Nemo Logo
Nemo
Docs Examples
Browse documentation
Home / Docs / datatypes
Core Concepts

Reference for Nemo’s native SQL types, storage bounds, aliases, and Python mappings.

Nemo Datatypes Reference

Nemo provides native, first-class support for 22+ standard relational and semi-structured SQL datatypes. All datatypes are strongly typed within the C++20 storage engine and seamlessly map to Python standard library primitives during queries, parameter bindings, and result fetching.

For a full guide to SQL functions for these types, see the SQL Built-in Functions Reference.

1. Complete Datatypes Reference Table

SQL Type Name Accepted Aliases Storage / Bounds Python Mapping Example Value
NULL — SQL Null representation None None
BOOLEAN BOOL 1 byte (true / false) bool True, False
TINYINT INT1 8-bit signed integer (-128 to 127) int 42
SMALLINT INT2, SHORT 16-bit signed integer (-32,768 to 32,767) int 1024, -32000
INTEGER INT, INT4 32-bit signed integer (-2.14B to 2.14B) int 100000
BIGINT INT8, LONG 64-bit signed integer (-9.22E18 to 9.22E18) int 9223372036854775807
REAL FLOAT, FLOAT4 32-bit IEEE 754 Single-Precision Float float 3.14159
DOUBLE PRECISION DOUBLE, FLOAT8 64-bit IEEE 754 Double-Precision Float float 2.718281828459045
NUMERIC(p, s) DECIMAL(p, s), DEC Fixed-point exact decimal representation decimal.Decimal / float 123456.78
VARCHAR(n) CHARACTER VARYING Variable-length string str 'code_alpha_99'
TEXT STRING Arbitrary-length UTF-8 encoded text str 'Unicode string text'
CHAR(n) BPCHAR(n) Fixed-length blank-padded character string str 'ABCDE'
BYTEA BLOB, BINARY Raw binary byte sequence bytes / bytearray b'\xde\xad\xbe\xef'
DATE — ISO Calendar Date (YYYY-MM-DD) datetime.date / str '2026-10-25'
TIME — Clock Time (HH:MM:SS) datetime.time / str '14:30:00'
TIMESTAMP DATETIME ISO Timestamp (YYYY-MM-DDTHH:MM:SSZ) datetime.datetime / str '2026-10-25T14:30:00Z'
TIMESTAMPTZ — ISO Timestamp with timezone offset datetime.datetime / str '2026-10-25T14:30:00+00:00'
INTERVAL — Elapsed duration / timespan datetime.timedelta / str '1 day 02:00:00'
UUID GUID 128-bit RFC 4122 UUID identifier uuid.UUID / str 'a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a11'
ARRAY T[] Ordered collection of elements list [1, 2, 3, 4, 5]
JSON — Structured JSON text with syntax verification dict, list, str {"user": "alice", "active": true}
JSONB — Binary decomposed JSON for high-speed extraction dict, str {"score": 98.5, "tags": ["db"]}
DICT MAP, HSTORE String-keyed key-value dictionary dict {"tier": 1, "status": "active"}
ENUM — Constrained set of named string constants str 'active', 'pending', 'closed'
BIT(n) — Fixed-length bit string str '10101010'
VARBIT(n) BIT VARYING Variable-length bit string str '110011'
INET — IPv4 or IPv6 host address str '192.168.1.1', '2001:db8::1'
CIDR — IPv4 or IPv6 network CIDR block str '10.0.0.0/24'
MACADDR — 48-bit Media Access Control hardware address str '08:00:2b:01:02:03'
RANGE INT4RANGE, etc. Contiguous interval with bounds [lower, upper) str '[1, 100)'

2. Structured Collection Types Comparison

Feature ARRAY DICT JSON / JSONB
Data Format Ordered list of items Flat string-keyed map Recursive multi-type document
Element Types Homogeneous values Flat key-value pairs Arbitrary mixed types
Nesting Linear list or matrix Single-level map Multi-level nested objects & arrays
Speed Profile Fastest vector math & index slicing O(1) attribute lookup Flexible path extractions
Ideal For Time series, tags, sensor metrics Entity attributes, key-value lookup REST APIs, payloads, configs

3. Dedicated Functions by Datatype Category

Nemo provides specialized SQL functions for each complex datatype. Detailed references and signatures are available in the SQL Built-in Functions Reference:

  • JSON & JSONB: json_extract(), json_value(), json_set(), json_insert(), json_remove(), json_keys(), json_length(), json_contains(), json_merge(), jsonb_extract(), jsonb_value(), jsonb_set().
  • ARRAY: array_length(), array_get(), array_slice(), array_append(), array_remove(), array_contains(), array_distinct(), array_sort(), array_sum(), array_avg().
  • DICT: dict_get(), dict_has_key(), dict_keys(), dict_values(), dict_size(), dict_set(), dict_remove(), dict_merge().
  • RANGE: range(), range_lower(), range_upper(), range_contains().
  • BIT: bit_length(), bit_count(), bit_and(), bit_or(), bit_xor(), bit_not().
  • NETWORK: host(), netmask(), is_ipv4(), is_ipv6().
  • UUID & BINARY: uuid_generate(), uuid_to_string(), hex_encode(), hex_decode(), base64_encode(), base64_decode().
  • example_3.py: Comprehensive 30 Native Datatypes Demonstration.
  • example_15.py: Numeric Datatypes Deep-Dive (TINYINT to BIGINT, REAL, DOUBLE).
  • example_16.py: Exact Financial Decimals (DECIMAL(p, s)).
  • example_17.py: Text, International Unicode, and Raw Binary Blobs (BYTEA).
  • example_18.py: Temporal Types & Timestamps (DATE, TIME, TIMESTAMP, INTERVAL).
  • example_19.py: Semi-Structured JSON & JSONB Documents.
  • example_20.py: Complex Collections: ARRAY and DICT.
  • example_21.py: System & Network Types (UUID, INET, CIDR, RANGE).
← Previous Article
Nemo Transactions Reference
Next Article →
Nemo SQL Reference & Grammar