Title: "Automatic" creation of searchable documents from Orac
Hi Kapil, (and fellow Zopatistas)

Well, it's been difficult, but I finally feel as if I'm on the verge or understanding this a bit better..

Thank you for taking the time to explain this.
I would think, it's a fairly simple thing I'm trying to do. Hehehe..  Famous last words.. huh?

So, anyway, I've been staying up till 2 am almost every night trying to absorb as much of this as I can, but I have an (extensive) web development, but not a programmer's background, and I'm very much a Zope newbie, so I need to ask you some more questions.. I hope you don't mind..

You can get a feel for what I am talking about by looking at any of the programs or workshops on the website I manage, www.msri.org.
Currently, I'm doing this site by rendering these fields of info through templates into static pages in Frontier, but I'd like to transition to Zope.
(Its the old webmaster's dream of the "self-maintaining site"  *LOL*)

So here's where I'm at so far.. I have built an Oracle 7 database to hold the event info.. Everything fine there, everything works.. now for the fun stuff..

I have several tables in this "event" section of the database,  programs, workshops and talks. (there actually is one other table for other kinds of events) Each kind of event is treated somewhat differently in the schema. It's a very simple setup with nothing exotic going on..  Logically, a given workshop is usually tied to a program..and a given talk is usually tied to a parent workshop.

I have been successful in setting up the basic web-database interaction. I made the basic insert and query forms that are described in the ZSQL Users Guide, and  as I said in my previous posting my next goal is to be able to create separate documents from the data in the individual records that are inserted into the tables.. I'd like to be able to search them along with the other Zope documents on the site, at once, (using a ZCatalog..hopefully) I also need to be able to quickly build several different kinds of automatically-maintained lists from SQL queries that home in on various subsets of the event data. (Ultimately, this system will be expanded to handle "people" data, which will involve some very large lists of people spanning over 20 years, so this system needs to have decent performance with large datasets...But, even then, the result set of a given query will always be quite small..)

Starting with (on the insert record side) I have an SQL Method for each event type that describes the permissable arguments, and the code describing the multi-field  insert.

(I have a sequence and trigger already working on the database side of things that implements a unique ID -the primary key- for each new program, so the ID is not one of the arguments passed through the ZSQL Method..should I change this? - I could do this some other way..)

I also have another SQL method for each type of event that does a select * from <event_type> so that I can implement pulldowns in the select form for the child workshops of a program. (This is really nice and is one of the reasons I think Zope will be great for building a highly useable web interface to this database,  widgets like this make it easy to keep referential integrity..)

Then, one (in the case of the event types without possible "parents") or both of the SQL Methods are called in an DTML Document input form. When the form is submitted a DTML Method is triggered, which right now only shows an acknowledging thank you to the person who submits the information.

So, if I understand you correctly, **I should then use that action to create a class that is tied to the primary key of the database table. Is there an example somewhere of this RDBMS table-keyed "automatic" creation of ZClasses or DTML documents?*

 I think ZClasses would be preferable because there are several kinds of event pages I'd like to generate for each kind of event.. (for example, sometimes we have short-term housing information we want to show, and a few weeks before a given event, as more information becomes available, I'd like to be able to have it added .. For example, as an event time nears, I almost always have to put up schedules of the talks in a given workshop..)

Anyway, if you have the time and can think of any handy pointers to documentation or relevant examples, Id appreciate it.. But even if not, you already have given me a good start..  It shouldn't be very hard, but even though I've been putting in a lot of time searching, finding information has been difficult for me.. The closest I could find to what I want to do is the (very)  few snippets of information available about "pluggable brains". But there isn't even a HOWTO on this, a question that should be a common question..  If anyone has any pointers for a beginner. I'd be really grateful..

Thank you..

Chris





I really appreciate your help..

Chris

[EMAIL PROTECTED]


Some considerations.

1. Duplicating data is bad.

2. Let the RDMBS do as much of the processing as possible.


A possible solution (TMTOWTDI):


create a dtml method 'create event'. its basically an encapsulation to
simulate triggers. in this method make your call to the insertion of the
data into the sql db. after that call you can create a ZClass or DTML
document with the property that stores the primary key of the data
record in the DB (if you do it as a dtml doc, you either want to clone,
or pass its body, check out the ZQR). this is your basic association of
your objects and the sql data. in a dtml method that lives above your
objects have a body that basically calls a retrieval sql on the db which
takes an arg for the id. The body of your dtml doc/ZClass(index_html)
calls this method for presentation of its data passing its property as
an arg. you can arrange your objects in whatever heirarchy you want as
they will acquire the dtml method and the sql methods for presentation
(or put it in the zclass). For deletion create a similiar trigger
encapsulation method, delete row, delete object.


caveat. DON't use this to display lots of events on a page (hideous
performance). you're much better off going to the db directly and
grabbing all the results you need perhaps with python brain for extra
presentation info.


Cheers

Kapil

Reply via email to