dtSearch has a high-end and very powerful search engine solution with numerous search conditions possible.
www.dtsearch.com. You can use com automation to access the search engine and indexes via VFP. Extremely fast.
Also, FoxWeb have a full text search program at
www.foxweb.com which is free at
http://www.foxweb.com/fwFullText/ . I have not tested the FoxWeb one as yet.
>Any suggestions on how to take about a couple million text blobs between 1K and 16K in lengh and index their contents so that a Google-like search can be asserted against them? These blobs are actually informational free-form text reports collected over a serveral year period, stored as a memo field one per record.
An essential factor is that the text needs to parsed and indexed such that the search engine can reasonably recognize a negative condition. Thus when I search on "seismic deficiency", I don't want the engine to return records that say "No seismic deficiency" or "Seismic deficiency is not present".
>Here is the setup:
>- Each text blob is actually a memo field in a VFP table.
>- The table has about 30 other fields. When a "search hit" occurs during a search of the memo field, some of the other fields need to be displayed.
>- I would like the engine to be able to handle boolean combinations that same way. Example with crude minus-sign syntax:
> "seismic deficiency" -"hydraulic overload"
>will find instances where "seismic deficiency" is spoken of as present or likely AND "hydraulic overload" is explicitly spoken of as not present or not likely (or not spoken of at all).
>
>Thanks very much!
In the End, we will remember not the words of our enemies, but the silence of our friends - Martin Luther King, Jr.