Run ProcessWire URL via command line

[DrQuincy]

My hosting environment does not allow for PHP to run for more than 120 seconds when it is a web request. However, there is no limit via command line PHP.

Is it possible take a ProcessWire URL such as http://domain.com/foo/bar/?bar=foo and run it as a PHP command line? Something like:

php /path/to/pw/site/index.php route=/foo/bar/ bar=foo

Is such a thing possible? Thanks.

[bernhard]

You can bootstrap PW and create your own command line scripts that have full access to the PW API: https://processwire.com/docs/front-end/include/

[DrQuincy]

Thanks, I've seen that. I just wondered if there was a way to — rather then get a page through the API which I could do per your link — get PW to actually process it as though it were a front-end request. If I do wire('page')->get(123) it will retrieve all the data but what I want is to trigger something that is the same as accessing the page through HTTP.

Is there a built-in function for this?

[bernhard]

Sure, you can access it via WireHttp, but then your 120s limit will apply because it will be seen as a regular web request...

$http = new WireHttp();
$http->get("https://your.site/your/url");

 

[DrQuincy]

Ah, yes, thanks for that.

I'll have a think about how to implement it.

[horst]

@DrQuincy what is this page doing that takes a lot of time?

If there is a loop operation that iterates over a lot of items, and you are able to implement an optional get param with the last processed item identifier, there is a good chance that this is all you need to do to make this work.

For this to work you would check if a start-identifier was given with the URL, if not start with the first item, otherwise with the #n item from the $input->get->myidentifiername.
When starting the script, store a timestamp and in each loop operation compare it with the current timestamp. Maybe when it reached the point greater or equal 100 seconds, do a redirect call to the same page, but with the current (last) processed loop item identifier: 

    ...
    if(time() >= $starttimestamp + 100) {
        $session->redirect($page->url . '?itemidentifiername=' . $lastProcessedItemID);
    }
    ...

Or, much easier, have you tested if the 120 second time limit can be reset from within a script? (calling set_time_limit(120) from within a loop can make it run endless on some server configurations)

 

[DrQuincy]

Thanks for your suggestion!

In this case it's for something that runs on cron in the night and syncs a website property listing with third-party listing software. It takes a long time to run as it needs to copy images over HTTP and then resize some of them so when the website is accessed during the day it's a lot faster.

Thanks for your suggestion though. I must admit I have never thought of doing it that way — it's a neat solution. If you run curl and then redirect via a session as you have done, does curl honour the redirect? I can't say I've never tried that but from when I have used curl in the command line I'm sure it just returns redirect headers, location, etc.

I just thought maybe there was a way with the API and get a page's rendered HTML as opposed to as an object to force ProcessWire to run the task as it would if you visited it via a browser.

I don't know, I thought maybe there'd be something like $pages->get('/path/to/page/')->render() that returns the page HTML. If so, from my command line script I could simply map the arguments to $_GET and also pass a path.

[Jan Romero]

1 hour ago, DrQuincy said:

does curl honour the redirect?

Only if you tell it to: https://curl.se/docs/manpage.html#-L
You’ll also have to increase the maximum number: https://curl.se/docs/manpage.html#--max-redirs

[DrQuincy]

Thanks Jan!