Re: [sqlite] FTS4 + unicode61 tokenizer performance issue

Ok, I'm working on a simple project, which will reproduce a problem. Valentine. 11.06.2013, в 17:30, Dan Kennedy написал(а): > On 06/11/2013 08:20 PM, Valentine Silvansky wrote: >> Hi all! >> >> I have a custom-built sqlite with FTS4 and unicode61 tokenizer for it. I've

Re: [sqlite] Obtain limit-less count of previous select

On Tue, Jun 11, 2013 at 9:08 AM, Igor Tandetnik wrote: > On 6/11/2013 9:01 AM, Simon Slavin wrote: > >> The only time that _prepare() will use a lot of memory is when it has to >> generate a temporary index because you have not created a table index >> suitable for the WHERE

Re: [sqlite] Obtain limit-less count of previous select

On Tue, Jun 11, 2013 at 09:08:42AM -0400, Igor Tandetnik scratched on the wall: > On 6/11/2013 9:01 AM, Simon Slavin wrote: > > The only time that _prepare() will use a lot of memory is when it has to > > generate a temporary index because you have not created a table index > > suitable for the

Re: [sqlite] sqlite3Fts3ExprParse and MAX_EXPR_DEPTH = 12

On Tue, Jun 11, 2013 at 9:10 AM, wrote: > Hello, > > we just updated the SQLite sources/DLL in our project from version > 3.7.15.2 to version 3.7.17. > After executing the update we got a failing unit test concerning a string > based database search > with a very

Re: [sqlite] FTS4 + unicode61 tokenizer performance issue

On 06/11/2013 08:20 PM, Valentine Silvansky wrote: Hi all! I have a custom-built sqlite with FTS4 and unicode61 tokenizer for it. I've got the only one table "messages": CREATE VIRTUAL TABLE messages USING FTS4(body TEXT, tokenize=unicode61); So, I populate this table with ~100k entries,

[sqlite] FTS4 + unicode61 tokenizer performance issue

Hi all! I have a custom-built sqlite with FTS4 and unicode61 tokenizer for it. I've got the only one table "messages": CREATE VIRTUAL TABLE messages USING FTS4(body TEXT, tokenize=unicode61); So, I populate this table with ~100k entries, each one is a sentence (or a short message) in english

[sqlite] sqlite3Fts3ExprParse and MAX_EXPR_DEPTH = 12

Hello, we just updated the SQLite sources/DLL in our project from version 3.7.15.2 to version 3.7.17. After executing the update we got a failing unit test concerning a string based database search with a very long string (but this string is present in our current SQLite DB). In fact the

Re: [sqlite] Obtain limit-less count of previous select

On 6/11/2013 9:01 AM, Simon Slavin wrote: The only time that _prepare() will use a lot of memory is when it has to generate a temporary index because you have not created a table index suitable for the WHERE and ORDER BY clauses. I'm pretty sure this would happen on the first _step(), not on

Re: [sqlite] Obtain limit-less count of previous select

On 10 Jun 2013, at 1:56pm, Daniel Hofmann wrote: > Background is, that I want to implement my paging entirely in sql in order to > save memory, because the complete result-data consumes a lot of memory. If you use _prepare(), _step(), _finalize() instead of using _exec()

Re: [sqlite] Download SQLlite

Jeffrey Hartman wrote: > I tried downloading the Zip files to install SQLLite but my system > does not recognize the file format. Windows has built-in .zip support. What happens when you double-click on such a file? Regards, Clemens ___ sqlite-users

Re: [sqlite] Computed columns in VIEWs return NULL but should be able to be typed! Any ideas?

Hi all, I'm starting to use SQLITE in GIS use cases. I'm facing this view typing column issue, which prevent my favourite client from interpreting correctly numeric data types. I'm using QGIS, so I won't be able to map numeric values (they fall back as text values). Is there anything new since

[sqlite] Download SQLlite

I tried downloading the Zip files to install SQLLite but my system does not recognize the file format. I'm using Windows 7 Express and it is a 64-Bit machine. Thanks. Jeff Hartman ___ sqlite-users mailing list sqlite-users@sqlite.org

[sqlite] Obtain limit-less count of previous select

Hi, Background is, that I want to implement my paging entirely in sql in order to save memory, because the complete result-data consumes a lot of memory. Question is: Why isn't there a way like the FOUND_ROWS()-function of MySQL (I didn't find any reading the docs and googling), to get the