Add an ensureRange() call to ChunkedStream_makeSubStream().

By checking if the data is all present before making a substream, we avoid
cases where we parse part of a stream and then throw a MissingDataException
part-way through, which forces us to later re-read the stream -- possibly
multiple times. This is a sizeable performance win for some cases when file
loading is slow (e.g. over the web).

src/core/chunked_stream.js

@@ -187,6 +187,8 @@ var ChunkedStream = (function ChunkedStreamClosure() {
     },
 
     makeSubStream: function ChunkedStream_makeSubStream(start, length, dict) {
+      this.ensureRange(start, start + length);
+
       function ChunkedStreamSubstream() {}
       ChunkedStreamSubstream.prototype = Object.create(this);
       ChunkedStreamSubstream.prototype.getMissingChunks = function() {