Smith, Randall wrote:
> Is it impossible to have references from temp tables to main tables?
Yes.
> If so, aren't TEMPORARY tables largely useless?
Only if you want to use foreign key constraints.
> Is there another idiom in SQLite for managing tables that are intended
> to have a short life or
Simon et al,
On Sat, Jul 23, 2016 at 2:20 AM, Simon Slavin wrote:
>
> On 23 Jul 2016, at 4:26am, Igor Korot wrote:
>
>> "Ambiguous expansion of macro MIN/MAX"
>
> Sorry, I don't recognise this problem. I would try Googling it.
Trying to google I found this:
http://stackoverflow.com/questions/2
On 2016/07/23 3:06 AM, Smith, Randall wrote:
Hi.
I'm creating a specialized index table to a persistent table that I need to
speed up a one-time operation. This seems like a great application for a
TEMPORARY table in SQLite, so the index will always be reliably cleaned up.
However, I can't
On 23 Jul 2016, at 2:24pm, R Smith wrote:
> No, it's not possible to forge permanent references to/from temporary tables
> - it defies the objective of being "temporary".
The other way around should fine, though: use a permanent table to ensure that
only legitimate values appear in a temporar
On 2016/07/23 3:30 PM, Simon Slavin wrote:
On 23 Jul 2016, at 2:24pm, R Smith wrote:
No, it's not possible to forge permanent references to/from temporary tables - it defies
the objective of being "temporary".
The other way around should fine, though: use a permanent table to ensure that
On 07/23/2016 08:49 PM, R Smith wrote:
On 2016/07/23 3:30 PM, Simon Slavin wrote:
On 23 Jul 2016, at 2:24pm, R Smith wrote:
No, it's not possible to forge permanent references to/from
temporary tables - it defies the objective of being "temporary".
The other way around should fine, though:
You can maintain a local conditional "index" table via temp trigger, as long as
there's no other processes modify the permanent table.
Create table t(id integer primary key, x);
Create temp table t1(id integer primary key);
Create temp trigger trig_1 after insert on t begin
Insert into t1 select
The next release of SQLite will be the "pi" release - version 3.14.
It will probably occur within the next two weeks.
Draft change log: https://www.sqlite.org/draft/releaselog/3_14_0.html
Code snapshot:
https://www.sqlite.org/snapshot/sqlite-snapshot-201607230522.tar.gz
Testing and (especially
> On 23 Jul 2016, at 3:33pm, Dan Kennedy wrote:
>
>> On 2016/07/23 3:30 PM, Simon Slavin wrote:
>>
>>> The other way around should fine, though: use a permanent table to ensure
>>> that only legitimate values appear in a temporary table.
>
> The trouble is that some other process may come alo
On 23 Jul 2016, at 4:16pm, Richard Hipp wrote:
> Testing and (especially) documentation work is on-going.
If there isn't already done, would it be possible to add an option to dbhash
which produces just the hash and not the path ? That would make it easier for
a shell script to compare two h
Mr. Hipp,
On Sat, Jul 23, 2016 at 11:16 AM, Richard Hipp wrote:
> The next release of SQLite will be the "pi" release - version 3.14.
> It will probably occur within the next two weeks.
>
> Draft change log: https://www.sqlite.org/draft/releaselog/3_14_0.html
>
> Code snapshot:
> https://www.sq
Is it too late to add a request?
There's been many questions and discussions around people not knowing
which Foreign Key constraint failure caused their transaction to abort
(and we have been over why that is not feasible), but perhaps the new
Trace hook is a great way to add an SQLITE_TRACE t
Dears,
(Writing from mobile, hope the HTML like formatting, which I can't seem to
override, won't be a total mess in this post).
Have not even yet got to the code to test it and review the new C API but I
already have a question about:
Added two new C-language interfaces: sqlite3_expanded_sql()
I created a function UUID() with sqlite3_create_function. There are
tables in my database which have default columns using the UUID
function. Do I run into trouble when I insert a row into a table with a
given UUID value without the defined function (Using the command line
interface)?
_
On 7/23/16, kna...@online.de wrote:
> I created a function UUID() with sqlite3_create_function. There are
> tables in my database which have default columns using the UUID
> function. Do I run into trouble when I insert a row into a table with a
> given UUID value without the defined function (Usi
Dears,
The sqlite3_trace() has been deprecated in favor of sqlite3_trace_v2().
Some '#ifndef SQLITE_OMIT_DEPRECATED' have been added (main.c for instance) and
shell.c has been updated to use the newer API. But some other files are
missing the #ifndef protection, leading to some issues trying
On 2016-07-23 8:16 AM, Richard Hipp wrote:
The next release of SQLite will be the "pi" release - version 3.14.
It will probably occur within the next two weeks.
Draft change log: https://www.sqlite.org/draft/releaselog/3_14_0.html
Code snapshot:
https://www.sqlite.org/snapshot/sqlite-snapshot
One thing I'd love to see addressed is an issue I reported back in May:
sqlite3_auto_extension + custom FTS5 tokenizer
http://thread.gmane.org/gmane.comp.db.sqlite.general/101685/focus=101697
http://mailinglists.sqlite.org/cgi-bin/mailman/private/sqlite-users/2016-May/066676.html
The workaround
18 matches
Mail list logo