Gateway renders pretty 404 pages if available

In the same way that an `index.html` file is rendered, if one is present, when a
path requested of the gateway is a directory, a `404.html` file is rendered if
the requested file is not present within the specified IPFS object.

`404.html` files are looked for in the directory of the requested path and each
parent until one is found, falling back on the previous error message.

License: MIT
Signed-off-by: JP Hastings-Spital <jphastings@gmail.com>

core/corehttp/gateway_handler.go

@@ -168,6 +168,8 @@ func (i *gatewayHandler) getOrHeadHandler(ctx context.Context, w http.ResponseWr
 		return
 	}
 
+	preventCache := false
+
 	// Resolve path to the final DAG node for the ETag
 	resolvedPath, err := i.api.ResolvePath(ctx, parsedPath)
 	switch err {
@@ -179,8 +181,16 @@ func (i *gatewayHandler) getOrHeadHandler(ctx context.Context, w http.ResponseWr
 		}
 		fallthrough
 	default:
-		webError(w, "ipfs resolve -r "+escapedURLPath, err, http.StatusNotFound)
-		return
+		// Look for a pretty 404 page in any parent folder of the requested file
+		resolved404Path, err404 := findPretty404(ctx, parsedPath, i)
+		if err404 != nil {
+			webError(w, "ipfs resolve -r "+escapedURLPath, err, http.StatusNotFound)
+			return
+		}
+
+		preventCache = true
+		resolvedPath = resolved404Path
+		log.Debugf("using pretty 404.html link for %s", escapedURLPath)
 	}
 
 	dr, err := i.api.Unixfs().Cat(ctx, resolvedPath)
@@ -191,6 +201,7 @@ func (i *gatewayHandler) getOrHeadHandler(ctx context.Context, w http.ResponseWr
 		defer dr.Close()
 	case coreiface.ErrIsDir:
 		dir = true
+		preventCache = true
 	default:
 		webError(w, "ipfs cat "+escapedURLPath, err, http.StatusNotFound)
 		return
@@ -259,7 +270,7 @@ func (i *gatewayHandler) getOrHeadHandler(ctx context.Context, w http.ResponseWr
 	// TODO: break this out when we split /ipfs /ipns routes.
 	modtime := time.Now()
 
-	if strings.HasPrefix(urlPath, ipfsPathPrefix) && !dir {
+	if strings.HasPrefix(urlPath, ipfsPathPrefix) && !preventCache {
 		w.Header().Set("Cache-Control", "public, max-age=29030400, immutable")
 
 		// set modtime to a really long time ago, since files are immutable and should stay cached
@@ -598,3 +609,20 @@ func webErrorWithCode(w http.ResponseWriter, message string, err error, code int
 func internalWebError(w http.ResponseWriter, err error) {
 	webErrorWithCode(w, "internalWebError", err, http.StatusInternalServerError)
 }
+
+func findPretty404(ctx context.Context, parsedPath coreiface.Path, i *gatewayHandler) (coreiface.Path, error) {
+	pathComponents := strings.SplitAfter(gopath.Dir(parsedPath.String()), "/")
+
+	for idx := len(pathComponents); idx >= 3; idx-- {
+		pretty404 := gopath.Join(append(pathComponents[0:idx], "404.html")...)
+		parsedPath, err := coreapi.ParsePath(pretty404)
+		if err == nil {
+			resolvedPath, err := i.api.ResolvePath(ctx, parsedPath)
+			if err == nil {
+				return resolvedPath, nil
+			}
+		}
+	}
+
+	return nil, errors.New("No pretty 404 in any parent folder")
+}