sorting by page title, with some complexity

[evan]

Hi,

Is there an easy way to add rules to the "sort" selector, or otherwise manipulate search results directly with PHP?  I'm trying to sort results by Page title, but I want to ignore any initial "the", "an", or "a."

I know I can export the results into an array, manipulate them, and then display them, but that would disrupt pagination, and seems clunky. Surely there's an elegant solution?

Thanks!

-evan

[Wanze]

Hi evan,

maybe this does work?

$pages->find('!title^=the |an |a , sort=title');

Just an idea. There are some limits on the length of words, i'm not an expert here.

If it doesn't work, I'd go with the array.

[kongondo]

Hi evan,

maybe this does work?

$pages->find('!title^=the |an |a , sort=title');

Just an idea. There are some limits on the length of words, i'm not an expert here.

If it doesn't work, I'd go with the array.

Not really. This excludes pages such as "their", "another", "about"

[Nico Knoll]

Maybe: $pages->find('!title^="the "|"an "|"a ", sort=title');   ?

[Wanze]

Not really. This excludes pages such as "their", "another", "about"

Tested? I added a space after the keywords in my selector for this cases

[evan]

Just to clarify, I'm not looking to exclude Pages that begin with the/a/an, but rather alphabetize them by the second word if they do start with any of those.  For example, given these Page titles:

Aguirre, the Wrath Of God
The Wild Blue Yonder
Cave of Forgotten Dreams
Stroszek
Lessons of Darkness
A Heart Of Glass

It should produce:

Aguirre, the Wrath Of God
Cave of Forgotten Dreams
A Heart Of Glass
Lessons of Darkness
Stroszek
The Wild Blue Yonder

What would be amazing if there was an optional field in find() that would let you call up a PHP function, in which you could manipulate the results.   Unless, of course, you can do something like that already.

-evan

[Soma]

Add a text field to add an alternative sort title to sort by it instead. No other way if you still want to keep it easy with pager.

[kongondo]

Tested? I added a space after the keywords in my selector for this cases

Yes, tested. Should have said that

[evan]

Add a text field to add an alternative sort title to sort by it instead. No other way if you still want to keep it easy with pager.

Yeah, I was thinking that.  I guess that'll have to do.

-evan

[Macrura]

what about a simple string replace on the search results (get rid of leading the, an, a)..

$sort_title = preg_replace("/^(the¦a¦an)\b\ */i", "", "Some title");

[evan]

what about a simple string replace on the search results (get rid of leading the, an, a)..

$sort_title = preg_replace("/^(the¦a¦an)\b\ */i", "", "Some title");

That'll work, but as soma said, I'll have to store that value in the Page in order for it to be alphabetized as I want it to.  Thanks!

-evan

[kongondo]

Maybe: $pages->find('!title^="the "|"an "|"a ", sort=title');   ?

Nope. Don't work. Tested.

[Martijn Geerts]

Maybe you can get InputfieldPageName module to the site/modules/ folder and change the JS name behaviour. 

( making the name fields javascript remove the / an / a / )

 

$title = "The tralalala";

$title = trim(preg_replace("/^(the¦a¦an)\b\ */i", "", $title));

$iLookLikeName = $sanitizer->pageName($title);

$pages->find("name=$iLookLikeName, sort=name");

[evan]

I ended up going with soma's suggestion, and just storing the alternate title as field sort_title in the Page.

Here's the function I made, triggered by the hourly LazyCron:

function createSortTitles(HookEvent $e) {
	$items = wire('pages')->find('template=item');
	foreach ($items as $item) {
		if ($item->sort_title=="") {
		$item->setOutputFormatting(false);
		$new_title = (preg_match( "/^(the|a|an)\s/i", $item->title)) ?
			substr(strstr($item->title," "), 1) : $item->title;
		$item->sort_title = $new_title;
		$item->save();
		}
	}
}

For some reason I couldn't get the selector sort_title='' to work correctly -- it would only return a portion of the Pages.  Hence the conditional instead.

