ALS, WebGenAI Docker updates

This release is available via pip and docker, including als and WebGenAI.

WebGenAI now supports a 'demo mode startup' that will create and run the genai_demo, which illustrates basic logic (note: the doc is generally correct, but is under revision):

# run local docker
https://apifabric.ai/admin-app/#/Home?demo=genai_demo

# WebGenAI website
http://localhost:8282/admin-app/#/Home?demo=genai_demo

To obtain the WebGenAI Docker:

docker pull apilogicserver/web_genai

Docker run instructions are provided on the WebGenAI Docker Repository.

14.01.00: N8N, Fixup missing attrs, Rebuild Test Data, Project Import/Merge

This major release is available via PyPi and Docker. It contains bug fixes, and new features listed below.

1. N8N

N8N enables integration with a wide range of systems. It is included as part of the nw_sample. Include code like the following in logic/declare_logic.py (or in logic/logic_discovery:

        if logic_row.is_inserted():
            send_n8n_message(logic_row=logic_row)

Simplified RowDictMapper

The sample illustrates usage such as:

    def fn_customer_workflow(row: Customer, old_row: Customer, logic_row: LogicRow):
        """ #als: Send N8N message 
        Webhook Workflow:  When Customer is inserted/updated - post n8n Webhook and call sendgrid email system

        Args:
            row (Customer): inserted / changed Customer
            old_row (Customer): n/a
            logic_row (LogicRow): bundles curr/old row, with ins/upd/dlt logic
         """
        
        if logic_row.is_inserted():
            send_n8n_message(logic_row=logic_row)



        """ #als: Send Kafka message - illustrate 3 alerting methods 
        See also:  send_kafka_message() in declare_logic.py.
        """
        

        kafka_producer.send_kafka_message(logic_row=logic_row,
                                            kafka_topic="customer",
                                            kafka_key=str(row.Id),
                                            msg="1. /integration.py: Kafka, sending logic_row=logic_row")

        kafka_producer.send_kafka_message(payload={"customer_id": row.Id, "balance": row.Balance},
                                            kafka_topic="customer",
                                            kafka_key=str(row.Id),
                                            msg="2. logic_discovery/integration.py: Kafka, sending payload=<dict>")

        kafka_producer.send_kafka_message(logic_row=logic_row,
                                            row_dict_mapper=Customer_Orders,
                                            kafka_topic="customer",
                                            kafka_key=str(row.Id),
                                            msg="3. logic_discovery/integration.py: Kafka, sending with declarative mapping (row_dict_mapper=Customer_Orders)")


    def fn_employee_workflow(row: Employee, old_row: Employee, logic_row: LogicRow):
        """
        Workflow:  When Employee is inserted = post n8n Webhook and call sendgrid email system
        """
        send_n8n_message(logic_row=logic_row,
                msg="1. /Webhook integration.py: n8n, sending Employee logic_row")        

 

2. Fixup Missing Attributes

Introducing new rules often requires adding attributes. These are now reported when you start your API Logic Project.

For WebG projects, you can use GenAI to discover missing attributes from your logic, and add them to your test sqlite database:

# Ask ChatGPT to rebuild the data model and test data - create missing attrs per the current rules
als genai-utils --fixup
# see results in genai_demo_with_logic/docs/fixup

# create a new project with the correct data model & test data
cd ..  # should be the manager
als genai --using=genai_demo_with_logic_fixed --project-name=genai_demo_with_logic_fixed --retries=-1 --repaired-response=genai_demo_with_logic/docs/fixup/response_fixup.json

 

3. Rebuild Test Data

Introducing new rules often requires adding attributes.. These need to be initialized with proper values that reflect the rules.

For WebG projects, you can use GenAI to update sqlite test databases:

cd <your project>
als genai-utils --rebuild-test-data 

 

4. Project Import/Merge

For multiple team (e.g. Line Of Business and Dev) working in parallel on the same project, LOB updates can be merged into the Dev project:

als genai-utils --import-genai --using=../wg_demo_no_logic_fixed

where --using is a project exported / unzipped from WegGenAI.

