Le 2011-04-10 15:44, Karol Danko a écrit :
Hi Mark, thank you for reply
I will use both POSIX (for WHERE) and PCRE for (match&replace)
The script will first select all rows matching WHERE fieldset
after that it will loop the results and handle the replace via PHP
(2nd fieldset MATCH and 3rd fieldset REPLACE)
(So to check all rows with PHP, the WHERE fieldset would be empty)
The problem however is the use of both POSIX and PCRE on the same page which
may be confusing,
but using only MySQL POSIX would make more complicated replaces impossible.
On the other hand using only PHP's PCRE and looping all the results may be
very inefficient on large result sets.
Indeed. About the interface, don't you think that introducing a new
syntax ($1 and #1) would be too confusing?
I think it would be better to offer a simpler syntax even if this means
having less functionality. Therefore you would keep the "standard" way
of using a backreference, a backslash followed by a digit.
On Sun, Apr 10, 2011 at 12:45 PM, Marc Delisle <marc(a)infomarc.info> wrote:
> Le 2011-04-10 05:35, Karol Danko a écrit :
>> Hi everybody,
>>
>> (I used phpmyadmin-users mailing first but I did not receive any response
> so
>> sorry for double-post)
>>
>> I'm new to phpmyadmin development, so far I have been just passive daily
>> user.
>> I think phpmyadmin is great and I would love to help making it even
> better.
>>
>> I'm working on regex-replace interface for phpmyadmin:
>>
>
http://sourceforge.net/tracker/index.php?func=detail&aid=1398891&gr…
>>
>>
>> I would love to hear your thoughts about this concept I have, I suggest
>> following UI:
>>
http://karoldanko.com/img/regex-replace.png
>
> Hi Karol,
> I'll have suggestions about the interface but first a question.
>
> Looking at
http://dev.mysql.com/doc/refman/5.5/en/regexp.html it seems
> that MySQL REGEXP is based on POSIX extended regular expressions and not
> PCRE (which supports back references).
>
>
> --
> Marc Delisle
>
http://infomarc.info
--
Marc Delisle
http://infomarc.info