Elasticsearch Concepts

Main Features

Distributed & Real-time: Data is indexed and searchable almost instantly across a cluster of nodes.
RESTful API: Interacts with the data using standard HTTP methods and JSON.
Full-Text Search: Powered by Apache Lucene, it handles complex text searches (fuzzy matching, highlighting, stemming).
Document-Oriented: Data is stored as JSON documents, making it schema-less (though schemas can be defined).
Inverted Index: Instead of storing strings, it stores an index of terms and which documents contain them, making searching extremely fast.
Scalability: Sharding and replication allow it to handle massive amounts of data and provide high availability.

Analyzers are used during indexing and searching to process text. An analyzer consists of:

Character Filters: Modify the text before tokenization (e.g., removing HTML tags).
Tokenizer: Splits the text into tokens (usually words).
Token Filters: Modify tokens (e.g., lowercasing, removing stop words like "the" or "and", stemming).

The default analyzer that works well for most languages. It tokenizes on word boundaries and removes punctuation.

You can define your own analyzers to handle specific cases, like indexing email addresses or product IDs.