Cypher MATCHでパターンを照合する

Welcome to the core of graph querying! The MATCH clause in Cypher is how you find specific patterns of data in your Neo4j graph.

Think of it like drawing the shape of the data you want to find. You describe nodes, relationships, and their properties, and Neo4j finds all instances that fit your description.

The simplest MATCH query finds any node in your graph. You represent a node using parentheses ().

Adding a variable inside, like (n), lets you refer to that node later, for example, to return its properties. We'll use LIMIT to show just a few results.

Graphs often have different types of nodes. In Cypher, these types are called labels. You can specify a label to find only nodes of a particular type.

Labels are placed after the node variable, separated by a colon: (variable:Label).

Nodes also have properties, which are key-value pairs describing them. You can filter nodes based on these properties directly within your MATCH clause.

Properties are placed inside curly braces {} after the label (or variable if no label): (node:Label {key: 'value'}).

Relationships define how nodes are connected. In MATCH, you represent a relationship using dashes and angle brackets: -- for undirected, --> for directed, and <-- for reverse directed.

You can also assign a variable to the relationship, like -[r]-, to work with it.

Just like nodes have labels, relationships have types. Specifying a relationship type helps you find specific kinds of connections between nodes.

The type is placed after the relationship variable, separated by a colon: -[r:TYPE]->.

Relationships are directional! The arrows in your MATCH pattern indicate the flow. If you omit arrows (e.g., (a)-[r]-(b)), the match will be undirected, finding connections regardless of their stored direction.

It's generally best practice to specify direction when you know it.

The real power of MATCH comes from combining nodes and relationships to describe complex patterns in your graph. You can chain multiple nodes and relationships together.

This allows you to find paths, connections through intermediate nodes, and more.

Sometimes, a part of your pattern might not always exist. OPTIONAL MATCH tries to find the pattern, but if it doesn't, the unmatched parts will be null instead of failing the whole query.

This is useful when you want to see existing connections while still returning the main node even if a specific connection is missing.

Great job! You've learned the fundamental skill of finding patterns in your graph using the MATCH clause.

• Use () for nodes, -- for relationships.
• Specify labels (:Label) for node types and types (:TYPE) for relationships.
• Filter by properties using {}.
• Arrows (-->, <--) define relationship direction.
• OPTIONAL MATCH helps find patterns that might not always exist without breaking the query.

Next, we'll explore how to refine your matched data using WHERE and ORDER BY!