What means exactly "Memory error"?

Bengt Richter bokr at oz.net
Sat Apr 26 03:09:42 CEST 2003


On Thu, 24 Apr 2003 14:06:20 +0200, "Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote:

>Hello, all.
>
>I made an application which uses 4DOM. I have a big (~10M) autogenerated HTML 
>file from an Oracle tool. The problem is to public it. Nobody want to open 
>10MB HTML file in their browser. 
>
>OK, so the main idea was to split it using frequently repeated <HR> tag. Then 
>I got ~1500 parts. OK, now I parse each part with HtmlLib and keep it tree 
>somewhere in the variable (in memory). Then after each part is parsed I want 
>to do a lot of changes inside the document to adapt it to using with Zope in 
>the cycle.
>
Are you essentially extracting data that originated from Oracle, so as to
be able to re-publish it wrapped in your own HTML? If so, can you connect to
Oracle and just get the data? Or does the HTML solve access and
interfacing problems you don't want to deal with?

Is the data frequently updated, so caching is not going to help?
Is there a time stamp in the HTML that you can use as a quick check
against repeat processing if you are serving many pages between updates?

How would just extracting the data from the HTML and putting it in your
own simplified data base work out? Could you generate your various pages
that way? Doesn't Zope support various data bases? How much traffic do
you have to support? Someone who uses/develops Zope can advise.

How much of the original 10mb winds up copied as is to your output?

It's hard to guess what a good approach would be, not knowing what
you are actually doing.

Just a few thoughts...

Regards,
Bengt Richter




More information about the Python-list mailing list