BeBanjo and Sequence use search engine technology to provide the features you rely on to find content. On 16 April we updated our applications to replace the Sphinx search engine with Elasticsearch.
Our intent was for this change to be transparent to users whilst laying the foundations for great new features, but there is one change in the search behaviour and other features worth mentioning. Read on!
This note describes what has changed in full, and as usual you can contact support or your technical account manager for further clarification.
Elasticsearch in BeBanjo and Sequence
Tips and tricks for productive searching
Orientation
In BeBanjo, there’s an omnipresent Search catalogue field:
No matter where you are in BeBanjo, performing a search with that field takes you to the Catalogue page. The Schedule, Rights, and Catalogue pages also include a Filter field.
In Sequence, the Overview page includes a search field that searches all Jobs, and the Work page has a search field used to filter the type of work (e.g. Pending) that you’re focussed on.
Phrasing is important
Here you can see I’ve searched, in BeBanjo, for star wars and the results might not be what you expect:
Even though Flash Gordon is way better than Star Wars, in this case it wasn’t what I was looking for.
However, if I search for “star wars” (note the quotes) here’s the result:
When searching for star wars the search engine will look for star or wars, but when searching for “star wars” (i.e. with quotes) the search engine will look for that exact phrase.
This is how searching has always behaved in BeBanjo and Sequence, but it’s a useful feature that you might not know about.
You might be wondering why Flash Gordon was returned at all, what’s he got to do with Star Wars? Here’s why:
Notice that the Short Description includes the words star and wars…
And lastly, remember you can combine text searches with filters. So, for example, finding all the great movies starring Melody Anderson is a snap!
Go wild!
If you often use the underscore character in the names of Assets and Catalogue entries, or in External IDs, you need to understand how that character is used by the search engine.
In the old search engine (Sphinx) an underscore was used as a word separator, but that’s not the case with the new search engine (Elasticsearch). Great, so what?!
Well, for example if I have two Titles names Boulevard_SD and Boulevard_HD then a search for boulevard will return no results:
However, if I search for boulevard (i.e. with an asterisk) then two results are returned:
The asterisk is a wildcard that tells the search engine to look for zero or more characters. And of course, searching for boulevard_sd or boulevard_hd would return one result.
What’s the Scope, BeBanjo?
In BeBanjo the search engine indexes the name, external ID, tags, licensor, and metadata for Titles and Series, and the name, external ID, tags, and metadata for Collections. For Assets, the name is indexed.
In Sequence the search engine indexes the title (name), external ID, content provider (licensor), assets, problems, and notes for Jobs.
Reference
movida#3157Replace Sphinx with Elasticsearchsequence#590Replace Sphinx with Elasticsearch