Hi,
Do you have experience with this warning ? Can anyone help me to get rid of
this warning.
--
Regards.
Thilina Buddika Abeyrathna,
Department of Computer Engineering,
Faculty Of Engineering,
University of Peradeniya,
Sri Lanka.
Hi all,
I merged some important changes to the master branch from the GSoC
project of Rouslan Placella, introducing a few new classes
(PMA_Response, PMA_Header, PMA_Footer, ..) that will handle html and
ajax responses, buffering output and generating the header, footer,
body. More information on how this works and how this impacts your
code can be found in the wiki page that Rouslan created [0].
Both Rouslan and me tested it and most bugs should be fixed now, but
don't hesitate to report any misbehaviour that might be related to the
introduction of these classes.
@ all GSoC students, if you're not doing this regularly already, this
might be a good time to update your branch with the recent changes in
the master branch (i.e. do a 'merge master' in your branch)
@ Rouslan:
There were no issues during merging.
After I merged, I noticed that the 'git revision' section on the main
page is broken. Can you have look at it?
About your wiki page, apart from some typo's, it looks very good. You
explained very clearly what the different classes do and how to use
them.
I think it might be a good idea to stress that using the PMA_Response,
etc. classes is preferred and the old style is deprecated/strongly
discouraged.
It might be useful to add some examples when to use certain
functionality, like disabling headers/footers, ...
I noticed you removed some unit tests related to global methods that
were replaced by methods in your classes. Do you plan to reintroduce
them when you will write unit tests for your new classes?
[0] http://wiki.phpmyadmin.net/pma/Generating_pages_and_ajax_responses_in_pma4
Kind regards,
Dieter Adriaenssens
Hi Chanaka,
git bisect tells me that your commit
be2dceb16471a2ffb69e0af48449277ea64fe89d broke a feature.
When browsing a table having a foreign key pointing to another table,
the value should be a link that permits to either reach this other table
or to choose another value.
--
Marc Delisle
http://infomarc.info
Chanaka,
about your question on long lines, I agree with you that this could be
put on two lines. Suggestion:
if ((PMA_strlen($wktval) > $GLOBALS['cfg']['LimitChars'])
&& ($_SESSION['tmp_user_values']['display_text'] ==
self::DISPLAY_PARTIAL_TEXT)
) {
--
Marc Delisle
http://infomarc.info
Hi all,
I'm trying to add a new page to the wiki that documents the usage of the
new framework for generating pages and ajax responses, as implemented in
my git tree. However, I'm unable to find an option in the wiki for
creating a new page. Is it disabled, or am I just missing something?
Bye,
Rouslan
Hi Marc,
While looking at "tbl_zoom_select.php":-
Line 191-201
if (isset($inputs) && ($inputs[0] != 'pma_null' || $inputs[1] !=
'pma_null')) {
$flag = 2;
for ($i = 0 ; $i < 4 ; $i++) {
if ($inputs[$i] != 'pma_null') {
$key = array_search($inputs[$i], $fields_list);
$tbl_fields_type[$i] = $fields_type[$key];
$tbl_fields_collation[$i] = $fields_collation[$key];
$tbl_fields_null[$i] = $fields_null[$key];
}
}
}
As the condition would be true in cases when post param $inputs is set and
has index 0 or 1 set to some column name. It would mostly happen if a half
or full filled search form is submitted, but there seems to be no further
use of $tbl_fields_collation & $tbl_fields_null. Though $tbl_fields_type
gets used afterwards with indexes >= 4.
I think as a result after the zoom plot is displayed, clicking at "show
search criteria" doesn't show any previously filled criteria.
I also didn't understand the use of $flag variable.
Thanks
--
Atul Pratap Singh
Senior Undergraduate Student
Institute Of Technology,
Banaras Hindu University, Varanasi,
India.
Hi Atul,
here is a bug that happens in your tree but not in upstream master. On a
table with some columns being nullable, do a zoom search, then click on
a data point to edit something.
You'll see that the generated UPDATE statement incorrectly sets all
nullable columns to NULL.
--
Marc Delisle
http://infomarc.info