Using Azure Search

When in need of using a search engine, it's always a good idea to use tested and well-known solutions that have been available on the market for a while. One of those solutions is Azure Search, which offers a search-as-a-service cloud solution with an API for developers, so users can focus on developing a working solution without the need to manage infrastructure or configuration. With the recent addition of the Cognitive Search API, we've been given an opportunity to enhance our search functionality with AI features, so it's possible to transform unstructured content into searchable content.

The following topics will be covered in this chapter:

• Using Azure Search in your projects
• Using full-text search for your needs
• Using linguistics analysis
• Using indexing, index definition, and indexers
• The new Cognitive Search API with indexing workloads...

Technical requirements

To perform exercises from this chapter, you will need the following:

• Access to an Azure subscription
• A tool for sending HTTP requests (cURL and Postman)

Creating an Azure Search service

Creating Azure Search is an easy task and should not take more than a few minutes. The important thing to remember here is that, if you have to recreate a service, there is no way to back up and restore data—if you make mistakes here, everything has to be done from scratch.

The reason to recreate would be changing the datatypes or other search attributes of the search index columns. While it will not be a problem during this exercise, have that in mind when creating a production workload.

Using Azure Portal

As with most services in Azure, we will start our journey by creating Azure Search in the portal. To do so, click on + Create a resource and search for Azure Search. On the introduction...

Full-text search in Azure Search

The power of Azure Search comes when you need to perform a full-text search to find relevant documents that will satisfy your query. This Azure service uses Apache Lucene under the hood, which is a well-known, high-performance search engine written in Java.

You can find more information about Lucene here: https://lucene.apache.org/core/. It is an open source project available to download for everyone.

In this chapter, you will learn how to perform a full-text search, what is the syntax, and how to recognize potential issues.

Sending a request

In the first section of this chapter, you created your Azure Search instance and saw Search explorer, which enables you to send simple queries. Now, we...

Linguistic analysis in full-text search

When using full-text search, you have to understand what are the rules for performing such operations. Search engines have to analyze search query lexically to be able to extract important information and pass it to a query tree. In this section, we will cover the most common lexical analysis, which is linguistic analysis, to help you understand how Azure Search works and how to perform correct queries.

Analyzers in Azure Search

To perform linguistic analysis, Azure Search supports a variety of different analyzers, which can be specified in the index definition. However, before we start defining one of those, you have to catch a glimpse of what we are talking about. When creating an...

Indexing in Azure Search

An index is one of the most important constructs in Azure Search. We defined it as a table that contains all imported documents with searchable data defined in it. At the beginning of this chapter, you learned how to create it and add or edit fields. In this section, you will learn a bit more about modifying it as an index is not a fixed being that cannot be altered and adjusted to your needs.

Importing more data

There is always a need to push more data to your index—your application grows, the storage of your documents becomes bigger and bigger, especially if you are creating a document repository, and you would like to be able to find what you are searching for even in the most recent documents...

Cognitive search – adding AI to the indexing workload

When creating and managing an index, you always have to make sure that you have selected all required fields and marked them as searchable where needed. Additionally, we are limited to the current service functionalities, so we cannot use things such as image or natural language processing. Fortunately, Azure Service is about to start supporting the Cognitive Search feature, which adds AI to your indexing operations, allowing for richer analysis using more vectors. In this chapter, you will learn how to configure it, so you can start using it from the very beginning.

Configuring Cognitive Search

Cognitive Search can be configured when importing data for your service...

Summary

Azure Search is a great service when you want to have your very own search solution and do not plan to maintain its infrastructure and configuration. With its flexibility and intuitiveness, you can quickly develop your application, leveraging features such as the push/pull model, scheduled indexing, or support for different kinds of data sources. Additionally, with the option to start from the free tier even for production workloads, you can gradually progress and scale your solution up adjusting costs based on real requirements. In Chapter 6, Mobile Notifications with Notification Hub, we will cover topics related to handling mobile and applications and push notification by using Azure Notification Hub.

Questions

1. What is an index?
2. What is the difference between the push and pull models?
3. Can an indexer be scheduled using a custom interval?
4. What analyzer does Azure Search use by default?
5. Can one implement a custom analyzer and use it in Azure Search?
6. What is the difference between partition and replica?
7. What is the name of a header used for authorizing requests to Azure Search?

Further reading

• Azure Search documentation: https://docs.microsoft.com/en-us/azure/search/
• Cognitive Skills for Azure Search: https://docs.microsoft.com/en-us/azure/search/cognitive-search-predefined-skills
• Apache Lucene documentation: https://lucene.apache.org/core/
• Filters in Azure Search: https://docs.microsoft.com/en-us/rest/api/searchservice/odata-expression-syntax-for-azure-search