[sqlite] Clarification of ON CONFLICT in trigger bodies?

2011-04-29 Thread Nico Williams
Long ago I noticed that the lang_createtrigger page says: "An ON CONFLICT clause may be specified as part of an UPDATE or INSERT action within the body of the trigger. However if an ON CONFLICT clause is specified as part of the statement causing the trigger to fire, then conflict handling policy

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Jay A. Kreibich
On Fri, Apr 29, 2011 at 04:12:13PM +, Joseph Stockman scratched on the wall: > It never reaches the xOpen call. In your xCreate() and xConnect() functions, make sure you're setting the sqlite3_vtab "zErrMsg" field to NULL. Other than that, run it in the debugger, dump the stack, and sta

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Dan Kennedy
On 04/29/2011 11:15 PM, Joseph Stockman wrote: > > We have tried a null pointer, a zero-length static, a dummy static and > strings created on the heap via sqlite3_mprintf. > All cause the same segmentation fault. Run your app under valgrind. Post the errors it finds if they don't make the proble

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Joseph Stockman
We have tried a null pointer, a zero-length static, a dummy static and strings created on the heap via sqlite3_mprintf. All cause the same segmentation fault. V. From: sqlite-users-boun...@sqlite.org [sqlite-users-boun...@sqlite.org] on behalf of Drake

Re: [sqlite] vacuum and rowids

2011-04-29 Thread Stephan Beal
On Fri, Apr 29, 2011 at 3:37 AM, Simon Slavin wrote: > The thing is, there's not problem with referring to rowid, or ROWID or any > of the other aliases when you do this. The only problem is possible > confusion for the programmer if you define a column with one of these names > which /isn't/ an

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Joseph Stockman
It never reaches the xOpen call. V. From: sqlite-users-boun...@sqlite.org [sqlite-users-boun...@sqlite.org] on behalf of Jay A. Kreibich [j...@kreibi.ch] Sent: Friday, April 29, 2011 10:47 AM To: General Discussion of SQLite Database Subject: Re: [sqlite]

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Jay A. Kreibich
On Fri, Apr 29, 2011 at 03:14:07PM +, Joseph Stockman scratched on the wall: > My apologies. > > We set idxNum to zero, idxStr to NULL, needToFreeIdxStr to zero, > orderbyConsumed to zero and estimatedCost to 1000. The sqlite3_index_info struct is zero-ed out before it is passed to your f

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Drake Wilson
Quoth Joseph Stockman , on 2011-04-29 15:14:07 +: > My apologies. No worries. > We set idxNum to zero, idxStr to NULL, needToFreeIdxStr to zero, > orderbyConsumed to zero and estimatedCost to 1000. But that still looks a bit wonky; I don't see anywhere in the virtual table docs that idxStr

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Joseph Stockman
My apologies. We set idxNum to zero, idxStr to NULL, needToFreeIdxStr to zero, orderbyConsumed to zero and estimatedCost to 1000. We never get to our xFilter method; sqlite throws the segmentation fault internally trying to process the result of the xBestIndex call. V.

Re: [sqlite] select * on empty virtual table

2011-04-29 Thread Drake Wilson
Quoth Joseph Stockman , on 2011-04-29 14:50:51 +: > Our trace shows that we enter our xBestIndex method, see that > both nConstraints and nOrderBy are zero, so we return without > modifying any of the structures. So essentially you're returning all undefined values? That doesn't seem like a

Re: [sqlite] vacuum and rowids

2011-04-29 Thread Nico Williams
On Thu, Apr 28, 2011 at 5:20 PM, Dave Hayden wrote: > On Apr 28, 2011, at 12:41 PM, Pavel Ivanov wrote: >>> After more poking, it appears that rowids might not be changed by a vacuum >>> if I have an index on the table. Is this true? If so, is it something I can >>> rely on going forward? >> >>

[sqlite] select * on empty virtual table

2011-04-29 Thread Joseph Stockman
I have looked for a reason or solution to this problem but can't seem to find one anywhere. We are running sqlite in a cross-platform environment (Windows, Linux and an embedded system). This error appears across the board. We are implementing our first set of virtual tables. After xCreate is c

Re: [sqlite] Latest ADO.NET driver does not install properly?

2011-04-29 Thread Mark N. Steele
Greetings, It sounds like you are running on a 64 bit implementation and visual studio looks for a 32 it implementation. Check which ODBC GUI sees your SQLlite ODBC driver Mark -Original Message- From: sqlite-users-boun...@sqlite.org [mailto:sqlite-users-boun...@sqlite.org] On Behalf O

[sqlite] Latest ADO.NET driver does not install properly?

2011-04-29 Thread Kostiainen Tom
Greetings all! I'm first starting to familiarize with SQLite. I first installed the latest sqlite-dotnet-x86-1007100.exe package from the sqlite.org site, but failed to get a connection working. First I noticed the Server Explorer did not list the driver for SQLite (my machine.config was unchan

Re: [sqlite] (UML) Visualisation of a SQLite schema ?

2011-04-29 Thread Fredrik Karlsson
Hi, Just thought I'd tell you that I went into solving this issue myself. The resulting script is here: https://github.com/dargosch/sqlite2dot * Requires Tcl and the sqlite3 package for Tcl. It generates a .dot file for use with Graphviz. Very simple script, but it solves the issue I wanted to