[sqlite] Sqlite & detect change in column

2014-07-07 Thread Micka
Hi, I know that there is a callback that can be called when a column is updated. But it doesn't give a bitmap of the columns changed /* ** Register a callback to be invoked each time a row is updated, ** inserted or deleted using this database connection. */ SQLITE_API void *sqlite3_update_hook(

Re: [sqlite] Sqlite in dead lock state when deleting records from the same table from different threads

2014-07-07 Thread mm.w
your sunglasses won't stop my gaze On Mon, Jul 7, 2014 at 4:21 PM, mm.w <0xcafef...@gmail.com> wrote: > comply to my request or stop, sorry but truth must be told at some point. > > > On Mon, Jul 7, 2014 at 6:02 AM, Srikanth Bemineni < > bemineni.srika...@gmail.com> wrote: > >> Hi, >> >> How can

Re: [sqlite] Sqlite in dead lock state when deleting records from the same table from different threads

2014-07-07 Thread mm.w
comply to my request or stop, sorry but truth must be told at some point. On Mon, Jul 7, 2014 at 6:02 AM, Srikanth Bemineni < bemineni.srika...@gmail.com> wrote: > Hi, > > How can I find this in the statement ? > > Will BEGIN immediate get an exclusive lock.? or like Igor specified if I > call "

Re: [sqlite] Basic SQLite/EF6 question

2014-07-07 Thread Joe Mistachkin
William Drago wrote: > > Yes, and I can see my tables, etc. too. > Ok. One thing to try is to GAC the System.Data.SQLite.EF6 assembly (which may require GAC'ing the Entity Framework 6 assembly as well). Per the rules for distributing Entity Framework, we cannot GAC it during the installation.

Re: [sqlite] Building for vxWorks

2014-07-07 Thread Jan Nijtmans
2014-07-07 14:44 GMT+02:00 Andy Ling : > Removing the SQLITE_ENABLE_LOCKING_STYLE I get ... > sqlite3.c: In function 'fileHasMoved': > sqlite3.c:25236: error: 'struct unixFileId' has no member named 'ino' > sqlite3.c: In function 'semCheckReservedLock': ... > sqlite3.c:26372: warning: unuse

Re: [sqlite] Autocommit in "with" query: bug or feature?

2014-07-07 Thread big stone
Curiousity question : Why are you basing your SQLfast next book on Python 2.7.6 rather than Python 3.3 (or 3.4) ? ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] sqlite & INNER JOIN & vtable

2014-07-07 Thread Hick Gunter
The type of join is unaffected by the type of table (native or virtual). Pretend all tables are native SQLite tables. You must return correct results from your xBestIndex function for the cost-based optimiser to select an efficient plan. -Ursprüngliche Nachricht- Von: Micka [mailto:mick

Re: [sqlite] sqlite & INNER JOIN & vtable

2014-07-07 Thread Pasha
Virtual tables are just like any other tables. The kind of join is the matter of you expectations from the query outcome. > Hi, > > In my case I've a normal sqlite table : toto > and an vtable : special > > This is the typical request that I have : > > SELECT toto.id, toto.name, special.info FRO

[sqlite] sqlite & INNER JOIN & vtable

2014-07-07 Thread Micka
Hi, In my case I've a normal sqlite table : toto and an vtable : special This is the typical request that I have : SELECT toto.id, toto.name, special.info FROM toto INNER JOIN special ON ( special.id = toto.specialid AND toto.test == TRUE ) I don't know if I should use LEFT JOIN, RIGHT JOIN or

Re: [sqlite] Sqlite in dead lock state when deleting records from the same table from different threads

2014-07-07 Thread Srikanth Bemineni
Hi, How can I find this in the statement ? Will BEGIN immediate get an exclusive lock.? or like Igor specified if I call "Delete * from where 0" will it be able to get an immediate lock on the table. Srikanth On Sun, Jul 6, 2014 at 9:21 PM, mm.w <0xcafef...@gmail.com> wrote: > what's the sys

Re: [sqlite] Building for vxWorks

2014-07-07 Thread Andy Ling
> -Original Message- > From: sqlite-users-boun...@sqlite.org [mailto:sqlite-users- > boun...@sqlite.org] On Behalf Of Jan Nijtmans > Sent: 07 July 2014 12:24 > To: General Discussion of SQLite Database > Subject: Re: [sqlite] Building for vxWorks > > 2014-07-04 11:52 GMT+02:00 Andy Ling :

Re: [sqlite] Building for vxWorks

2014-07-07 Thread Jan Nijtmans
2014-07-04 11:52 GMT+02:00 Andy Ling : > I now have sqlite running under vxWorks. As nobody offered any suggestions, I > thought I would share my experience in the hope it will help others and maybe > feed some changes back into the main code. > semaphone.h is included if SQLITE_ENABLE_LOCKI

Re: [sqlite] Error xBestIndex returned an invalid plan

2014-07-07 Thread Hick Gunter
The values on the RHS of constraints are generally not known until (bound values) or after (joins) the query is executed. At that point, having a better estimate is moot. -Ursprüngliche Nachricht- Von: Dominique Devienne [mailto:ddevie...@gmail.com] Gesendet: Montag, 07. Juli 2014 10:39

Re: [sqlite] Error xBestIndex returned an invalid plan

2014-07-07 Thread Dominique Devienne
On Mon, Jul 7, 2014 at 10:25 AM, Hick Gunter wrote: > "fred" is only known if your select statement references only that one table > and the value is supplied as a literal. Known to whom? AFAIK, even in that case, http://www.sqlite.org/vtab.html#xbestindex does not know about 'fred'. Which fiel

Re: [sqlite] Error xBestIndex returned an invalid plan

2014-07-07 Thread Hick Gunter
"fred" is only known if your select statement references only that one table and the value is supplied as a literal. If the value is a bound variable or part of an (explicit or implicit) join expression the value(s) (there may be more than one) will not be known until well after the sqlite3_pre

Re: [sqlite] Can't make SQLite work

2014-07-07 Thread Simon Slavin
On 7 Jul 2014, at 4:21am, mm.w <0xcafef...@gmail.com> wrote: > as we start low a more simple solution if recent features are needed Nothing to do with MacPorts is simple. If you've messed up your shell path and 'which sqlite3' doesn't give an answer, just download the executable from the SQLit

Re: [sqlite] Autocommit in "with" query: bug or feature?

2014-07-07 Thread Simon Slavin
On 7 Jul 2014, at 6:49am, big stone wrote: > If I understood well, the "autocommit" idea was to not "hold" the database > file by default, and ease multi-user access to the file database. Not executing "BEGIN" is also the most frequent error that SQL programmers make. Just like even the most