Tuesday, March 11, 2008, 12:34:02 PM, you wrote: JS> Neville Franks wrote: >> Hi John, >> >> Tuesday, March 11, 2008, 11:51:57 AM, you wrote: >> >> JS> Neville Franks wrote: >> >>>>Hi John, >>>> >>>>Tuesday, March 11, 2008, 10:18:30 AM, you wrote: >>>> >>>>JS> Neville Franks wrote: >>>> >>>> >>>>>>Tuesday, March 11, 2008, 8:48:05 AM, you wrote: >>>>>> >>>>>>JAK> On Tue, Mar 11, 2008 at 08:37:27AM +1100, Neville >>>>>>Franks scratched on the wall: >>>>>> >>>>>> >>>>>> >>>>>>>>If I use sqlite3_get_table() and update/insert/delete one of the tables >>>>>>>>in the original query, does the results set get updated. ie. If I >>>>>>>>iterate through the original results from sqlite3_get_table() will I >>>>>>>>see the changes, or do I need to run the sqlite3_get_table() query >>>>>>>>anew? >>>>>>>> >>>>>>>>If I do see the changes does this also apply to an sqlite3_get_table() >>>>>>>>query that was on a VIEW? >>>>>> >>>>>> >>>>>>JAK> From <http://www.sqlite.org/c3ref/free_table.html>: >>>>>> >>>>>>JAK> The sqlite3_get_table() interface is implemented as a wrapper >>>>>>around >>>>>>JAK> sqlite3_exec(). The sqlite3_get_table() routine does not have >>>>>>access >>>>>>JAK> to any internal data structures of SQLite. It uses only the >>>>>>public >>>>>>JAK> interface defined here. As a consequence, errors that occur in >>>>>>the >>>>>>JAK> wrapper layer outside of the internal sqlite3_exec() call are not >>>>>>JAK> reflected in subsequent calls to sqlite3_errcode() or >>>>>>JAK> sqlite3_errmsg(). >>>>>> >>>>>>JAK> In other words, "No." You need to run it again. >>>>>> >>>>>>JAK> -j >>>>>> >>>>>>Jay, >>>>>>Thanks, I assumed that would be the case. Dynamically updating GUI's >>>>>>on SQL DB updates appears to be challenging. >>>>>> >>>> >>>>JS> Are you using Windows and the WIN API? If so you can just set up >>>>JS> callbacks to do it. >>>> >>>>I am using C++ and Windows. Do you mean use "triggers"? >>>> >> >> JS> No. If you have the data in some form of windows control like a >> JS> listview when you get a notify messge telling you it is changed you can >> JS> fire a callback to perform an Sqlite update. It is a bit tedious to >> JS> program. It is quite a while since I wrote such a program so I can only >> JS> give you a big picture. >> >> JS> By making the edit phase a transaction you can give the user the option >> JS> of commiting changes or rolling back. >> >> >> Thanks, but that isn't what I'm trying to do. >> >> I'm displaying tree's which are built from SQL queries and I ideally >> want the tree control to reflect updates to the underlying tables >> without have to do queries all over again and rebuild the trees from >> scratch. >> JS> In that case a trigger activating a user function which updates the tree JS> would be a way of doing that. Depends on number of user, processes etc. JS> We use an XML output from the DB to do that is a very ditributed way.
I'm creating XML as well. I've written my own tree control that interfaces to a data source, without knowing nor caring what the data source is. It doesn't store any data itself. I was hoping I could use SQLite as a data source, but that won't work. I'll start by recreating the XML whenever the underlying SQL changes and if that proves too slow, I'll look at in situ updates to the XML DOM. There can be 10's of thousands of nodes in the trees. I'll have to run some tests with large trees to determine how long it takes to create same. -- Best regards, Neville Franks, http://www.surfulater.com http://blog.surfulater.com _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
