Re: Time for Jade to process SUBDOCs

["W. Eliot Kimber" <eliot@xxxxxxxxxxxxxx>]

At 02:45 PM 11/17/98 -0500, Chuck Darney wrote:
>I'm now using Jade to process a document containing many SUBDOCs.  The
>processing now takes from 4 to 6 times as long (8 minutes instead of
>1.5).  I assumed there would be some overhead involved, but wasn't
>expecting this.  Are there any suggestions for reducing processing time?

The time most likely reflects that the data has to be parsed twice: once to
create a single instance and once to process that instance. Coupled with
the writing of the single instance, this doubles or triples the I/O time,
which is often the most time-consuming part of the process.  Also, as each
subdoc presumably has its own DOCTYPE declaration, you're parsing the DTD
declarations once for each subdoc.  If the DTD is especially large (e.g.,
full Docbook), this could add significant processing time.

One way to avoid the DTD parsing overhead is to not use DTDs for the
subdocs. Jade is happy with this.  You can also trim down the DTD used for
the subdocs to just those declarations it actually needs (sometimes this
doesn't help much because of the large numbers of leaf-level elements in
some DTDs, where you can't predict which subset a given document will need).

Cheers,

E.
--
<Address HyTime=bibloc>
W. Eliot Kimber, Senior Consulting SGML Engineer
ISOGEN International Corp.
2200 N. Lamar St., Suite 230, Dallas, TX 75202.  214.953.0004
www.isogen.com
</Address>


 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist