From: <mc...@ci...> - 2003-01-21 17:53:10
|
In-Reply-To: <E18...@sc...> Systems Librarian <po...@li...> wrote: > I have been getting some strange results on htdig and have been unable > to determine why. Any suggestions will be appreciated! > > Try searching TAX FORMS at > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl. Dale Poulter In that page you have <input type="hidden" name="method" value="ALL"> In subsequent pages you have <select name="method"><option value="and">All /*...*/ which is correct for the name and CASE of the method value, and works. I've not read the code to find out whether the match_method setting in your config file takes effect when you specify an invalid value, or whether the invalid value triggers a hard-coded default to "or". So best fix the .pl :-) Mike |
From: Gilles D. <gr...@sc...> - 2003-01-21 18:29:04
|
According to Mike Holderness: > Systems Librarian <po...@li...> wrote: > > I have been getting some strange results on htdig and have been unable > > to determine why. Any suggestions will be appreciated! > > > > Try searching TAX FORMS at > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl. Dale Poulter > > In that page you have > <input type="hidden" name="method" value="ALL"> > > In subsequent pages you have > <select name="method"><option value="and">All /*...*/ > which is correct for the name and CASE of the method value, and works. > > I've not read the code to find out whether the match_method > setting in your config file takes effect when you specify > an invalid value, or whether the invalid value triggers a > hard-coded default to "or". So best fix the .pl :-) Good catch, Mike. You're more observant than the rest of us. It helps to be more specific in a problem report about exactly what the problem is, or a lot of us will miss it. htsearch does indeed default to the "or" method, if the "method" input parameter or "match_method" attribute is not explicitly set to either "boolean" or "and" (lower-case only!). By changing the "ALL" in the URL of the results page from the above search, to "and", the number of matches dropped from 530 to 37, which is likely closer to what Dale wanted. Is that right, Dale, or is there another unspecified problem that's escaped all our eyes? If it's phrase matching you're after, see http://www.htdig.org/FAQ.html#q1.9 -- Gilles R. Detillieux E-mail: <gr...@sc...> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba Winnipeg, MB R3E 3J7 (Canada) |
From: Dale P. <Poulter@LIBRARY.Vanderbilt.edu> - 2003-01-21 19:18:03
|
YES! Thanks, that did seem to resolve most of my problems. I am still getting a bad result when search NABOKOV. Searching from http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I get two listings for "Sigaux Project: Box 28 Content list" the second listing links to box54 and does not contain the word NABOKOV. The index is rebuilt each evening. Any suggestions? Thanks. > According to Mike Holderness: > > Systems Librarian <po...@li...> wrote: > > > I have been getting some strange results on htdig and have been > > > unable to determine why. Any suggestions will be appreciated! > > > > > > Try searching TAX FORMS at > > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl. Dale > > > Poulter > > > > In that page you have > > <input type="hidden" name="method" value="ALL"> > > > > In subsequent pages you have > > <select name="method"><option value="and">All /*...*/ > > which is correct for the name and CASE of the method value, and > > works. > > > > I've not read the code to find out whether the match_method > > setting in your config file takes effect when you specify > > an invalid value, or whether the invalid value triggers a > > hard-coded default to "or". So best fix the .pl :-) > > Good catch, Mike. You're more observant than the rest of us. It > helps to be more specific in a problem report about exactly what the > problem is, or a lot of us will miss it. > > htsearch does indeed default to the "or" method, if the "method" input > parameter or "match_method" attribute is not explicitly set to either > "boolean" or "and" (lower-case only!). By changing the "ALL" in the > URL of the results page from the above search, to "and", the number of > matches dropped from 530 to 37, which is likely closer to what Dale > wanted. Is that right, Dale, or is there another unspecified problem > that's escaped all our eyes? If it's phrase matching you're after, > see http://www.htdig.org/FAQ.html#q1.9 > > -- > Gilles R. Detillieux E-mail: <gr...@sc...