| From: | Andrew Dunstan <andrew(at)dunslane(dot)net> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | db(at)zigo(dot)dhs(dot)org, ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp>, Bruce Momjian <bruce(at)momjian(dot)us>, pgsql-patches(at)postgresql(dot)org |
| Subject: | Re: UTF8MatchText |
| Date: | 2007-05-21 16:59:33 |
| Message-ID: | 4651CFF5.6010709@dunslane.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers pgsql-patches |
Tom Lane wrote:
> Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
>
>> But why are we doing that CHAREQ?
>>
>
> To avoid the cost of the recursive call, just like it says.
>
>
>> If it succeeds we'll
>> just do it again when we recurse, I think.
>>
>
> If you move the other two cases then you could advance t and p before
> entering the recursion.
>
>
>
Yeah. Since I have removed the "_" case I believe it's now safe there to
use BYTEEQ/NextByte, and since they are sufficiently cheap it's not
worth worrying about.
Attached is a patch version that I think draws together all the threads
of discussion so far. It's in fact quite a lot simpler than the existing
code, with no special UTF8 case - this should improve LIKE/ILIKE
processing for all charsets.
More eyeballs please for nasty corner cases.
cheers
andrew
| Attachment | Content-Type | Size |
|---|---|---|
| like.patch | text/x-patch | 25.5 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Karl O. Pinc | 2007-05-21 17:02:29 | Re: COPY into a view; help w. design & patch |
| Previous Message | Jim C. Nasby | 2007-05-21 16:23:57 | Re: COPY into a view; help w. design & patch |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Peter Eisentraut | 2007-05-21 17:11:50 | Re: xpath_array with namespaces support |
| Previous Message | Jeff Davis | 2007-05-21 15:24:21 | Re: Synchronized Scan |