uofast-orm 1.1.0

Simple ORM for U2 Unidata databases

uofast-orm

Simple, fast ORM for U2 Unidata/UniVerse databases via uopy.

Install

pip install uofast-orm

Quick Start

import uopy
from uofast_orm import UopyModel

class Customer(UopyModel):
    _file_name = "CUSTOMERS"
    _field_names = ["NAME", "EMAIL", "PHONE", "ADDRESS"]
    _field_map = {
        "full_name": "NAME",
        "email_address": "EMAIL",
        "phone_number": "PHONE",
        "street_address": "ADDRESS",
    }

with uopy.connect(host="...", user="...", password="...", account="...") as session:
    # Create
    c = Customer(session)
    c.full_name = "Jane Smith"
    c.email_address = "[e-mail]"
    c.create("CUST001")

    # Read
    c = Customer(session, record_id="CUST001")
    print(c.full_name)

    # Update
    c.phone_number = "555-1234"
    c.update()

    # Query
    results = c.select('NAME = "Jane Smith"')

    # Delete
    c.delete()

Models

Subclass UopyModel and set class attributes:

Attribute	Required	Description
_file_name	Yes	U2 file name
_field_names	Yes*	List of DICT field names
_field_map	No	{"python_name": "DICT_NAME"} mapping
_enable_cache	No	Enable LRU cache (default: False)
_cache_max_size	No	Max cached records (default: 100)

* Can be omitted when _field_map is provided.

Named Fields with Alternative DICT

When multiple data files share a single DICT:

from uofast_orm import SmartFile, patch_uopy_file

# Option A: monkey-patch uopy.File globally (zero call-site changes)
patch_uopy_file()

# Option B: use SmartFile directly
with SmartFile("ORDERS") as f:
    rec = f.read_named_fields("ORD001", ["CUSTNO", "AMOUNT"],
                              dict_file="SHARED_DICT")
    f.write_named_fields("ORD001", {"AMOUNT": 999}, dict_file="SHARED_DICT")

Code Generator

Auto-generate model classes from U2 DICT definitions (requires Ollama):

uofast-generate CUSTOMERS \
  --host 192.168.1.10 --user admin --password secret \
  --account /u2/MYACCOUNT \
  --output customer_model.py

Falls back to template-based generation if Ollama is unavailable.

API Reference

UopyModel

Method	Description
load()	Load record from database
create(record_id)	Create new record
update()	Update existing record
delete()	Delete record
save()	Create or update (auto-detects)
get(field, default)	Get field value (supports property or DB name)
set(field, value)	Set field value (supports property or DB name)
to_dict(use_property_names)	Export as dict
select(stmt, limit, offset)	Run SELECT statement, return model list
read(record_id)	Read single record, return model instance
read_many(ids, batch_size)	Batch read, return model list
clear_cache()	Clear LRU cache (classmethod)

Requirements

• Python 3.8+
• uopy (U2 Unidata/UniVerse client)
• requests (for code generator Ollama calls)

License

MIT