Re: [sqlite] SQLite JDBC driver performance
Christopher Masonwrites: > [Apologies for posting this here, but the sqlitejdbc list appears to be > defunct.] Google deleted it without warning or explanation: http://www.zentus.com/sqlitejdbc/google-group-disappearance.html It has been relocated here: https://lists.hcoop.net/listinfo/sqlitejdbc - a ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Re: [sqlite] Error message from RAISE just plain text ?
On 20 Jun 2009, at 5:21pm, Dennis Cote wrote: > From the parser you can see the allowed raise function syntax. It honestly didn't occur to me to read the source code. Thanks, Dennis. Simon. ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Re: [sqlite] Error message from RAISE just plain text ?
Simon Slavin wrote: > On 15 Jun 2009, at 2:37am, Simon Slavin wrote: > > >> The examples for the error text I've found are all simple text >> strings, for instance >> >> RAISE(ROLLBACK, 'delete on table "foo" violates foreign key constraint >> "fk_foo_id"') >> >> What I want to do is more like >> >> RAISE(ROLLBACK, 'Attempt to delete author '||old.name||' who has >> books.') >> >> but that doesn't work. >> > > No responses to this ? Can someone confirm for me that error messages > from 'RAISE' have to be just pre-assigned text strings and can't be > expressions ? > > Simon. > Simon, From the parser you can see the allowed raise function syntax. expr ::= RAISE LP raisetype COMMA nm RP Where nm is the error message. The nm nonterminal can be replaced by the following: nm ::= id nm ::= STRING nm ::= JOIN_KW and the id can bereplaced by; id ::= ID id ::= INDEXED So you can see that the parser is basically designed to only accept a STRING at that position, not an expression which is what would be required to allow concatenation and other function calls. I assume that the other possible values (join keyword etc) would be caught and reported as errors. HTH Dennis Cote ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Re: [sqlite] Error message from RAISE just plain text ?
On 15 Jun 2009, at 2:37am, Simon Slavin wrote: > The examples for the error text I've found are all simple text > strings, for instance > > RAISE(ROLLBACK, 'delete on table "foo" violates foreign key constraint > "fk_foo_id"') > > What I want to do is more like > > RAISE(ROLLBACK, 'Attempt to delete author '||old.name||' who has > books.') > > but that doesn't work. No responses to this ? Can someone confirm for me that error messages from 'RAISE' have to be just pre-assigned text strings and can't be expressions ? Simon. ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Re: [sqlite] async io and locks
Async I/O fits perfectly with my app: I don't need the durable guarantee. But I do need the ability to activate it with a pragma since it isn't a C app. Are there plans to make async I/O available via pragma? Jim On 6/20/09, Danwrote: > > I think we have quite different approaches. > > The SQLite asynchronous IO backend is queueing IO, not queries. It > intercepts > the equivalent of the write() system call. It does not overload query > processing > in any way. The things on the write-queue are blobs of data to be > written > to specific offsets within the database file, not SQL queries. > > It's more complex than that of course. But the point is that the write- > queue is > a layer between SQLite and the OS, not between the user and SQLite. > > Dan. > > > ___ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > -- Software first. Software lasts! ___ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users