We have started using queries of the form
SELECT ... WHERE ... AND local_relpath LIKE ...
and I was curious about the performance of LIKE. So I tried a large
database:
$ sqlite3 wcx.db "select count(*) from nodes"
377021
and a LIKE query:
$ time sqlite3 wcx.db "select count(*) from nodes w
* Philip Martin:
> We have started using queries of the form
>
>SELECT ... WHERE ... AND local_relpath LIKE ...
>
> and I was curious about the performance of LIKE.
LIKE is ASCII-case-insensitive in SQLite, and the indexes are
case-sensitive by default, so SQLite can't do the usual range
opti
It looks like the issue
http://subversion.tigris.org/issues/show_bug.cgi?id=3727 is because of
the commit done in r880559.
Since this is a conscious change, should this issue be ignored?
Thanks and Regards
Noorul
Florian Weimer writes:
> * Philip Martin:
>
>> We have started using queries of the form
>>
>>SELECT ... WHERE ... AND local_relpath LIKE ...
>>
>> and I was curious about the performance of LIKE.
>
> LIKE is ASCII-case-insensitive in SQLite, and the indexes are
> case-sensitive by default, s
Noorul Islam K M writes:
> It looks like the issue
> http://subversion.tigris.org/issues/show_bug.cgi?id=3727 is because of
> the commit done in r880559.
>
> Since this is a conscious change, should this issue be ignored?
>From the log message:
r880559 | hwright | 2009-11-12 18:21:32 + (T
Philip Martin writes:
> I need something like
>
> local_relpath = 'zig1/zag27
> OR (local_relpath > 'zig1/zag27/' AND local_relpath < 'zig1/zag270')
>
> and that is as slow as LIKE. Adding that "local_relpath =" is the
> problem, without it I get the children in 0.006s. With it I get
* Philip Martin:
>> and that is as slow as LIKE. Adding that "local_relpath =" is the
>> problem, without it I get the children in 0.006s. With it I get the
>> path and the children but it takes 0.35s.
>
> SQL is tricky: I can run a query using either
>
> local_relpath = 'zig1/zag27
>
> or
On Thu, Oct 28, 2010 at 16:42, Daniel Shahaf wrote:
> C. Michael Pilato wrote on Wed, Oct 27, 2010 at 10:11:20 -0400:
>> On 10/27/2010 09:41 AM, Ivan Zhakov wrote:
>> > 2. Let's think about users: they have different needs for Subversion:
>> > just a client, client and server etc. There are differ
On Thu, 2010-10-28 at 14:28 -0400, C. Michael Pilato wrote:
> On 10/28/2010 01:58 PM, Greg Stein wrote:
> > On Thu, Oct 28, 2010 at 13:04, C. Michael Pilato
> > wrote:
> >> On 10/28/2010 12:50 PM, Hyrum K. Wright wrote:
> >>> On Thu, Oct 28, 2010 at 11:48 AM, wrote:
> Author: cmpilato
> >>
Florian Weimer writes:
> It seems an optimizer issue. Which version of SQLite do you use?
I was using 3.6.21-2~bpo50 on Debian/stable. I've just built a local
3.7.3-1 and get the same result.
The database has 377021 rows. The exact commands are:
sqlite3 wcx.db "select count(*) from nodes
Philip Martin writes:
> sqlite3 wcx.db "select count(*) from nodes where wc_id = 1 and
> local_relpath = 'zag1/zag27'"
Typo, that should be 'zig1/zag27'.
--
Philip
Philip Martin writes:
> I suppose I will have to try using EXPLAIN.
For the fast query
"select ... where wc_id = ...
and local_relpath > ... and local_relpath < ..."
explain starts:
0|Ttace|0|0|0||00|
1|Integer|1|1|0||00|
2|String8|0|2|0|zig1/zag27/|00|
3|String8|0|3|0|zig1/za
http://subversion.apache.org/docs/community-guide/releasing.html#custom-releases
refers and links to svncorp's trademarks policy.
Should it link to ASF's trademarks policy?
On Thu, 2010-10-28 at 13:55 -0400, Greg Stein wrote:
> On Thu, Oct 28, 2010 at 12:36, Hyrum K. Wright
> wrote:
> > On Thu, Oct 28, 2010 at 11:29 AM, Mark Phippard wrote:
> >> On Thu, Oct 28, 2010 at 12:23 PM, Hyrum K. Wright
> >> wrote:
> >>> On Thu, Oct 28, 2010 at 10:09 AM, Julian Foad
> >>>
Philip Martin wrote on Fri, Oct 29, 2010 at 08:58:02 +0100:
> Can we treat the string as a bytes and just increment and carry?
No, this might invalidate the UTF-8 sequence. If a character spans
multiple bytes, then those bytes have header bits of the form /^1+0/
(in regex), so incrementing the by
Daniel Shahaf writes:
> Philip Martin wrote on Fri, Oct 29, 2010 at 08:58:02 +0100:
>> Can we treat the string as a bytes and just increment and carry?
>
> No, this might invalidate the UTF-8 sequence. If a character spans
> multiple bytes, then those bytes have header bits of the form /^1+0/
>
* Philip Martin:
>> I suppose I will have to try using EXPLAIN.
