From: Gilles D. <gr...@sc...> - 2001-12-05 19:49:56
|
According to Geoff Hutchison: > At 4:05 PM -0500 11/30/01, tr...@ma... wrote: > >a) how do you restrict the search to meta tagged keywords? > > As per the FAQ. Keep in mind that for 3.2 code, you don't have to do > any reindexing--you'll just want to make a "keyword-only.conf" config > file that sets the scoring of all the various _factor attributes to 0 > *except* the keywords_factor. Note that scores of 0 don't (yet) cause search result to be suppressed, so it doesn't quite restrict the search exclusively to meta keywords, but it gives them a whole lot more weight in the score. I've clarified this point in the FAQ. > >b) can searches produce results from both the meta tag keywords and > >the document? > > By default they do. But remember that the excerpt shown in the search > results is from the document--it won't display the keywords though > they are used in the retrieval and scoring. > > >c) can searches be restricted to the document only? > > By a similar means to the FAQ--just set the keywords_factor to 0. > > >where is the "keywords_factor", (did i miss it?)? > > There are *many* attributes that are not specified in the sample > htdig.conf that's installed. See http://www.htdig.org/confindex.html > for more (and http://www.htdig.org/dev/htdig-3.2/confindex.html for > 3.2-specific docs.) See also http://www.htdig.org/FAQ.html#q4.18 Also, to get the most up to date documentation for pre-release snapshots, you should look at the files in the htdoc subdirectory of the source distribution. htdoc/attrs.html will document all the current attrinbutes. > >when it says "in 3.2 you will be able to..." is that present in the > >snapshot of b4 now? if so, how does it work? > > As I said earlier, you *can* change the scoring on the fly--no need > to reindex with releases after 3.2.0b2 or so. On the other hand, you > cannot currently directly restrict a search to a keyword, e.g.: > > bar AND title=foo I clarified that too in the FAQ, without getting into specifics of how the restrictions might be specified. ... and later Ted asked... > > where can i read up on how all that works... hmm.. maybe there is no > > "layman's" place for that, ;) so i ask: > > > > if a search *only* found a (1 or more) match from a meta tagged > > keyword how would that result be displayed? by url only? You're right that there's not a lot of htdig documentation at a layman's level, but at the same time most of the existing documentation isn't over most people's heads. I strongly recommend you take the time to read through it. Many of the questions you asked are dealt with quite clearly and directly in the documentation. For example, htdig.html mentions the standard for robots exclusion and points to the reference document where this is explained in detail. The whole thread on robots.txt would have been mostly unnecessary if you'd looked at this document. Also, there are tons of configuration attributes in attrs.html. While I don't suggest you sit down and read it from start to finish, I do recommend you skim through it for ideas on what is possible, and take the time to search it and the FAQ thoroughly whenever you have a "how can I configure it to do ..." type of question. Most of these questions can be answered by a good read of these two documents. To answer your question about what's displayed in searches matching only meta keywords, see http://www.htdig.org/attrs.html#no_excerpt_text and http://www.htdig.org/attrs.html#no_excerpt_show_top Finally, it rarely hurts to try things out for oneself. I see a lot of questions on this list that go something like "what does htsearch do if I do this...", to which the obvious answer is "have you even tried it, and if so, what do you see when you do that...". It's far to easy to get into a mindset where you "just ask the experts" rather than trying to figure it out, but these trivial questions take time away from more substantial questions. Testing htsearch is very easy to do once you've gotten your databases built. -- Gilles R. Detillieux E-mail: <gr...@sc...> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil Dept. Physiology, U. of Manitoba Phone: (204)789-3766 Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930 |