[xml] Introduction Ideas and Brain Dump
- From: James Michael DuPont <mdupont777 yahoo com>
- To: xml gnome org
- Subject: [xml] Introduction Ideas and Brain Dump
- Date: Thu, 20 Jun 2002 06:42:28 -0700 (PDT)
Hello,
My name is James Michael DuPont,
I have been working on creating an XML/PERL/SQL
interface into the GCC intermediate tree
representation, the introspector.sf.net project.
Over the course of my playing with XML I have come up
with some ideas that I wanted to talk to you all
about.
I have been thinking about using the in-memory DOM of
the LIBXML for doing a couple of following things.
I hope that you find these ideas good.
Many are probably really old and not interesting,
but for me I find them interesting.
Maybe this can all be implemented by modules jus
passing trees around?
The idea is to be able to plug in a set of backends
and post processors into libxml2 and be able to work
with that as a data transfer system.
1. Transfer of data between programs without external
representation. That would allow for an XML link
interface between GPL programs without needing to
transfer the data into a file. Perl could link to it
as a lib and be called back. That would eliminate any
copyright and linking issues that I currently have
with the gcc.
2. Persistent storage of data structures.
the ability to map the data in memory onto a database
access layer. Right now I use perl and
DBI/DBD::Postgres and Mysql.
I found the following links that look interesting :
http://www.xmldb.org/index.html
XUpdate http://www.xmldb.org/xupdate/
3. Visualisation of data structures,
by transforming the XML into a graph language and
visualization by VCG or GraphVis.
availabe at
http://xml.coverpages.org/xgmml.html or
http://www.xml.com/pub/r/723
4. Diffing of database structures.
Once a packet of data is in a DOM we can diff them.
The diff could be made on the structure and on the
content. Maybe some form of cvs or diff modification?
This is done by a couple of programs :
http://www.logilab.org/xmldiff/
http://diffxml.sourceforge.net/
5. Storing a transformation routine between a C data
structure and a XML dom object representing it.
That will come done to keeping the XML document
that is normally outputted just as a Text file
instanciated in memory and X-ref-ed to the memory
structures in the the program.
C OBJECT -->points to---> XML Representation in
memory --> points back to --> C Object
The XML object can be updated and queried with effects
on the C object, and vice versa. Both are just
different views and encodings on the same abstract
Some form of notification mechanism would be cool, as
well as a URI PUT and GET mechanism.
Anyway, I hope to hear from you, and look forward to
working with you.
Mike
=====
James Michael DuPont
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]