if you create a table use following statement (script generated from MS SQL
Server 2000)
CREATE TABLE [XTollData] (
[DutyID] [char] (32) NOT NULL ,
[CarNumber] [char] (10) NULL
);
SQLite3_Column_decltype will treat DutyID as data type 'char' but not 'char(32)'
On Tue, 2007-03-20 at 02:44 +0200, Rafi Cohen wrote:
> Hi, I'm calling system from a C application in order to invoke sqlite3
> database, then continue with commands: .separator "," and .import file
> table, all in a single string as argument to system.
> sqlite3 is indeed invoked with the correct
Hello:
I recently upgraded to the latest version of SQLite 3.13 and ever since
then I have been experiencing numerous SQLITE_MISUSE[21] error message.
But when I run the same code under Win32 or the PocketPC Emulator the
code works fine.
Now previously I was using SQLite 3.08 or 3.09 and
Hi RBS,
I use SQLite as a data manipulator, not as a database. I get data
from a
server database, dump to SQLite, manipulate the data and finally
dump to
Excel. As this is reporting software speed is important, so I will
go with
the fastest method.
OK, I have to ask. What do you then do
Hi, I'm calling system from a C application in order to invoke sqlite3
database, then continue with commands: .separator "," and .import file
table, all in a single string as argument to system.
sqlite3 is indeed invoked with the correct database, but the problem is
that the .import command is
Hi there,
I want to see if there is anyone using sqlite-3.3.3 with perl-DBD-SQLite
in a mod_perl environment where multiple threads will be present.
The FAQ indicates the default sqlite3 does not have multi-threading
option enabled. But I want to stick with the system library as much as
On Mon, 2007-03-19 at 01:46 +0200, Dimitris Servis wrote:
> 2007/3/19, guenther <[EMAIL PROTECTED]>:
> > On Sun, 2007-03-18 at 23:51 +0200, Dimitris Servis wrote:
> > > in my wildest dreams... if you read carefully, *each* file is about
> > > 100-200MB. I now end up wit ha collection of 100-200
It should save some time. How much is questionable.
Why would sqlite have to bind the Pointer bound variables? Isn't the strategy
of binding to associate a variable with a statment? Why should I have to
continually re-associate the bindings with a statement thats allready been
prepared and
ken-33 <[EMAIL PROTECTED]> wrote:
> Anyone thoughts?
>
I would rather not complicate the existing API
unnecessarily by add bells and whistles that can
be easily implemented using simple by wrappers.
>
> ken-33 wrote:
> >
> > Question for the list,
> >
> > I'd like to optimize my code,
Anyone thoughts?
ken-33 wrote:
>
> Question for the list,
>
> I'd like to optimize my code, using the following pseudo code as an
> example.
>
> ===
> int i = 0 ;
> char str[20];
>
> sqlite3_prepare_v2( "insert into t1 values (?,?)" )
>
I don't see how your modified version is any better than just putting
the sqlite3_bind_int() inside the loop. You've superficially lifted
some code out of the loop, but sqlite3_step() is going to have to go
through and bind all of the "pointer bound" variables in your
suggested API, so it won't
Hubertus wrote:
Well now, if this isn't something!
fist I wann thank you for your quick reply. Finally I have to wait for my result
satisfying 0.6 sec. This is great. Now I can add lots more data.
What I did:
- I added a second index for the column "campId" and did the analyze trick
(I had
Anyone
Ken <[EMAIL PROTECTED]> wrote: Question for the list,
I'd like to optimize my code, using the following pseudo code as an example.
===
int i = 0 ;
char str[20];
sqlite3_prepare_v2( "insert into t1 values (?,?)" )
sqlite3_bind_int ( i )
> Would it also be acceptable (assuming you want the protection offered
> by the temporary tables) to do ... "PRAGMA temp_store = MEMORY" on
> the database?
>
The temp_store pragma only effects the placement of temporary
tables. But the rollback journal is not a table. The
temp_store pragma
[EMAIL PROTECTED] wrote:
Siebe Warners <[EMAIL PROTECTED]> wrote:
Hi all,
I'm using sqlite on an embedded system with flash EEPROM as the medium.
Performance of the flash is not too good, so i'm trying to avoid file
system access where possible without running risk of database corruption.
"Rich Rattanni" <[EMAIL PROTECTED]> wrote:
> On 3/19/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> > Siebe Warners <[EMAIL PROTECTED]> wrote:
> > >
> > > When i perform a sequence of commands like:
> > > BEGIN;
> > > UPDATE t SET col="val1" WHERE key="key1";
> > > .