> > Spinal Cord Research Centre WWW: > http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba > Winnipeg, MB R3E 3J7 (Canada) > > > ------------------------------------------------------- > This SF.net email is sponsored by: Scholarships for Techies! > Can't afford IT training? All 2003 ictp students receive scholarships. > Get hands-on training in Microsoft, Cisco, Sun, Linux/UNIX, and more. > www.ictp.com/training/sourceforge.asp > _______________________________________________ htdig-general mailing > list <htd...@li...> To unsubscribe, send a > message to <htd...@li...> with a > subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html Dale Poulter Systems Librarian Library Information Technology Services Vanderbilt University Suite 700 110 21st Avenue South Nashville, TN 37240 (615)343-5388 (615)343-8834 (fax) po...@li... |
From: Gilles D. <gr...@sc...> - 2003-01-21 19:36:06
|
According to Dale Poulter: > YES! Thanks, that did seem to resolve most of my problems. I am still getting a bad > result when search NABOKOV. Searching from > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I get two listings for > "Sigaux Project: Box 28 Content list" the second listing links to box54 and does not > contain the word NABOKOV. The index is rebuilt each evening. Any suggestions? Are there any documents with links to either of these two pages, which might contain NABOKOV in the link description text? -- Gilles R. Detillieux E-mail: <gr...@sc...> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba Winnipeg, MB R3E 3J7 (Canada) |
From: Dale P. <Poulter@LIBRARY.Vanderbilt.edu> - 2003-01-21 19:47:04
|
Not that I have seen. Box 28 does contain NABOKOV but not box54. Any idea why the page title is incorrect? The same problem occurs with the last hit Box 62. Thanks,. > According to Dale Poulter: > > YES! Thanks, that did seem to resolve most of my problems. I am > > still getting a bad result when search NABOKOV. Searching from > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I > > get two listings for "Sigaux Project: Box 28 Content list" the > > second listing links to box54 and does not contain the word NABOKOV. > > The index is rebuilt each evening. Any suggestions? > > Are there any documents with links to either of these two pages, which > might contain NABOKOV in the link description text? > > -- > Gilles R. Detillieux E-mail: <gr...@sc...> > Spinal Cord Research Centre WWW: > http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba > Winnipeg, MB R3E 3J7 (Canada) Dale Poulter Systems Librarian Library Information Technology Services Vanderbilt University Suite 700 110 21st Avenue South Nashville, TN 37240 (615)343-5388 (615)343-8834 (fax) po...@li... |
From: Jim C. <li...@yg...> - 2003-01-21 22:30:33
|
On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > YES! Thanks, that did seem to resolve most of my problems. I am > still getting a bad > result when search NABOKOV. Searching from > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I get > two listings for > "Sigaux Project: Box 28 Content list" the second listing links to > box54 and does not > contain the word NABOKOV. The index is rebuilt each evening. Any > suggestions? It looks like a database problem. A title and excerpt for one document are clearly being associated with a link for another. Are you rebuilding the index from scratch every night or just updating the index? If the latter, you might try rebuilding from scratch and checking to see if the problem reoccurs. Jim |
From: Dale P. <Poulter@LIBRARY.Vanderbilt.edu> - 2003-01-22 22:09:35
|
Correct. The rebuild last night corrected most of the problems. I am still getting a couple of bad hits on tax forms but believe it will be corrected in the morning (hopefully). Thanks for the help! > According to Jim Cole: > > On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > > > YES! Thanks, that did seem to resolve most of my problems. I am > > > still getting a bad result when search NABOKOV. Searching from > > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I > > > get two listings for "Sigaux Project: Box 28 Content list" the > > > second listing links to box54 and does not contain the word > > > NABOKOV. The index is rebuilt each evening. Any suggestions? > > > > It looks like a database problem. A title and excerpt for one > > document are clearly being associated with a link for another. Are > > you rebuilding the index from scratch every night or just updating > > the index? If the latter, you might try rebuilding from scratch and > > checking to see if the problem reoccurs. > > I'm unable to reproduce the problem. At least not today. I get 6 > matches for NABOKOV, and 5 of the 6 matching pages to indeed contain > that name. The 4th match, > http://www.library.vanderbilt.edu/central/locations.html, gives a File > Not Found error right now. > > -- > Gilles R. Detillieux E-mail: <gr...@sc...> > Spinal Cord Research Centre WWW: > http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba > Winnipeg, MB R3E 3J7 (Canada) Dale Poulter Systems Librarian Library Information Technology Services Vanderbilt University Suite 700 110 21st Avenue South Nashville, TN 37240 (615)343-5388 (615)343-8834 (fax) po...@li... |
