"P Kishor" wrote...
> On 2/8/08, jose isaias cabrera <[EMAIL PROTECTED]> wrote: >> >> "P Kishor" replied... >> >> >> > On 2/7/08, jose isaias cabrera <[EMAIL PROTECTED]> wrote: >> >> >> >> "P Kishor" asked... >> >> >> >> >> >> >I don't understand your question at all. >> >> >> >> I will answer it at the end. >> >> >> >> > On 2/7/08, jose isaias cabrera <[EMAIL PROTECTED]> wrote: >> >> >> >> >> >> Greetings. >> >> >> >> >> >> I would like to have the results of a select be returned sorted in >> >> >> an >> >> >> specific way. Let me show you what I mean: >> >> >> >> >> >> sqlite> SELECT PSubClass FROM LSOpenJobs WHERE subProjID = 2190 >> >> >> GROUP >> >> >> BY >> >> > >> >> > But, first... why are you using GROUP BY above when you have no >> >> > aggregate function (Count, Max, Min, Sum, etc.) in your query? >> >> >> >> Because of this: >> >> >> >> sqlite> SELECT PSubClass FROM LSOpenJobs WHERE subProjID = 2190; >> >> Pre-Proc >> >> Post-Proc >> >> DOC-Trans >> >> DTP >> >> Pre-Proc >> >> Pre-Proc >> >> Pre-Proc >> >> Pre-Proc >> >> Pre-Proc >> >> Post-Proc >> >> Post-Proc >> >> Post-Proc >> >> Post-Proc >> >> Post-Proc >> >> DOC-Trans >> >> DOC-Trans >> >> DOC-Trans >> >> DOC-Trans >> >> DOC-Trans >> >> DTP >> >> DTP >> >> DTP >> >> DTP >> >> DTP >> >> PM >> >> PM >> >> PM >> >> PM >> >> PM >> >> PM >> >> sqlite> >> > >> > In that case you need SELECT DISTINCT, not GROUP BY >> >> Oh, ok, thanks for the help. Weird how the ORDER BY works... I know >> what >> each of these (ORDER BY and DISTINT) do, however is the usage of one of >> these faster than the other? >> >> > >> > >> >> >> >> >> >> > >> >> >> PSubClass; >> >> >> DOC-Trans >> >> >> DTP >> >> >> PM >> >> >> Post-Proc >> >> >> Pre-Proc >> >> >> sqlite> >> >> >> >> >> >> What I would like is to have the SELECT result be, >> >> >> >> >> >> Pre-Proc >> >> >> Post-Proc >> >> >> DOC-Trans >> >> >> DTP >> >> >> PM >> >> >> >> >> >> is this possible? Yes, I know I can sort it in the program, but >> >> >> how >> >> >> can >> >> >> I >> >> >> get this special sort from the DB? >> >> >> >> >> > >> >> > What is the logic above? Without you telling us why Pre-Doc is >> >> > before >> >> > Post-Proc is before DOC-Trans, and so on, how can we guess what your >> >> > logic is for the sort? In other words, please explain what makes the >> >> > above sort "special"? >> >> >> >> Aaah, ok. I am preparing an html file and in it I will list these >> >> PSubClasses in this special logical display sequence: >> >> 1. Pre-Proc >> >> 2. Post-Proc >> >> 3. DOC-Trans >> >> 4. DTP >> >> 5. PM >> >> >> >> I hope this explains it... >> >> >> > >> > Nope, it doesn't at all. You logic is not clear to me, and even if it >> > is, it won't be clear to the poor program that you want to write. If >> > you want a specific order which is not dependent on a collation >> > sequence, alpha order, etc., then enter your prescribed ordinal-ity in >> > a separate column and ORDER BY that column. >> >> Ok Punkish, no problem. Thanks for trying. You have helped me many >> times >> and I am thankful. Sometimes I am very hard to understand. However, >> other >> folks in the list were able to understand what I meant and they have >> provided a few solutions for me. Thanks for the help. And the poor >> program >> understands me correctly and it is working as desired .-). It is a good >> thing computers understand me. Which I can not say the same for >> people... >> .-) > > > Hi José, I apologize if my email came off as brusque and rude > sounding. That was not the intent at all. I really mean what I say > above, and what I suggest above is exactly what has been suggested by > others you have replied to you as well. The logic that may be very > clear in our brain has to be explicit for the program. Using ORDER BY > depends on alphabetical sort or a numeric sort. If you want something > else (for example, importance of a job title), you have to convert > that hierarchy into something that the ORDER BY mechanism can > understand. But, you have to do that explicitly. > > Remember, there is no inherent concept of order in an rdbms. Each row > is independent of each other. Order is imposed *after* your result set > has been extracted, and it only changes the presentation of the result > set, not the original data. There is no inherent concept of "previous > row" or "next row" unlike a spreadsheet. Puneet, once again, you hit the spot. You are a good teacher. Thanks. You are correct. I should have been more specific and I being under time constraint I did not realized my vagueness, and for that I apologize. I tried googleling how to sort in a "special" way with sqlite, but was not successful. Don't worry about the brusque and rude sounding. Knowing you, through your post, I thought you were joking with me. .-) Thanks again. josé _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
