Ultipa Python Driver
Official Python driver for Ultipa graph database (GQL).
Requirements
- Python 3.10+
Installation
pip install ultipa # base driver
pip install ultipa[ml] # + graph ML data loaders (torch); see gqldb.ml
Note: This is the v6.x driver for Ultipa Graph. If you are using Ultipa v4.x, please install ultipa==4.5.2. If you are using Ultipa v5.x, please install ultipa==5.2.1.
Quick Start
from gqldb import GqldbClient
from gqldb.config import GqldbConfig
# Create client
config = GqldbConfig(
hosts=["localhost:60061"],
username="admin",
password="password",
default_graph="myGraph"
)
client = GqldbClient(config)
client.login("admin", "password")
# Execute query
response = client.gql("MATCH (n) RETURN n LIMIT 10")
print(f"Rows: {len(response.rows)}")
print(f"Columns: {response.columns}")
# Close
client.close()
Features
- GQL query execution with parameters
- Streaming results for large datasets
- Transaction support (begin, commit, rollback)
- Graph management (create, drop, list)
- Bulk import for high-throughput loading
- Algorithm support (CALL algo.pagerank, algo.degree, etc.)
- Graph ML data loaders for PyTorch Geometric / DGL —
gqldb.ml(GraphLoader,NeighborLoader,EdgeLoader) - Type hints support
- Health checks
Errors
Every driver error is a GqldbError. An error from the server carries the
engine's error code in e.code (for example 5020) and a stable name in
e.reason (for example "FULLTEXT_INDEX_LOADING"); both are 0 and "" when
the server sent none. They come from a structured detail on the gRPC error,
never from the message text. EngineCode and ErrorReason name the common
ones, and error_detail(exc) reads them from any exception.
from gqldb import GqldbError, EngineCode
try:
client.gql("INSERT (:Person {name: 'Alice'})")
except GqldbError as e:
if e.code == EngineCode.WRITES_COMMITTED:
pass # stored although the call failed: do not run it again
A gql / gql_stream call with QueryConfig(read_only=True) and no
transaction_id is sent again when the server answers 5020 (a fulltext index
is still loading): up to retry_count times (default 3), waiting retry_delay
(default 0.1 s) and doubling, capped at 2 s, all within the call's timeout.
retry_count=0 turns it off. After signing in again, or when a result is too
large for gql, a call is sent again only if the server refused it before it
ran or it is read_only outside a transaction: a request that may have
written is never sent again. An answer marked partly stored
(partly_stored(e): a statement failed after earlier statements of the
request had committed, or a [5024] answer) is never sent again, a 5020
included; never run such a request again as it is. A write conflict (3011,
WriteConflictError: nothing stored) is never sent again, with or without a
transaction_id: run the whole transaction again yourself, or let
with_transaction(..., retry_on_conflict=N) do it. See
GUIDE.md.
Documentation
See Quick Start for detailed usage.
For the graph ML data loaders (streaming features / subgraphs into PyTorch
Geometric / DGL for GNN training), see gqldb/ml/README.md.
License
MIT License
Metadata
Release files for ultipa 6.2.15
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| ultipa-6.2.15.tar.gz | 209.0 kB | Without provenance | Details for ultipa-6.2.15.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| ultipa-6.2.15-py3-none-any.whl | Python 3 | none | any | Details for ultipa-6.2.15-py3-none-any.whl |
Total release size: 453.7 kB
Release files / ultipa-6.2.15.tar.gz
| Download URL | ultipa-6.2.15.tar.gz |
|---|---|
| Size | 209.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
a4dbd6c1c24f2d4898df3ae012b01a60173f9bde0b7789b1e7d34b9a7b4c0c1b
|
|
BLAKE2b-256 checksum How to use checksums |
ddb4a73de39394abe5c0078a1acb8ad677c985d0f544e40733cc64dba9731a09
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.14
|
Release files / ultipa-6.2.15-py3-none-any.whl
| Download URL | ultipa-6.2.15-py3-none-any.whl |
|---|---|
| Size | 244.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
6a53d726333ab6f7e3736bbc0c0764213ad910e2477ae2c4e76cb1a2fd463c02
|
|
BLAKE2b-256 checksum How to use checksums |
f8fd2926a89d14e112f91d85b5542f943dde362e31de602290d76ad0b9cad25d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.14
|
Security for ultipa 6.2.15
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.