From: Gilles D. <gr...@sc...> - 2003-01-22 22:40:10
|
According to Jim Cole: > On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > > YES! Thanks, that did seem to resolve most of my problems. I am > > still getting a bad > > result when search NABOKOV. Searching from > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I get > > two listings for > > "Sigaux Project: Box 28 Content list" the second listing links to > > box54 and does not > > contain the word NABOKOV. The index is rebuilt each evening. Any > > suggestions? > > It looks like a database problem. A title and excerpt for one document > are clearly being associated with a link for another. Are you > rebuilding the index from scratch every night or just updating the > index? If the latter, you might try rebuilding from scratch and > checking to see if the problem reoccurs. I'm unable to reproduce the problem. At least not today. I get 6 matches for NABOKOV, and 5 of the 6 matching pages to indeed contain that name. The 4th match, http://www.library.vanderbilt.edu/central/locations.html, gives a File Not Found error right now. -- Gilles R. Detillieux E-mail: <gr...@sc...> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba Winnipeg, MB R3E 3J7 (Canada) |
From: Dale P. <Poulter@LIBRARY.Vanderbilt.edu> - 2003-01-23 13:57:55
|
Thanks. The error appears this morning. I am confused as to why the first document does not contain Nabokov and contains the incorrect title. Any ideas? > According to Jim Cole: > > On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > > > YES! Thanks, that did seem to resolve most of my problems. I am > > > still getting a bad result when search NABOKOV. Searching from > > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I > > > get two listings for "Sigaux Project: Box 28 Content list" the > > > second listing links to box54 and does not contain the word > > > NABOKOV. The index is rebuilt each evening. Any suggestions? > > > > It looks like a database problem. A title and excerpt for one > > document are clearly being associated with a link for another. Are > > you rebuilding the index from scratch every night or just updating > > the index? If the latter, you might try rebuilding from scratch and > > checking to see if the problem reoccurs. > > I'm unable to reproduce the problem. At least not today. I get 6 > matches for NABOKOV, and 5 of the 6 matching pages to indeed contain > that name. The 4th match, > http://www.library.vanderbilt.edu/central/locations.html, gives a File > Not Found error right now. > > -- > Gilles R. Detillieux E-mail: <gr...@sc...> > Spinal Cord Research Centre WWW: > http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba > Winnipeg, MB R3E 3J7 (Canada) > > > ------------------------------------------------------- > This SF.net email is sponsored by: Scholarships for Techies! > Can't afford IT training? All 2003 ictp students receive scholarships. > Get hands-on training in Microsoft, Cisco, Sun, Linux/UNIX, and more. > www.ictp.com/training/sourceforge.asp > _______________________________________________ htdig-general mailing > list <htd...@li...> To unsubscribe, send a > message to <htd...@li...> with a > subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html Dale Poulter Systems Librarian Library Information Technology Services Vanderbilt University Suite 700 110 21st Avenue South Nashville, TN 37240 (615)343-5388 (615)343-8834 (fax) po...@li... |
From: Gilles D. <gr...@sc...> - 2003-01-23 23:52:45
|