As expected, the slow query doesn't use the index. I think you should
bring this up on SQLite user mailing list.
--
Florian Weimer
BFK edv-consulting GmbH http://www.bfk.de/
Kriegsstraße 100 tel:
On Fri, Oct 29, 2010 at 3:57 AM, Noorul Islam K M wrote:
>
> It looks like the issue
> http://subversion.tigris.org/issues/show_bug.cgi?id=3727 is because of
> the commit done in r880559.
>
> Since this is a conscious change, should this issue be ignored?
Not necessarily. I recall at least one p
On 10/29/2010 06:46 AM, Daniel Shahaf wrote:
> http://subversion.apache.org/docs/community-guide/releasing.html#custom-releases
> refers and links to svncorp's trademarks policy.
>
> Should it link to ASF's trademarks policy?
Yup.
--
C. Michael Pilato
CollabNet <> www.collab.net <> Dis
On 10/29/2010 05:29 AM, Ivan Zhakov wrote:
> I've added note to the packages that requires registration in r1028656.
Thanks, Ivan.
--
C. Michael Pilato
CollabNet <> www.collab.net <> Distributed Development On Demand
signature.asc
Description: OpenPGP digital signature
Our current help texts explain syntax of commands quite well,
but they don't really explain the semantics. This makes svn help
fairly useless for people who just want to get going with their
first steps in Subversion.
The Subversion book on the other hand explains the semantics very well,
but it l
On 10/29/2010 11:57 AM, Stefan Sperling wrote:
> Thoughts? Comments?
I've not reviewed your suggested help text. But merely seeing the size of
makes me wonder ... is it time for 'svn help SUBCOMMAND --verbose' ?
I, for one, would love to offer two levels of built-in help. And if the
second leve
On Fri, Oct 29, 2010 at 12:01:41PM -0400, C. Michael Pilato wrote:
> On 10/29/2010 11:57 AM, Stefan Sperling wrote:
> > Thoughts? Comments?
>
> I've not reviewed your suggested help text. But merely seeing the size of
> makes me wonder ... is it time for 'svn help SUBCOMMAND --verbose' ?
Sure, n
On Oct 29, 2010, at 18:13 , Stefan Sperling wrote:
> On Fri, Oct 29, 2010 at 12:01:41PM -0400, C. Michael Pilato wrote:
>> On 10/29/2010 11:57 AM, Stefan Sperling wrote:
>>> Thoughts? Comments?
>>
>> I've not reviewed your suggested help text. But merely seeing the size of
>> makes me wonder ..
On Fri, 2010-10-29 at 12:06 +0100, Julian Foad wrote:
> On Thu, 2010-10-28 at 13:55 -0400, Greg Stein wrote:
> > On Thu, Oct 28, 2010 at 12:36, Hyrum K. Wright
> > wrote:
> > > On Thu, Oct 28, 2010 at 11:29 AM, Mark Phippard
> > > wrote:
> > >> On Thu, Oct 28, 2010 at 12:23 PM, Hyrum K. Wright
>
C. Michael Pilato wrote on Fri, Oct 29, 2010 at 12:01:41 -0400:
> On 10/29/2010 11:57 AM, Stefan Sperling wrote:
> > Thoughts? Comments?
>
> I've not reviewed your suggested help text. But merely seeing the size of
> makes me wonder ... is it time for 'svn help SUBCOMMAND --verbose' ?
Don't you
C. Michael Pilato wrote on Fri, Oct 29, 2010 at 11:16:36 -0400:
> On 10/29/2010 06:46 AM, Daniel Shahaf wrote:
> > http://subversion.apache.org/docs/community-guide/releasing.html#custom-releases
> > refers and links to svncorp's trademarks policy.
> >
> > Should it link to ASF's trademarks policy
> So the proposed help text below centers around the most common use
> cases.
Wow... this looks great. It will be nice to have this at the command line.
I read through it all... only found one small nit:
> If several commits to trunk were related to the fix, multiple
> revisions
> to
On Thu, Oct 28, 2010 at 08:42, Daniel Shahaf wrote:
>...
> As the page stands now, the "professionally supported and certified" is
> just lots of words that add nothing over "maintained by". And
> information-freeness is Bad.
I disagree on this one. "Maintained by Jeremy Whitlock" says something
Greg Stein wrote on Fri, Oct 29, 2010 at 15:17:39 -0400:
> On Thu, Oct 28, 2010 at 08:42, Daniel Shahaf wrote:
> >...
> > As the page stands now, the "professionally supported and certified" is
> > just lots of words that add nothing over "maintained by". And
> > information-freeness is Bad.
>
>
On Fri, Oct 29, 2010 at 17:00, Daniel Shahaf wrote:
> Greg Stein wrote on Fri, Oct 29, 2010 at 15:17:39 -0400:
>> On Thu, Oct 28, 2010 at 08:42, Daniel Shahaf wrote:
>> >...
>> > As the page stands now, the "professionally supported and certified" is
>> > just lots of words that add nothing over
31 matches
Mail list logo
