Skip to content

SchemaInspector API

SchemaInspector discovers your ArangoDB structure at runtime. See the Schema Introspection guide for usage patterns.

from odin import SchemaInspector, inspect_arango_schema

Constructor

SchemaInspector(db, max_sample_docs: int = 5)
Parameter Type Default Description
db StandardDatabase required A connected python-arango database
max_sample_docs int 5 Documents sampled per collection to discover fields

get_schema_map

get_schema_map(refresh: bool = False) -> dict

Returns the full schema map. Cached after the first call; pass refresh=True to rebuild.

schema = inspector.get_schema_map()
schema["database_name"]   # str
schema["collections"]     # list of document collections
schema["edges"]           # list of edge collections

Shape:

{
    "database_name": "my_graph",
    "collections": [
        {"name": "ExtractedEntities", "type": "document",
         "count": 12345, "fields": ["_key", "name", "type", ...]},
        ...
    ],
    "edges": [
        {"name": "ExtractedRelationships", "count": 45678,
         "from_collections": ["ExtractedEntities"],
         "to_collections": ["ExtractedEntities"],
         "fields": ["_from", "_to", "relation", ...]},
        ...
    ],
}

System collections (names starting with _) are skipped.

get_collection_info

get_collection_info(collection_name: str) -> dict | None

Returns the schema entry for a document or edge collection, or None if not found.

info = inspector.get_collection_info("ExtractedEntities")
info["fields"]    # discovered field names

get_edge_info

get_edge_info(edge_collection: str) -> dict | None

Returns the schema entry for an edge collection, including its from_collections and to_collections.

edge = inspector.get_edge_info("ExtractedRelationships")
edge["from_collections"]   # e.g. ["ExtractedEntities"]
edge["to_collections"]     # e.g. ["ExtractedEntities"]

inspect_arango_schema

inspect_arango_schema(db, output_file: str = "schema.json") -> dict

Convenience one-call helper: builds the schema map and optionally writes it to output_file as JSON.

from odin import inspect_arango_schema

inspect_arango_schema(db, output_file="schema.json")

Use the exported file to prime an agent, generate documentation, or validate structures across environments.

Data classes

The inspector returns plain dictionaries, backed internally by these dataclasses:

Dataclass Fields
CollectionSchema name, type, count, fields
EdgeSchema name, count, from_collections, to_collections, fields
SchemaMap database_name, collections, edges