Binary Data

Rick Jelliffe ricko at
Mon Feb 23 17:22:52 GMT 1998

From: Michael Emmel <mike at>

>This says to me that binary data is required to either be encoded to ascii
>be included, or have Mime type boundries for XML tags with  binary data
> not containing the mime boundries included.
>In the document or be obtained from a ascii normalized external URI link.

Binary data can only be included in a parseable entity if it is first
in some way which
1) does not contain delimiters which may cause false triggering
2) does not contain any characters which the XML "SGML declaration"
says are unused (or shunned).
Base64 is one such encoding. Other encodings may be more efficient
if you have a 16-bit data stream.

The way to signal you are using an encoding is to use an element
with a notation attribute.

If you embed binary data with MIME type boundaries, you no longer
have a parseable XML entity, you have a MIME multipart file which
can be processed to generate an XML entity.

>There is no way to tell a XML arser to skip x number of  arbitrary bytes of
>embedded unparsed entity  data which is consumed by the "application"  and
>restart the parser
>at the next valid section.

An XML parser is not interested in the contents of a non-XML-parseable
entity. Indexing into binary data is either done before the parser (i.e. by
embedding the appropriate instructions in the system identifier of the
entity) or by the application after the parser.

>Am I wrong ???

What do you mean "restart the parser"?  Parsing continues after an entity

Rick Jelliffe

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at
Archived as:
To (un)subscribe, mailto:majordomo at the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at

More information about the Xml-dev mailing list