03-augmented-search

Augmented-Search tutorial

Building on top of the previous tutorial, we're going to:

• start an Elasticsearch single-node cluster
• Install augmented-search
• Configure the database connection via a groovy script
• Configure augmented-search to use that connection
• Enable augmented-search on digitall
• Index Digitall

Instructions

To get started, execute the following commands:

cd ~
git clone https://github.com/Jahia/provisioning-tutorials.git
cd provisioning-tutorials/03-augmented-search
docker-compose up --renew-anon-volumes

While the elasticsearch cluster is starting, you will first see the MariaDB container booting up and Jahia creating the necessary tables and continue with its startup.

After startup

At the end of startup (give it a minute or two), open a browser to Digitall's home page at http://localhost:8080.

Fiddling with GraphQL API

You can also play with our GraphQL API:

Query: As an authenticated user (root), fetch 2 hits across all documents (no search terms) in the EDIT workspace

curl --request POST \
  --url http://localhost:8080/modules/graphql \
  --header 'Content-Type: application/json' \
  --header 'Origin: http://localhost:8080' \
  --header 'authorization: APIToken kgHNm05iQV61I+GY3X5HVr13i866HAAsyou8G+eGubk=' \
  --data '{"query":"query {\n  search(q: \"\", workspace: EDIT) {\n    results(size: 2) {\n      hits {\n        displayableName\n      }\n    }\n  }\n}"}'

Response:

{"data":{"search":{"results":{"hits":[{"displayableName":"Singh"},{"displayableName":"Digitall Network Expands To Transportation Industry"}]}}}}

You might notice that we searched in the EDIT workspace, which is only accessible to an authenticated user, look at what happens if you run the same query without providing a token:

Query: As guest, fetch 2 hits across all documents (no search terms) in the EDIT workspace

curl --request POST \
  --url http://localhost:8080/modules/graphql \
  --header 'Content-Type: application/json' \
  --header 'Origin: http://localhost:8080' \
  --data '{"query":"query {\n  search(q: \"\", workspace: EDIT) {\n    results(size: 2) {\n      hits {\n        displayableName\n      }\n    }\n  }\n}"}'

Response:

{"data":{"search":{"results":{"hits":[]}}}}

Now try changing the workspace to LIVE, which should work fine as guest.

Query: As guest, fetch 2 hits across all documents (no search terms) in the LIVE workspace

curl --request POST \
  --url http://localhost:8080/modules/graphql \
  --header 'Content-Type: application/json' \
  --header 'Origin: http://localhost:8080' \
  --data '{"query":"query {\n  search(q: \"\", workspace: LIVE) {\n    results(size: 2) {\n      hits {\n        displayableName\n      }\n    }\n  }\n}"}'

Response:

{"data":{"search":{"results":{"hits":[{"displayableName":"all-Movies"},{"displayableName":"all-Med"}]}}}}

PS: Since we're not searching on a particular search term, the exact result order between the two workspaces might differ (as they do in the example above).

What did we learn ?

In this tutorial got much closer to a complex real-world scenario in which we took a released version of Jahia and customized it to our own use case by installing additional modules, and performing some GraphQL queries.

This provisioning script uses a few new commands when compared to the previous tutorial:

• include to execute another provisioning script. As you probably noted we were able to run the provisioning script from Tutorial #2 by including it in this script.
• addMavenRepository to install modules from another Nexus repository (in this particular case, Jahia Store)
• enable to enable a set module with a set siteKey

Next

In the next tutorial, we're going to add jExperience and forms to the mix click here.