phpMyEdit - Viewing Bug Advanced Details Go Back ]
25 core / general major always 2003-02-12 23:36 2006-01-23 01:46
anonymous IIS  
nepto windows 2000 server  
normal php4.  
closed 0.6.1  
all fixed  
none  
none  
tons of errors when display_errors set to ON
Everytime I turn on display_errors in php this script spits out hundreds of errors of the type:

Notice: Undefined index: filters in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2869

Notice: Undefined index: triggers in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2870

Notice: Undefined index: logtable in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2871

Notice: Undefined index: page_name in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2873

Notice: Undefined index: cgi in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2911

Notice: Undefined index: cgi in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 2912

Notice: Undefined index: sfn in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 1259

Notice: Undefined index: operation in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 1259

Notice: Undefined index: operation in c:\inetpub\wwwroot\testing\phpMyEdit-5.2\phpMyEdit.class.php on line 1263
edit the php.ini to set display_errors = On and then the errors display on any table page produced by the phpmyeditsetup
0000025-notices.diff.txt (10 KB) 1970-01-01 01:33
0000025-error_report.diff.txt (1 KB) 1970-01-01 01:33

Bug Notes
nepto   
2003-02-16 02:13   
Well, this is confirmed. These are notices. If we wanted to have them all fixed, code will be probably huge, what may result into more longer parsing time. However, I will see if it is possible to eliminate such notices, or it is very difficult goal.
nepto   
2003-04-08 19:51   
Some error_reporting() calls were added into phpMyEdit for preventing before notices reporting. Can you confirm, that these warnings do not occur anymore?
anonymous   
2003-07-13 18:01   
Same problem here!!
nepto   
2003-07-14 01:34   
Yeah, this note really helps me! :-/

Could you at least tell me what version you are using? Did error_reporting() call in core class solve this problem?
hbernard   
2004-04-17 12:22   
(I went on this when looking on #227)

this bug can be closed in two ways

1. Have a real deactivation of E_NOTICE

current PME tried to deactivate E_NOTICE, but wasn't successful.

Surprisingly, when prefixed by @, function error_reporting doesn't seem to do anything but retrieving current level.

Uploaded error_report.diff patch fixes this.

2. Fix all occurences of E_NOTICE
As you said in a note above, this is lot of work.

However, I made a ten minutes patch which is a start of such clean up. It fixes immediate notices when using basic options.
(uploaded notices.diff)
I don't think this will make PME much slower.

So, nepto, which way do you prefer ?

1. => apply patch error_report.diff, resolve this bug. PME won't care anymore about E_NOTICE
2. => apply patch notices.diff, tell PME to output E_NOTICES, keep this bug open some time to see what was missed in this patch.

because #227 is about E_Warning, it has to be handled specifically.

I got no opinions on this. Both solutions are OK for me. I'm only sure we can fix it quickly, depending on the way you want.
nepto   
2004-04-17 13:20   
I think we can simply go with [1].

It is really strange, that @ before error_reporting() makes this function do nothing...
nepto   
2004-04-17 13:35   
Simplier version commited into CVS.

Please reopen this bug-report when needed.