For anyone doing any 6.x development, I spent this afternoon making coder (http://drupal.org/project/coder) compatible with Drupal HEAD (6.x). I also added a new upgrade review for 5.x to 6.x (which is available in both the HEAD and DRUPAL-5 branches). The 6.x upgrade review is better than my past upgrade reviews, and (I think) catches everything except the url() and l() argument changes. (I hope to add these checks before the code freeze.) While I'm announcing the coder 6.x availability, I should also mention that I created a core style review patch (http://drupal.org/node/134493). If you're interesting in making Drupal core coding standards compliant (http://drupal.org/node/318), then review of this patch would be helpful. It's mostly spaces - add an extra one here - remove an extraneous one there; but it's a moderate sized patch- 173 changes throughout core. Doug Green 904-583-3342 www.douggreenconsulting.com Bringing Ideas to Life with Software Artistry and Invention... Providing open source software political solutions
For anyone doing any 6.x development, I spent this afternoon making coder (http://drupal.org/project/coder) compatible with Drupal HEAD (6.x). I also added a new upgrade review for 5.x to 6.x (which is available in both the HEAD and DRUPAL-5 branches). The 6.x upgrade review is better than my past upgrade reviews, and (I think) catches everything except the url() and l() argument changes. (I hope to add these checks before the code freeze.)
Note that the issue for the l() url() change contains a PHP script that can automatically convert all l()/url() calls in Drupal code. Perhaps you could adapt it? Steven
Thanks Steven! I updated coder, but used a slightly different regex. The warning references your script. One question however, I'm not sure that you script works on cases with functions as arguments. When I tried to borrow the regex, it didn't work in the matching phase. I didn't run any tests, but the regex was off when I used it in coder. Does your script properly ignore? return l(format_date($query->created, 'custom', 'F, Y'), "$arg/$query->name"); If it doesn't work, you might want to look at the regex in coder_6x.inc, which I "think" works better. Doug Green 904-583-3342 www.douggreenconsulting.com Bringing Ideas to Life with Software Artistry and Invention... Providing open source software political solutions -----Original Message----- From: development-bounces@drupal.org [mailto:development-bounces@drupal.org] On Behalf Of Steven Wittens Sent: Monday, April 09, 2007 10:31 PM To: development@drupal.org Subject: Re: [development] Announce Coder 6.x
For anyone doing any 6.x development, I spent this afternoon making coder (http://drupal.org/project/coder) compatible with Drupal HEAD (6.x). I also added a new upgrade review for 5.x to 6.x (which is available in both the HEAD and DRUPAL-5 branches). The 6.x upgrade review is better than my past upgrade reviews, and (I think) catches everything except the url() and l() argument changes. (I hope to add these checks before the code freeze.)
Note that the issue for the l() url() change contains a PHP script that can automatically convert all l()/url() calls in Drupal code. Perhaps you could adapt it? Steven
On 10.04.2007, at 16:05, Doug Green wrote:
One question however, I'm not sure that you script works on cases with functions as arguments. When I tried to borrow the regex, it didn't work in the matching phase. I didn't run any tests, but the regex was off when I used it in coder. Does your script properly ignore?
Maybe the multi regex parser for JavaScript compression from http:// drupal.org/node/119441 can be used for this to prevent overlapping matches. Konstantin Käfer – http://kkaefer.com/
One question however, I'm not sure that you script works on cases with functions as arguments. When I tried to borrow the regex, it didn't work in the matching phase. I didn't run any tests, but the regex was off when I used it in coder. Does your script properly ignore?
return l(format_date($query->created, 'custom', 'F, Y'), "$arg/$query->name");
The regexp doesn't care about the contents of arguments, it only cares about balanced parentheses. I did run into weird problems where the regexp would fail to match if I processed an entire module at once. I had to split it up at known non-match points for it to pick up everything. Everything pointed to a bug in PCRE with recursive regular expressions. Steven Wittens
participants (3)
-
Doug Green -
Konstantin Käfer -
Steven Wittens