Cypher In Nexyron

Consumer View

Cypher is the main query language you should think in when you use Nexyron today.

It is not only a graph-pattern language in this repo. The implemented surface mixes:

• graph matching and mutation
• abstraction and feature/intervention/scenario/snapshot/dataset registry surfaces
• temporal clauses for event-style filtering
• document-style map and list processing
• geospatial functions plus temporal-property-based location workflows
• vector and similarity functions
• procedures for projections, search, and graph algorithms

That matters because the useful queries are often hybrid. A real Nexyron query might:

1. match users and documents through graph edges
2. filter interactions by time
3. inspect nested document metadata
4. rank by vector similarity
5. call a graph algorithm or semantic procedure

Current Runtime Reality

The code-grounded rule in this repo is simple:

• embedded execution is currently Cypher-first
• Session::execute_language() and server query handling only accept cypher
• Cypher query text may be one statement or a semicolon-delimited script; scripts execute statements left-to-right and return the final statement result
• script statements can pass the immediately previous result into the next statement with FROM PREVIOUS AS alias
• older comments that imply broad multi-language parity should not be treated as authoritative without code verification
• Nexyron-specific temporal clauses and nexyron.* procedures are extensions on top of that Cypher surface, not evidence of formal openCypher/TCK compliance

How To Read This Tree

Use this documentation in three layers:

• Clauses: the pipeline-building pieces such as MATCH, WITH, CALL, TIME, MERGE, RETURN, statement forms, and DDL clauses
• Functions: scalar and collection helpers for graph, document, temporal, type-conversion, and vector work
• Procedures And Algorithms: row-producing operations such as graph projections, semantic retrieval, and graph analytics

Supporting overviews:

• Clauses overview
• Functions overview
• Procedures and algorithms overview
• Surface inventory and coverage

The only query language documented here is Cypher because that is the only language accepted on the current runtime path.

Why The Documentation Is Structured Per Item

The clauses and procedures in Nexyron are too composable for shallow glossary-style docs to be enough.

For example:

• TIME is only really useful when you understand where it sits relative to pattern matching and aggregation
• CALL matters because it is how algorithms and semantic retrieval re-enter a normal Cypher pipeline
• cosine_similarity matters more when you see how it mixes with graph filters and document metadata
• community algorithms only become useful when you understand how they differ from one another

That is why the per-item pages aim to describe:

• what the feature does
• when to use it
• how it differs from nearby features
• realistic use cases
• real limitations