmt.Sprintf("%d", rdr.Len()))
- if attachment {
- w.Header().Set("Content-Disposition", "attachment")
- }
+ basename := path.Base(filename)
+ applyContentDispositionHdr(w, r, basename, attachment)
- w.WriteHeader(http.StatusOK)
- _, err = io.Copy(w, rdr)
+ modstr, _ := collection["modified_at"].(string)
+ modtime, err := time.Parse(time.RFC3339Nano, modstr)
if err != nil {
- statusCode, statusText = http.StatusBadGateway, err.Error()
+ modtime = time.Now()
+ }
+ http.ServeContent(w, r, basename, modtime, rdr)
+}
+
+func applyContentDispositionHdr(w http.ResponseWriter, r *http.Request, filename string, isAttachment bool) {
+ disposition := "inline"
+ if isAttachment {
+ disposition = "attachment"
+ }
+ if strings.ContainsRune(r.RequestURI, '?') {
+ // Help the UA realize that the filename is just
+ // "filename.txt", not
+ // "filename.txt?disposition=attachment".
+ //
+ // TODO(TC): Follow advice at RFC 6266 appendix D
+ disposition += "; filename=" + strconv.QuoteToASCII(filename)
+ }
+ if disposition != "inline" {
+ w.Header().Set("Content-Disposition", disposition)
}
}