Re: Sorting by highest number of matches unique to a field
This WebDNA talk-list message is from 2003
It keeps the original formatting.
numero = 52132
interpreted = N
texte = Thanks Brian, I'm leaning toward something like that after having thought about this for awhile. This sounds like it would be something that is so simple to do with a normal search and sort, but it's not. Is it just me, or does it seem like the summary should take place after items are sorted? On Friday, August 8, 2003, at 03:34 PM, Brian Fries wrote: > You could store the total number of messages in a thread in the first > message for that thread (msg_rank=0, if I understand you scheme > correctly) - say a "msg_count" field. When you add a message to the > thread, increment "msg_count" in the first thread message. Then you > could sort with > > [search > db=db/ > messages.db&eqmsg_rankdatarq=0&allhit=1&demsg_countsort=1&msg_counttype > =num] > > No need to summ since you only find the first message per thread. > > - brian > > On Friday, August 8, 2003, at 11:17 AM, Robert Wade wrote: > >> I've tried this. In fact, I do have a ranking system that is exactly >> how you described. Every post in the board belongs to a [msg_id] >> (which identifies the thread), and every post also gets a [msg_rank] >> (starting with 0). >> >> I tried a search like the following, hoping that it would work, but >> the problem is that the results are summarized BEFORE they are >> sorted. Therefore, it uses the first match of the msg_ranks within >> that particular msg_id and uses it for the summ. Ideally, it would >> sort the results FIRST, then summarize them using the top result >> dependent upon your sort order. Seems like sort and summ could play >> together a little nicer. >> >> [search >> db=db/ >> messages.db&nemsg_rankdatarq=[blank]&demsg_ranksort=1&msg_ranktype=num >> &msg_idsumm=T] >> >> >> On Friday, August 8, 2003, at 12:29 PM, WebDna @ wrote: >> >>> I have been looking into index searching alot lately. It may not be >>> exactly >>> what you want, but it may be along the right premise to getting what >>> you >>> need. >>> >>> Add an additional field to the DB or make a new DB that is >>> associated with >>> the main topic ID and it will contain a number. Starting out with 0 >>> and then >>> increase by one everytime someone post a reply to it. >>> >>> This would allow you to show the number of replies to a specific post >>> without having to search and sum. >>> >>> Is this possible for you? >>> >>> >>> ----- Original Message ----- >>> From: "Robert Wade"
>>> To: "WebDNA Talk" >>> Sent: Friday, August 08, 2003 10:12 AM >>> Subject: Re: Sorting by highest number of matches unique to a field >>> >>> >>>> Kenneth, Thanks. I just tried a slimmed down search with no sort >>>> specified... >>>> >>>> [search >>>> db=db/messages.db&nemsg_iddatarq=[blank]&msg_idsumm=T&max=10] >>>> [founditems] >>>> ... >>>> [/founditems] >>>> [/search] >>>> >>>> What you are saying, is that this would list the [msg_id] at the top >>>> that had the most matches? It didn't. >>>> >>>> this search seemed to randomly, pick out 10 results. I could not >>>> find >>>> a pattern to the sort at all. It just grabbed 10 of them out of >>>> seemingly anywhere in the db., of those the highest number was 10 >>>> matches. I know that one [msg_id] in particular has over 200 posts, >>>> and there are many that have more than 10. >>>> >>>> I thought that the default sort was "best match" not "most matches" >>>> In >>>> otherwords the results that "best match" the search criteria, not >>>> the >>>> results that have the most matches...? >>>> >>>> -- Confused. >>>> >>>> >>>> On Friday, August 8, 2003, at 11:27 AM, Kenneth Grome wrote: >>>> >>>>>> I was hoping to get a clearer understanding of the possibility of >>>>>> sorting a search by entries that have the most number of matches >>>>>> first. >>>>>> > -- Brian Fries, BrainScan Software -- > http://www.brainscansoftware.com -- > > > ------------------------------------------------------------- > This message is sent to you because you are subscribed to > the mailing list . > To unsubscribe, E-mail to: > To switch to the DIGEST mode, E-mail to > > Web Archive of this list is at: http://webdna.smithmicro.com/ > ------------------------------------------------------------- This message is sent to you because you are subscribed to the mailing list . To unsubscribe, E-mail to: To switch to the DIGEST mode, E-mail to Web Archive of this list is at: http://webdna.smithmicro.com/
Associated Messages, from the most recent to the oldest:
Thanks Brian, I'm leaning toward something like that after having thought about this for awhile. This sounds like it would be something that is so simple to do with a normal search and sort, but it's not. Is it just me, or does it seem like the summary should take place after items are sorted? On Friday, August 8, 2003, at 03:34 PM, Brian Fries wrote: > You could store the total number of messages in a thread in the first > message for that thread (msg_rank=0, if I understand you scheme > correctly) - say a "msg_count" field. When you add a message to the > thread, increment "msg_count" in the first thread message. Then you > could sort with > > [search > db=db/ > messages.db&eqmsg_rankdatarq=0&allhit=1&demsg_countsort=1&msg_counttype > =num] > > No need to summ since you only find the first message per thread. > > - brian > > On Friday, August 8, 2003, at 11:17 AM, Robert Wade wrote: > >> I've tried this. In fact, I do have a ranking system that is exactly >> how you described. Every post in the board belongs to a [msg_id] >> (which identifies the thread), and every post also gets a [msg_rank] >> (starting with 0). >> >> I tried a search like the following, hoping that it would work, but >> the problem is that the results are summarized BEFORE they are >> sorted. Therefore, it uses the first match of the msg_ranks within >> that particular msg_id and uses it for the summ. Ideally, it would >> sort the results FIRST, then summarize them using the top result >> dependent upon your sort order. Seems like sort and summ could play >> together a little nicer. >> >> [search >> db=db/ >> messages.db&nemsg_rankdatarq=[blank]&demsg_ranksort=1&msg_ranktype=num >> &msg_idsumm=T] >> >> >> On Friday, August 8, 2003, at 12:29 PM, WebDna @ wrote: >> >>> I have been looking into index searching alot lately. It may not be >>> exactly >>> what you want, but it may be along the right premise to getting what >>> you >>> need. >>> >>> Add an additional field to the DB or make a new DB that is >>> associated with >>> the main topic ID and it will contain a number. Starting out with 0 >>> and then >>> increase by one everytime someone post a reply to it. >>> >>> This would allow you to show the number of replies to a specific post >>> without having to search and sum. >>> >>> Is this possible for you? >>> >>> >>> ----- Original Message ----- >>> From: "Robert Wade" >>> To: "WebDNA Talk" >>> Sent: Friday, August 08, 2003 10:12 AM >>> Subject: Re: Sorting by highest number of matches unique to a field >>> >>> >>>> Kenneth, Thanks. I just tried a slimmed down search with no sort >>>> specified... >>>> >>>> [search >>>> db=db/messages.db&nemsg_iddatarq=[blank]&msg_idsumm=T&max=10] >>>> [founditems] >>>> ... >>>> [/founditems] >>>> [/search] >>>> >>>> What you are saying, is that this would list the [msg_id] at the top >>>> that had the most matches? It didn't. >>>> >>>> this search seemed to randomly, pick out 10 results. I could not >>>> find >>>> a pattern to the sort at all. It just grabbed 10 of them out of >>>> seemingly anywhere in the db., of those the highest number was 10 >>>> matches. I know that one [msg_id] in particular has over 200 posts, >>>> and there are many that have more than 10. >>>> >>>> I thought that the default sort was "best match" not "most matches" >>>> In >>>> otherwords the results that "best match" the search criteria, not >>>> the >>>> results that have the most matches...? >>>> >>>> -- Confused. >>>> >>>> >>>> On Friday, August 8, 2003, at 11:27 AM, Kenneth Grome wrote: >>>> >>>>>> I was hoping to get a clearer understanding of the possibility of >>>>>> sorting a search by entries that have the most number of matches >>>>>> first. >>>>>> > -- Brian Fries, BrainScan Software -- > http://www.brainscansoftware.com -- > > > ------------------------------------------------------------- > This message is sent to you because you are subscribed to > the mailing list . > To unsubscribe, E-mail to: > To switch to the DIGEST mode, E-mail to > > Web Archive of this list is at: http://webdna.smithmicro.com/ > ------------------------------------------------------------- This message is sent to you because you are subscribed to the mailing list . To unsubscribe, E-mail to: To switch to the DIGEST mode, E-mail to Web Archive of this list is at: http://webdna.smithmicro.com/
Robert Wade
DOWNLOAD WEBDNA NOW!
Top Articles:
Talk List
The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...
Related Readings:
Converting characters to numbers ... (1997)
Sendmail question (2002)
Which GUI HTML editors work with WC ? (1997)
Purchased cart being overwritten - still !?? (1997)
can WC render sites out? (1997)
[WriteFile] problems (1997)
[WebDNA] Video Request: "Faces of WebDNA" (2010)
Ultimate BB (2001)
add to cart within a page? (1997)
Generating Report Totals (1997)
RE: WebCat name recognition (was MacFinder -- a new WebDNAweb site) (1998)
[OT] Half completed orders (2005)
Setting up shop (1997)
Writing [raw] to a file (2000)
Nested tags count question (1997)
Cart Number sequence (1997)
PARAMETER vs. OPERATOR (was The BUG is BACK ...) (1998)
Security problems (2000)
[OT] USA States (2003)
Help with WebCatalog (1998)