What are the steps I need to do to enable text indexing of a CLOB Column containing RTF data?  I understand this includes having an external module for the filtering
the RTF text, but I have no idea what this module should look like.

Has anyone done this (he asks nervously)?

Regards,

SteveG

Steve


I've put text filter & word generator modules in the extensions ng

Roy Lambert [Team Elevate]

Roy,

> I've put text filter & word generator modules in the extensions ng

I guess, he wants to know how to create such a module ( so do I )

Actually the (excellent) help files don't help very much in this case.

Uli

> I guess, he wants to know how to create such a module ( so do I )

> Actually the (excellent) help files don't help very much in this case.

Uli is correct.  That's exactly what I want to know.  While the help files are great, they don't go into enough detail about this.  Any help or guidance would be greatly
appreciated.

Thanks.

SteveG

Hello All,

>> I guess, he wants to know how to create such a module ( so do I )
>
>> Actually the (excellent) help files don't help very much in this case.
>
> Uli is correct.  That's exactly what I want to know.  While the help files are great, they don't go into enough detail about this.  Any help or guidance would be greatly
> appreciated.
>
> Thanks.
>
> SteveG
>

I'm interested too ... In my case i want for tRichView.

--
Charalampos Michael - [Creation Power] - http://www.creationpower.gr

> I'm interested too ... In my case i want for tRichView.

Me too.

Steve


OK.

Point 1
The component you want it for is not important. What the text filters and word generators do is take a string and alter it to reflect what you want to have stored in the full text index. You can think of it as kicking in AFTER your component (in my case WPTools, memo & edit components or their DB equivalent, plus email from the web via Synapse) has said to the nice DB engine "here you go have some data"

Point 2
The easiest way to write your own from scratch is to use the templates provided by Tim in D2006 File | New | Other | ElevateDB

Point 3
There are two components you'll need - the text filter and the word generator. Text filters are the ones that strip out all the unwanted stuff (eg formatting commands) and word generators actually extract discreet words that will be used. If, as an example, you're not concerned with leaving formatting in RTF but want to remove all the "naughty" words you can ignore producing a text filter and just write the word generator

The end result will be one or more dll's to drop into the directory containing the config file.


Being a bighead I'd suggest the best way is to download the ones I've written and have a look. They are as efficient and fleximple as I could make them, however, if you can improve on the performance of the rtf or html filters I'd be very grateful if you'd post the code to the extensions ng.

I agree about the lack of info in the help. Search the ngs under my name and fti or full text indexing or text filter or word generator and you'll see a fair bit of discussion between myself and Tim.

Roy Lambert [Team Elevate]

Uli,

<< Actually the (excellent) help files don't help very much in this case. >>

We kind of avoided this topic on purpose since we want to make sure that
someone has got enough experience to do so before attempting to create these
types of modules.  It's very easy to mess up an EDB system by writing one
improperly.

We do, however, provide template projects to speed up the process:  you'll
find them under File/New/Other/ElevateDB/ElevateDB Text Filter Module.  In
the project source for the text file module you'll find comments that tell
you how to fill in the code properly.  Text filtering is actually the
easiest of the bunch because it is just a matter of looking at the filter
type, filtering the input text, and then setting the output text
accordingly.

--
Tim Young
Elevate Software
www.elevatesoft.com

Tim,

> We kind of avoided this topic on purpose since we want to make sure that
> someone has got enough experience to do so before attempting to create these
> types of modules.  It's very easy to mess up an EDB system by writing one
> improperly.

OK, I see.

> We do, however, provide template projects to speed up the process:  you'll
> find ...

Thank you. At the moment I am not at home but I remember that I was
looking for just changing the "stop words" from English to German and
couldn't find how to do it.

Regards Uli

Uli


The best one for that is the word generator. You can either hard code the words into the dll or do as I do and load from a text file when the dll is initialised.

Roy Lambert [Team Elevate]