> I believe that it is appropriate for contrib/ because it is a good demo > of FTI-like capabilities. But nothing more, yet. For at least a couple > of reasons: > > 1) It generates the "index" as a table, not a PostgreSQL index or > index-like thing. > > 2) It has a hardcoded list of non-indexed words. This should come from a > table, to allow it to be tuned to the application requirements. > > Comments? > > - Thomas > In general.. a) Considering that I was coding up the same thing with triggers and such, things could only get better. b) Check out MSSQL 7's capabilities and weep. c) It would be a start. One its in the tree, it gets used more, gets improved.. It would be a while yet before 7.2 starts, plenty of time then to develop it further. Regards John
- [HACKERS] Full text Indexing -out of contrib and into ma... John Huttley
- Re: [HACKERS] Full text Indexing -out of contrib an... Don Baccus
- Re: [HACKERS] Full text Indexing -out of contri... Bruce Momjian
- Re: [HACKERS] Full text Indexing -out of co... The Hermit Hacker
- Re: [HACKERS] Full text Indexing -out o... Mitch Vincent
- Re: [HACKERS] Full text Indexing -... Bruce Momjian
- Re: [HACKERS] Full text Indexing -out o... Don Baccus
- Re: [HACKERS] Full text Indexing -out o... john huttley
- Re: [HACKERS] Full text Indexing -out of contrib an... Lamar Owen
- Re: [HACKERS] Full text Indexing -out of contri... Thomas Lockhart
- Re: [HACKERS] Full text Indexing -out of co... john huttley
- Re: [HACKERS] Full text Indexing -out o... Hannu Krosing
- RE: [HACKERS] Full text Indexing -out of contrib an... Magnus Hagander