Hi, On Tue, Jun 3, 2014 at 5:11 PM, Chirayu Chiripal &lt;chirayu.chiripal(a)gmail.com > Hi, > > > > On Tue, Jun 3, 2014 at 12:15 PM, Dhananjay Nakrani < > dhananjaynakrani(a)gmail.com> >> Hello everyone, >> >> I'm trying to fix the failing tests for the code I developed. I need a >> little help from the community in it. >> >> I am providing appropriate new data in the data provider in >> 'PMA_Error_Handler_Test::providerForTestHandleError()' [0]. That data >> contains a string with empty single quotes ('') in it (see, Line#98 [1] >> ). So, I escaped it with backslash (\) and passed it as (\'\'). But that >> gives a very weird behavior. See Line#342 [2] of the log where it starts >> repeating the string again in itself. It seems to be going in recursion >> somewhere. >> >> Am I doing anything wrong Or is it just a bug in PHPUnit? >> Is there anyway I can fix this?? >> >> >> [0]: >> https://github.com/dhananjay92/phpmyadmin/blob/dhananjay-gsoc-dev/test/clas… >> [1]: >> https://github.com/dhananjay92/phpmyadmin/blob/dhananjay-gsoc-dev/test/clas… >> [2]: https://travis-ci.org/dhananjay92/phpmyadmin/jobs/26636648#L342 >> >> >> Thanks, >> Dhananjay Nakrani. >> >> > I don't think its a bug in PHPUnit. Its happening because of backtrace > information part in error handler due to which whatever parameters you are > passing to testGetDispErrorsForDisplayFalse() will reappear in $output in > backtrace section. Moreover, the parameters will be converted using > htmlspecialchars(). As the 6th parameter ($output_hide) of the function > contains html tags it will get converted by htmlspecialchars and then added > to the $output (this is the thing that appears at L342 of [2] you mentioned > before). So, you have to change your 6th parameter of data set #0 so that > it does not contain html tags. Note that you cannot even replace your > parameter with htmlspecialchars() equivalent string like string containing > &lt; or &gt; because each '&' of that string will again be replaced by > &amp; again and again. Hence, change your test data accordingly. > > Ohh, Okay. Now understood the problem. Thank you very much. Thanks, Dhananjay Nakrani.