Trying to understand identifiers and collections in MATCH and WHERE

Question

I am trying to understand to what kind of cypher "data structure" certain identifiers or expressions correspond to, depending on how and where they are used. Below I list examples I encountered. Please tell me if I got it right (in the comments) or if I am missing something.

MATCH (a:MYTYPE { label:'l_a' })
// a corresponds to a collection of nodes

MATCH (b:MYTYPE { label:'l_b' })
// so does b

MATCH p=(a)-[sp1:CF*]->(b)-[sp12:CF]->(c)
// p corresponds to a collection of paths
// a and b correspond to a collection of nodes 
// (or does the previous MATCH of a and b change something?)
// sp1 corresponds to a collection of collections of relationships
// sp12 corresponds to a collection of relationships
// c corresponds to a collection of nodes

WHERE ( p = ... )
// Here, the p corresponds to a path, i.e. there must be a path or (I don't know) on the right side of the =
WHERE ( a = ... )
// a corresponds to a node, i.e. there must be a node on the right side of the =
WHERE ( sp1 = ... )
// sp1 corresponds to a collection of nodes, i.e. there must be a collection of relationships on the right side

//BONUS:
WHERE ( (e)-[sp2:CF*]->(f) ) = ...
// there must be a collection of collections of paths on the right side of the =

Answer 1

I think the easiest way to answer all these questions is by passing identifiers to functions that will throw an error telling you what it expected and what it actually received. I think you should also be careful about how you use the word collection, as it is not correct.

Nodes, Relationships, Paths

MATCH (n) RETURN n;

n is a Node.

MATCH ()-[r]-() RETURN r;

r is a Relationship.

MATCH p = ()-[]-()

p is a Path.

Collections

MATCH (n) WITH COLLECT(n) AS c RETURN c;

c is a Collection<Node>.

MATCH ()-[r]-() WITH COLLECT(r) AS c RETURN c;

c is a Collection<Relationship>.

MATCH p = ()-[]-() WITH COLLECT(p) AS c RETURN c;

c is a Collection<Path>.

Variable Length Paths

MATCH p = ()-[r*..2]-() RETURN p, r;

p is a Path.

r is a Collection<Relationship>.

And to refer to your specific example:

MATCH p = (a)-[sp1:CF*]->(b)-[sp12:CF]->(c)

p is a Path.

a is a Node.

sp1 is a Collection<Relationship>.

b is a Node.

sp12 is a Relationship.

c is a Node.

And I'm not sure what you're asking about the WHERE clauses. Perhaps you can clarify with an edit to your question.