performance / sgml-parse

[Stephan Finkler <sf@xxxxxxxxxx>]

In the following example we parse one sgml file with a
specified dtd and the active dsssl takes a lot of additionaly 
files with the same dtd.
To parse the same dtd for a lot of 'sub-files' (i.e. over 500) cost 
a lot of performance.

To solve this problem with Subdoc's or Entities is (I think) not
possible, because we decide later with the DSSSL wich Sub-Sgml must be
included and
wich not.

Is is possible that jade has the dtd in cache and need only to parse 
the instance-part?

I hope somebody can help me.

Greetings Stephan Finkler
sf@xxxxxxxxxx


-- example.sgm ----------------------------
   <!DOCTYPE example SYSTEM "example.dtd">
   <example>
      <part LINK="1.sgm">
      <part LINK="2.sgm">
      <part LINK="3.sgm">
   </example>
-------------------------------------------

-- example.dtd ----------------------------

   <!ELEMENT example - O (part)*>
   <!ELEMENT part - O  EMPTY>
   <!ELEMENT subpart - O  (#PCDATA)>
   <!ATTLIST (part) LINK CDATA #IMPLIED>
-------------------------------------------

-- example.dsl ----------------------------
   <!doctype style-sheet PUBLIC "-//James Clark//DTD DSSSL Style Sheet//EN">
   
   (element part
      (process-node-list (sgml-parse (attribute-string "LINK"
(current-node))))
   )
-------------------------------------------

-- 1.sgm ----------------------------------
   <!DOCTYPE subpart SYSTEM "example.dtd"> 
   <subpart> hello 111 </subpart>
-------------------------------------------

-- 2.sgm ----------------------------------
   <!DOCTYPE subpart SYSTEM "example.dtd"> 
   <subpart> hello 222 </subpart>
-------------------------------------------

-- 3.sgm ----------------------------------
   <!DOCTYPE subpart SYSTEM "example.dtd"> 
   <subpart> hello 333 </subpart>
-------------------------------------------


result:
 hello 111  hello 222  hello 333 




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