GenAI Logic with "any" cardinality rules

This release is available for PyPi and Docker, both apilogicserver/api_logic_server and apilogicserver/web_genai.

There are many improvements to GenAI, including the Landing Page with fun graphics (see the video at GenAI-Logic, such as support for "any" cardinality rules.

GenAI Informal and Multi-Rule Logic

This release (pip and docker) sees signficant advances in declaring logic via GenAI.

For example, instead of dot notation, you can create prompts like this:

Create a system with customers, orders, items and products.

Include a notes field for orders.

Use LogicBank to enforce the Check Credit requirement:
    1. The Customer's balance is less than the credit limit
    2. The Customer's balance is the sum of the Order amount_total where date_shipped is null
    3. The Order's amount_total is the sum of the Item amount
    4. The Item amount is the quantity * unit_price
    5. The Item unit_price is copied from the Product unit_price

You can even infer multiple rules from a single line; create a project with a prompt like this:

System for Departments and Employees.

LogicBank:
1. Sum of employee salaries cannot exceed department budget

This creates a running system: a database, an API, an Admin App, and logic. From the Manager:

als genai --using=system/genai/examples/emp_depts/emp_dept.prompt

The logic is non-trivial:

1. A Department.total_salaries is created
2. Two rules are created in logic/declare_logic.py

    # Logic from GenAI: (or, use your IDE w/ code completion)

    # Aggregate the total salaries of employees for each department.
    Rule.sum(derive=Department.total_salaries, as_sum_of=Employee.salary)

    # Ensure the sum of employee salaries does not exceed the department budget.
    Rule.constraint(validate=Department, as_condition=lambda row: row.total_salaries <= row.budget, error_msg="Total salaries ({row.total_salaries}) exceed the department budget ({row.budget})")

    # End Logic from GenAI

This support is in technology-preview state.

DBML with Table Descriptions

This is mainly an internal release (PyPy and docker):

1. SRA (Safrs React Admin) 10-22
2. For projects created with GenAI, the DBML now shows the table descriptions

Natural Language Logic - Beta

Natural Language Logic has matured to the Beta stage. There are still occasional issues with proper initialization of derived columns, but is usually correct. It's a good idea to check your test data...

This release (pip and docker) also addresses:

1. Improved Singularization: table names like Address now result in classes like Address (not Addres)
2. Logic Learning is much-improved, with a new customizable /system/genai/learning_requests
3. Improved provisions for logic hallucinations in creating logic
   4, You can use Natural Language to add logic to existing projects, though see the provisos noted in the doc.
4. Projects created from GenAI now show "completed' prompts, ie, include system-supplied prompt engineering, which makes them easier to understand and debug
5. The project readme is improved

Natural Language Logic

As of release 11.2.10, you can declare Natural Language Logic when you create projects, and for existing projects.

This is a technology preview, not yet available in docker or Web/GenAI.

Bug Fix - Constraint Reporting in Admin App

Fixes improperly reported constraints - available on PyPy and docker.

Dup Rule Detection; GenAI Hardening

This release fixes dup rule detection, and some new coding styles from ChatGPT that previously caused errors, including:

Decimal Test Data

invoice1 = Invoice(client_id=client1.id, project_id=project1.id, amount=Decimal('1000.0'), issued_date=datetime.datetime.utcnow())

This is fixed by adding import decimal to system/genai/create_db_models_inserts/create_db_models_prefix.py.

Improper syntax for datetime

time_entry2 = TimeEntry(employee_id=2, task_id=2, start_time=datetime.datetime.now(), end_time(datetime.datetime.now() + datetime.timedelta(hours=3), hours_worked=3.0)

The missing equal sign in end_time(datetime.datetime.now() is specifically checked and repaired.

Database Diagrams, GenAI Conversations

This release provides:

• Database Diagrams
• GenAI Enhancements

It is available via both pip and docker.