zyxdb 0.1.20

Python bindings for ZYX graph database engine

zyxdb

Python bindings for ZYX, a high-performance embeddable graph database engine with Cypher query support.

Installation

pip install zyxdb

Build from source

cd bindings/python
pip install -e ".[test]"

Quick Start

import zyxdb

# Open a database (creates if not exists, auto-closes on exit)
with zyxdb.Database("/tmp/movies") as db:

    # Create nodes — returns node ID (int)
    alice = db.create_node("Person", {"name": "Alice", "age": 30})
    bob   = db.create_node("Person", {"name": "Bob",   "age": 25})
    carol = db.create_node("Person", {"name": "Carol", "age": 35})

    # Create edges between nodes
    db.create_edge(alice, bob,   "FRIENDS_WITH")
    db.create_edge(bob,   carol, "FRIENDS_WITH")

    # Cypher works too — with parameterized queries
    db.execute("CREATE (m:Movie {title: $title, year: $year})",
               title="The Matrix", year=1999)

    # Transactions — all or nothing
    with db.begin_transaction() as tx:
        tx.execute(
            "MATCH (p:Person {name: 'Alice'}), (m:Movie {title: 'The Matrix'}) "
            "CREATE (p)-[:RATED {score: 9.5}]->(m)"
        )
        tx.execute(
            "MATCH (p:Person {name: 'Bob'}), (m:Movie {title: 'The Matrix'}) "
            "CREATE (p)-[:RATED {score: 10.0}]->(m)"
        )
        tx.commit()

    # Query with iteration
    for row in db.execute(
        "MATCH (p:Person)-[r:RATED]->(m:Movie) "
        "RETURN p.name AS person, m.title AS movie, r.score AS score"
    ):
        print(f"{row['person']} rated {row['movie']} {row['score']}/10")

    # Read-only transaction for safe reads
    with db.begin_read_only_transaction() as tx:
        for row in tx.execute("MATCH (p:Person) RETURN count(p) AS cnt"):
            print(f"Total people: {row['cnt']}")

    # Shortest path
    path = db.get_shortest_path(alice, carol)
    names = [n.properties["name"] for n in path]
    print(" -> ".join(names))  # Alice -> Bob -> Carol

Supported Value Types

Python	ZYX
None	null
bool	bool
int	int64
float	double
str	string
list[float]	vector (embeddings)
list[str]	string list
list (mixed)	heterogeneous list
dict	map

API Reference

Database(path, *, open=True)

• execute(cypher, **params) — Execute Cypher query with keyword parameters
• begin_transaction() — Start a read-write transaction
• begin_read_only_transaction() — Start a read-only transaction
• create_node(label, props) — Create a node, return its ID (label can be str or list[str])
• create_nodes(label, props_list) — Batch create nodes, return list of IDs
• create_edge(src_id, dst_id, edge_type, props) — Create edge between two nodes by ID
• create_edges(edge_type, edges) — Batch create edges
• get_shortest_path(start_id, end_id, max_depth=15) — Find shortest path, return list of Node
• bfs(start_id, visitor) — Breadth-first traversal
• save() — Flush to disk
• close() — Close the database
• has_active_transaction — Whether there is an active transaction

Transaction

• execute(cypher, **params) — Execute within transaction
• commit() — Commit changes
• rollback() — Roll back changes
• is_active — Whether transaction is still active
• is_read_only — Whether read-only

Result

• Iterable: for row in result yields Record objects
• columns / column_names — List of column names
• is_success — Whether query succeeded
• error — Error message or None
• duration — Query execution time (ms)
• fetchone() — Fetch next row or None
• fetchall() — Fetch all remaining rows
• fetchmany(size) — Fetch up to size rows
• single(strict=True) — Return the single result row
• scalar() — Return first column of first row
• data() — Return all rows as list of dicts
• to_df() — Convert to pandas DataFrame

License

Apache License 2.0