1. Store each document in an independent file. Create a table with file names
2. Process each document (indexing) by parsing it for words. Create a table of words and setup a link table for each word and relate to file names (foriegn key) table.
3. At search time, look up the word and link table to get a list of documents. Present the findings.
4. When user selects a choice (a document), present the document after parsing the document and inserting highlight attributes around search words.
5. This is just a simple idea and does not explain phrases, or complex logic of words.
For improved performance, external actions offloading the process to multiple computers can be considered. Another way to leverage multiple computers is to use SOAP, or XML-RPC (as excellently documented on Scott Cadillac's site). You may require multiple Witango licenses for this approach.
Sri Amudhanar
Maxys Corporation
Ashburn, VA 20147, USA
+1 703 729 0600
[EMAIL PROTECTED]
www.maxys.com
Authorized Witango Reseller.
Hosting, Training & Development.
John Shaw wrote:
Hello again all, I have about 2000 single page transcripts. My org wants user to be able to search the bodies of these transcripts. I'm using SQL2000 for the db, but I'm guessing these searches will take a lot of time. So, I was thinking I should use some sort of indexing process. Like creating a table of words used for searches and ID's of transcripts that contain the word. That way, I could check that table before doing the big search.Does anyone have ideas or advice? Thanks, John Shaw ________________________________________________________________________ TO UNSUBSCRIBE: send a plain text/US ASCII email to [EMAIL PROTECTED] with unsubscribe witango-talk in the message body
________________________________________________________________________
TO UNSUBSCRIBE: send a plain text/US ASCII email to [EMAIL PROTECTED]
with unsubscribe witango-talk in the message body