According to Dale Poulter: > Thanks. The error appears this morning. I am confused as to why the first > document does not contain Nabokov and contains the incorrect title. Any ideas? All I'm getting now is the no matches page. However, I did notice you're running 3.2.0b3, which is horribly buggy and outdated. You should either be running 3.1.6 (the latest stable release), or a recent 3.2.0b4 development snapshot (http://www.htdig.org/files/snapshots/) if you need the 3.2 beta features. I may have inadvertently mislead you earlier, as FAQ 1.9 was sorely out of date as well. I've just updated it. > > According to Jim Cole: > > > On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > > > > YES! Thanks, that did seem to resolve most of my problems. I am > > > > still getting a bad result when search NABOKOV. Searching from > > > > http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for NABOKOV I > > > > get two listings for "Sigaux Project: Box 28 Content list" the > > > > second listing links to box54 and does not contain the word > > > > NABOKOV. The index is rebuilt each evening. Any suggestions? > > > > > > It looks like a database problem. A title and excerpt for one > > > document are clearly being associated with a link for another. Are > > > you rebuilding the index from scratch every night or just updating > > > the index? If the latter, you might try rebuilding from scratch and > > > checking to see if the problem reoccurs. > > > > I'm unable to reproduce the problem. At least not today. I get 6 > > matches for NABOKOV, and 5 of the 6 matching pages to indeed contain > > that name. The 4th match, > > http://www.library.vanderbilt.edu/central/locations.html, gives a File > > Not Found error right now. -- Gilles R. Detillieux E-mail: <gr...@sc...> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba Winnipeg, MB R3E 3J7 (Canada) |
From: Dale P. <Poulter@LIBRARY.Vanderbilt.edu> - 2003-01-31 16:40:40
|
Giles, Thanks. I have tried several times with a couple of the snapshots to compile 3.2.0b4 but continually get an error with the Z_DEFAULT_COMPRESSION undeclared in mp_cmpr.c line 140. I am compiling without zlib (also have tried with) using gnumake and gcc on solaris 8. Any suggestions as to the cause of this problem? I searched the archive and notice another report of the problem but no solutions. Thanks for your assistance. -Dale > According to Dale Poulter: > > Thanks. The error appears this morning. I am confused as to why > > the first document does not contain Nabokov and contains the > > incorrect title. Any ideas? > > All I'm getting now is the no matches page. However, I did notice > you're running 3.2.0b3, which is horribly buggy and outdated. You > should either be running 3.1.6 (the latest stable release), or a > recent 3.2.0b4 development snapshot > (http://www.htdig.org/files/snapshots/) if you need the 3.2 beta > features. I may have inadvertently mislead you earlier, as FAQ 1.9 > was sorely out of date as well. I've just updated it. > > > > According to Jim Cole: > > > > On Tuesday, January 21, 2003, at 12:19 PM, Dale Poulter wrote: > > > > > YES! Thanks, that did seem to resolve most of my problems. I > > > > > am still getting a bad result when search NABOKOV. Searching > > > > > from http://lib11.library.vanderbilt.edu/diglib/a-to-z.pl for > > > > > NABOKOV I get two listings for "Sigaux Project: Box 28 Content > > > > > list" the second listing links to box54 and does not contain > > > > > the word NABOKOV. The index is rebuilt each evening. Any > > > > > suggestions? > > > > > > > > It looks like a database problem. A title and excerpt for one > > > > document are clearly being associated with a link for another. > > > > Are you rebuilding the index from scratch every night or just > > > > updating the index? If the latter, you might try rebuilding from > > > > scratch and checking to see if the problem reoccurs. > > > > > > I'm unable to reproduce the problem. At least not today. I get 6 > > > matches for NABOKOV, and 5 of the 6 matching pages to indeed > > > contain that name. The 4th match, > > > http://www.library.vanderbilt.edu/central/locations.html, gives a > > > File Not Found error right now. > > > -- > Gilles R. Detillieux E-mail: <gr...@sc...> > Spinal Cord Research Centre WWW: > http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba > Winnipeg, MB R3E 3J7 (Canada) > > > ------------------------------------------------------- > This SF.NET email is sponsored by: > SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! > http://www.vasoftware.com > _______________________________________________ htdig-general mailing > list <htd...@li...> To unsubscribe, send a > message to <htd...@li...> with a > subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html Dale Poulter Systems Librarian Library Information Technology Services Vanderbilt University Suite 700 110 21st Avenue South Nashville, TN 37240 (615)343-5388 (615)343-8834 (fax) po...@li... |