-evan

[Pete]

With just a little more work you could put that function into a module that runs on page save which would then be ideal for your needs

For example, something like this:

<?php
class SortPageTitles extends WireData implements Module {

    /**
     * getModuleInfo is a module required by all modules to tell ProcessWire about them
     *
     * @return array
     *
     */
    public static function getModuleInfo() {

        return array(

            // The module's title, typically a little more descriptive than the class name
            'title' => 'Sort Page Titles',

            // version: major, minor, revision, i.e. 100 = 1.0.0
            'version' => 100,

            // summary is brief description of what this module is
            'summary' => 'Sorts pages by title by second word if first word is "a, the, an" etc.',
            
            // Optional URL to more information about the module
            'href' => 'http://processwire.com/talk/topic/3306-sorting-by-page-title-with-some-complexity/',

            // singular=true: indicates that only one instance of the module is allowed.
            // This is usually what you want for modules that attach hooks.
            'singular' => true,

            // autoload=true: indicates the module should be started with ProcessWire.
            // This is necessary for any modules that attach runtime hooks, otherwise those
            // hooks won't get attached unless some other code calls the module on it's own.
            // Note that autoload modules are almost always also 'singular' (seen above).
            'autoload' => true,
            );
    }

    /**
     * Initialize the module
     *
     * ProcessWire calls this when the module is loaded. For 'autoload' modules, this will be called
     * when ProcessWire's API is ready. As a result, this is a good place to attach hooks.
     *
     */
    public function init() {

        // add a hook after the $pages->save, to sort certain pages if necessary
        $this->pages->addHookAfter('save', $this, 'sorttitles');
    }

    /**
     * Hooks into the pages->save method and sorts pages with a certain template based on certain criteria
     *
     */
    public function sorttitles($event) {
        $page = $event->arguments[0];
        
        // Only run if the page we just saved has the "item" template
        if ($page->template == 'item') {
            $items = wire('pages')->find('template=item');
            foreach ($items as $item) {
                if ($item->sort_title=="") {
                $item->setOutputFormatting(false);
                $new_title = (preg_match( "/^(the|a|an)\s/i", $item->title)) ?
                    substr(strstr($item->title," "), 1) : $item->title;
                $item->sort_title = $new_title;
                $item->save();
                }
            }
        }
    }
}

I've not tested it, but it's based on the helloworld module, so if you download it (attached) and put it in your /site/modules folder it should work for you, and hopefully be a good example of how easy it is to write modules too

SortPageTitles.module

[evan]

Hey, nice!  I did notice that there was a page->save() hook after the fact, maybe I'll try that out.  It was just quicker for me to do LazyCron, as I'd done it before.  But it's always good to know a little more about PW!

-evan

[Pete]

No probs. On a few sites I've ended up creating a little site-specific module like this to check for various templates and do certain things after page save among other things and they can be really useful.

[Ryan Pierce]

Whoa. What perfect timing Pete - just came here looking for this exact same thing. Should definitely be submitted to Modules.

[Pete]

Well it's sort of specific to the template/fields evan has on his site, so no point in submitting it really unless you're reproducing his exact same structure.

All it is is his function inside the default helloworld module that comes with every PW install (check inside site/modules directory, and read more here: http://processwire.com/api/modules/ ). You'll be up and running with your own little helper modules in no time

[Soma]

... and end up in endless loop land. ;-)

Better just save the field only $item->save('sort_title') so the hook doesnt get called again and again and...

[Pete]

... and end up in endless loop land. ;-)

Better just save the field only $item->save('sort_title') so the hook doesnt get called again and again and...

Ah yes, Soma is of course correct that since it runs on page save and saves the page at the end it will go on FOREVER. Like I said, it wasn't tested

Changing $item->save to $item->save('sort_title') will of course work like he says.

[evan]

If you hook onto page->save(), it's no longer necessary to loop over pages->find():

    public function sorttitles($event) {
        $page = $event->arguments[0];
        
        // Only run if the page we just saved has the "item" template
        if ($page->template == 'item') {
                $new_title = (preg_match( "/^(the|a|an)\s/i", $page->title)) ?
                    substr(strstr($page->title," "), 1) : $page->title;
                $page->sort_title = $new_title;
                $page->save('sort_title');
                }
            }
        }
    }
}

-evan