> > > UPDATE t SET
On 3/19/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
Siebe Warners <[EMAIL PROTECTED]> wrote:
> Hi all,
>
> I'm using sqlite on an embedded system with flash EEPROM as the medium.
> Performance of the flash is not too good, so i'm trying to avoid file
> system access where possible without
Hubertus wrote:
Dear John,
You might also look at using the Sqlite date format rather than seperate
columns for year, month etc.
That was what I considered first. The problem was, that other people
are also supposed to use this datbase. Some of them use Fortran and they
said that it's easier
Did you ever determine the cardinality of the campID field? I'm guessing its
pretty good since your query is now .6 seconds.
Lets say your cardinality was low, ie say less than .3 (arbitrary number).
Then using the index to perform you data lookups would probably be slower than
just reading
Dear John,
> You might also look at using the Sqlite date format rather than seperate
> columns for year, month etc.
That was what I considered first. The problem was, that other people
are also supposed to use this datbase. Some of them use Fortran and they
said that it's easier to compile the
Well now, if this isn't something!
fist I wann thank you for your quick reply. Finally I have to wait for my result
satisfying 0.6 sec. This is great. Now I can add lots more data.
What I did:
- I added a second index for the column "campId" and did the analyze trick
(I had this column in the
On 3/19/07, Hubertus <[EMAIL PROTECTED]> wrote:
Dear list,
sorry to just come up with another performance question. I build a yet small
database with one table. It has about 650.000 rows, 75 columns and
has at the moment about 650 Mb. It runs on a Intel Pentium M with 2 GHz. The
Laptop runs Suse
First, you have your double and single quotes mixed up. SQL uses single
quotes for literals.
Second, you are performing row scans and not using any of your indices.
You will do better if you have an index on the column you specify in
your search.
You might also look at using the Sqlite
Siebe Warners <[EMAIL PROTECTED]> wrote:
> Hi all,
>
> I'm using sqlite on an embedded system with flash EEPROM as the medium.
> Performance of the flash is not too good, so i'm trying to avoid file
> system access where possible without running risk of database corruption.
> So the database
I'll give 3.3.13 a try but it will be for debugging only for now. The
project using the database is in the stage of avoiding all risk, so i'm
not sure i will be able to convince the people to swap databases at this
point.
So were there any updates wrt transactions and temporary files?
Siebe Warners <[EMAIL PROTECTED]> wrote:
> I'm using sqlite 3.2.7.
Have you tried 3.3.13?
--
D. Richard Hipp <[EMAIL PROTECTED]>
-
To unsubscribe, send email to [EMAIL PROTECTED]
Hi all,
I'm using sqlite on an embedded system with flash EEPROM as the medium.
Performance of the flash is not too good, so i'm trying to avoid file
system access where possible without running risk of database corruption.
So the database file and the journal files need to be written to the
Hubertus wrote:
This is the schema:
CREATE TABLE 'data'('nr' INTEGER PRIMARY KEY,
'year' INTEGER,
'month' INTEGER,
'day' INTEGER,
'sec' REAL,
'campId' TEXT,
'flightNr' INTEGER,
'1' REAL,
...
'71' REAL
);
You should use double quotes around column and table names. Single
Hubertus uttered:
Dear list,
sorry to just come up with another performance question. I build a yet small
database with one table. It has about 650.000 rows, 75 columns and
has at the moment about 650 Mb. It runs on a Intel Pentium M with 2 GHz. The
Laptop runs Suse 10.2 and does basicly
Looks like it is going to do a full scan of the entire database to complete
that querry based upon your where clause.
Are you always accessing the data by campID? What is the cardinality of
campId data?
Depending upon that it might be worth while putting and index on CampID.
Hubertus
Dear list,
sorry to just come up with another performance question. I build a yet small
database with one table. It has about 650.000 rows, 75 columns and
has at the moment about 650 Mb. It runs on a Intel Pentium M with 2 GHz. The
Laptop runs Suse 10.2 and does basicly nothing but this
Rafi Cohen <[EMAIL PROTECTED]> wrote:
1. When I prepare a select statement for later execution, may I
use a question mark instead of a table and later "bind" different
table names
No. You can only use a parameter where an expression would be valid.
2. In "order by" clause may I use a column
Hi, 1. When I prepare a select statement for later execution, may I use
a question mark instead of a table and later "bind" different table
names with the same cxonstruct according rto some condition?
for example, sqlite3_prepare_v2(db, "select * from ? where.",...);
2. In "order by" clause
Hello John,
this is extremely helpful. Thanks a lot!!!
Dimitris
34 matches
Mail list logo