Facets are currently hardcoded in https://github.com/yawik/Solr/blob/master/src/Solr/Filter/JobBoardPaginationQuery.php#L88

We should make this configurable

10 jobs found, which have to expire ...
Done                 [-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------->] 100%             

at least on jobs.yawik.org

Currently, the sort order is hardcoded descending by date.

We want to make this configurable. For example this would make the

< $randString = rand(1,100);
< $query->addSortField('random'.$randString, SolrQuery::ORDER_DESC);

The current solr schema has a couple of redundant fields. This issue collects things, which are related to cleanup the schema.

One question, which comes to my mind.

• Is JTS is really needed?

Hallo,

ist es momentan möglich Facet Queries zu nutzen?

Ich hatte folgendes Szenario:
Ich wollte neben den Standard-Filtern einen Filter haben, der die Jobs nach dem publishDate filtert. Dazu sollen aber Filterqueries genutzt werden. In meinem Fall sieht es z.b. so aus:

$query->addFacetQuery('datePublishStart:[NOW-1DAY TO NOW]') // 24h
            ->addFacetQuery('datePublishStart:[NOW-3DAYS TO NOW]') // 3 Tage
            ->addFacetQuery('datePublishStart:[NOW-7DAYS TO NOW]') // 7 Tage
            ->addFacetQuery('datePublishStart:[NOW-30DAYS TO NOW]'); // 30 Tage

Ich habe als Lösung die Klasse \Solr\Facets erweitert, um auch Facet Queries auszulesen.

Facets Fields können ja momentan über die Konfigurationsdatei definiert werden. Ist das für Queries auch geplant?

VG Stefanie Drost

When using the --batch option, the console command solr index job does not run at all.

The module should be compatible with the latest Solr version.

So you can filter the results by country also.

@fedys @TiSiE

adding a jobosting without a fulltext to solr makes no sense. So what can we do, if a job is activated and no html is given.

My idea is to simply fetch the html. If fetching fails, the job is not inserted.

Can we/should we do this in the solr module?

Keep in mind that all the time consuming tasks like fetching pages, inserting into solr, sending mails... should be moved into some queing system

It should be possible to configure default filter queries which are used for every search request regardless of user input via the search form.

Solr provides the ability to define such filter queries so we should use this.
All we have to do is:

• Provide options to configure filter queries
• Use these options to set up Solr filter queries prior to send the search query.
• tests

can we use our new viewHelper when saving the "location" in solr?

The value, which is stored in Solr:

https://github.com/yawik/Solr/blob/master/src/Entity/JobProxy.php#L358

It has been shown that an index job takes a long time, if an optimize is done after each commit

Let's remove the optimize

Docs: http://yawik.readthedocs.io/en/latest/modules/solr/index.html

https://travis-ci.org/yawik/Solr/builds/280462993#L1353

Solr\Paginator\JobsBoardPaginatorFactory::setCreationOptions() in /home/travis/build/yawik/yawik/module/Solr/test/SolrTest/Paginator/JobsPaginatorFactoryTest.php on line 88

Adding jobs with the SimpleImport Module are not added automatically into Solr (if Solr module is used).

As @TiSiE has mentioned in yawik/SimpleImport#3 the Solr module needs a "preinsert" hook