Today I enabled the allow_url_include configuration option in PHP on Apache on ProgClub MemberNet. This allows me to include a source file directly from svn, e.g.:
require_once( 'https://www.progclub.org/svn/pcrepo/slib/trunk/src/uri.php' );
Today I enabled the allow_url_include configuration option in PHP on Apache on ProgClub MemberNet. This allows me to include a source file directly from svn, e.g.:
require_once( 'https://www.progclub.org/svn/pcrepo/slib/trunk/src/uri.php' );
Finally got around to making my MemberNet web-page look nice and pretty. Also integrated the look and feel with all of my PHP tests.
I learned about the PHP get_called_class function today. Basically it returns the name of the class that the function that’s running is defined in. Handy for figuring out which class a static method is running in. Compare with get_class.
The file function parses a text file into an array, and the file_get_contents function returns a text file as a string.
Today I discovered the user_agent php.ini configuration option. Basically it allows you to specify the user agent PHP uses when it sends HTTP requests for files. I was screen scraping some data from Wikipedia (processing URI schemes) and it was replying with a 403 error, presumably because they’ve banned the default PHP user agent. Anyway I just changed my user agent to a copy of my one from Firefox and things started working. Pretty handy option!
I used the CSS overflow Property today and set it to ‘auto’ so that my code fit on the screen.
pre { overflow: auto; }
I used the CSS :last-child and :after selectors to hack the content: property on my menu lists. Basically items in the menu have a pipe ‘|’ between them, except no pipe before or after the list. I did that in CSS with:
#menu li:after { content: ' | '; } #menu li:last-child:after { content: ''; }
Thanks to this handy document How To Set Cron to Run Every 5 Minutes I now my crontab configured so that my jj5-test repo gets updated every five minutes. It’s the sort of thing I generally do in a post-commit hook, but in this case that won’t work owing to the way the servers are configured (the files are in my account and not owned by the www-data user the commit-hook runs as).
So to configure cron I issued the command:
$ crontab -e
And then to update my svn working copy:
# m h dom mon dow command */5 * * * * cd /home/jj5/web/test && svn update > /dev/null
On my list of things to do is read the document Handling character encodings in HTML and CSS from the W3C. For some reason I can’t quite bring myself to concentrate on it right now.
Found some instructions about How to open file links in Firefox 1.5 and above and I’m going to try it out just as soon as I can reset my browser.
Basically you add something like the following to your user.js file which is in C:\Documents and Settings\[User Name]\Application Data\Mozilla\Firefox\Profiles\[Profile]\user.js.
user_pref("capability.policy.policynames", "localfilelinks"); user_pref("capability.policy.localfilelinks.sites", "http://www.example.com"); user_pref("capability.policy.localfilelinks.checkloaduri.enabled", "allAccess");