Re: cvs: pear /Validate/Validate IS.php /Validate/tests validate_IS.phpt

[snip] > > > + /* Same configuration as last time? No need to go further then */ > > + if (count($postCodes) && $dataFile == $lastFile && > > + (($strong && $lastUrl == $url) || !$strong)) { > *** (!$strong || $lastUrl == $url)) { > is more efficient Noted. > + return in_array($postCode, $postCodes); > > + } > ****************** > > + /* Sanity check the data file */ > > + if (!is_readable($dataFile)) { > > + return false; > > } > ****************** > I would simply not do that, you could imagine to run only on $url, no > $file. > Check it better as last mean after trying url, see below It doesn't make any sens to me to run without datafile as backup.. ..but, I guess if you tought about it someone else will to :) [snip] > > > - if (!count($postCodes) && file_exists($file)) { > > - $postCodes = file($file); > > + if (!count($postCodes) && file_exists($dataFile)) { > *** if (!count($postCodes) && $dataFile && is_readable($dataFile)) { > You take the file as last mean, if url was not used or did not work is_readable() makes more sense then file_exists, true, but the extra && $dataFile I can't see any use for since is_readable will reutrn false no matter what $dataFile contains (well, unless its readable file :)) > + $postCodes = file($dataFile); > > + $lastFile = $dataFile; > > } > > - if (is_array($postCodes) && in_array($postcode, $postCodes)) { > > + if (count($postCodes) && in_array($postCode, $postCodes)) { > > return true; > > } > > Haha, interesting discussion ... but I like the autosave feature based > on the writable or not file: > I mean in normal case, the web server will not have it , so no risk. > But then, if the admin runs it in a cli as root, then he *will* update > the file. > a+ > -- > toggg > Thanks -Hannes