What I find interesting is the results of some quick tests I did: http://gyazo.com/90b9a2ed6c5ec21086ac1968e4e86122.png
(screen shot'd to keep the formatting in place [it seems that the list manager cobbles up some of the syntax/formatting of what I had sent initially]) The results seem inconsistent to me. i.e. in the 2nd search (for the phrase '4th parameter'), the 2nd result presented has multiple instances of the phrase therein, yet it still presented the content from the other field, outside of any ellipses (no, I am not dabbling with the text, just extracting it from the returned table). More testing is in order, I feel... -t P.S. is it possible that what I am seeing are bread crumbs from debugging efforts during FTS dev (or undocumented features), where it is reporting which the conditional field triggered the hit and that which content was pulled from? Might be very useful info, if this is the case, and the docs are just pending update.?. hmmm... -----Original Message----- From: Tod Wulff [mailto:tod.wu...@gmail.com] Sent: Wednesday, December 29, 2010 01:20 PM To: 'General Discussion of SQLite Database' Subject: RE: [sqlite] FTS4 snippets contains unintended content? Dan, Thank you, sir, for your response. Honestly, I don't really know. I had understood from the docs that all FTS snippet content returned would be wrapped in the 'ellipses' text (4th parameter of the snippet construct), with possible ellipses separating multiple token content in the event that a concatenated phrase couldn't be found, per this guidance: "After the M fragments have been located, where M is between two and four as described in the paragraphs above, they are joined together in sorted order with the "ellipses" text separating them. The three modifications enumerated earlier are performed on the text before it is returned." I guess that either I have a flawed understanding of how it works, or something is amiss. Frankly, I had considered what you postulated, but without the text being wrapped in the ellipses, I find myself a bit skeptical.?. More to follow, after some further testing tonight and over the holiday break. -t -----Original Message----- From: sqlite-users-boun...@sqlite.org [mailto:sqlite-users-boun...@sqlite.org] On Behalf Of Dan Kennedy Sent: Wednesday, December 29, 2010 06:05 AM To: sqlite-users@sqlite.org Subject: Re: [sqlite] FTS4 snippets contains unintended content? > RIM_API_FTS_DB MATCH 'bug report API_Version:374 OR API_Version:UAR' > > > [03:41]<Schema> 1: [3.7.4] Release History Of SQLite ( http://j.mp/fHI3HG ) > ...generator bug introduced in version 3.6.14 . This bug ...status() > interface, in order to report out the amount of. 374 I think that is an expected result. The snippet() function is returning two fragments of text (since it cannot find a single fragment that contains all the terms in your query). The second fragment of text contains a single token - "374". Dan. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
