pancham 0.19.3

No project description provided

Pancham Data Pipelines

Pancham simplifies the process of reading and processing data with Pandas.

What is it for

A common scenario in data migration projects is to need to take exports from one system, apply some transformation rules and write the data somewhere else. This could be files, a database or an API. All of this can be done with Pandas but we would need to write a lot of code, Pancham is here to make that process easier.

flowchart LR;
    s[Source Files]-->t[Transformation];
    t-->o[Output Files];

Key features

• Load files
• Rename fields
• Validate data types and not-null columns
• Apply python functions to transform values
• Return data as a Pandas DataFrame
• Write data to a database

Supported Source Files

• Excel
• SQL
• YAML
• CSV
• JSON

Supported Output

• SQL
• CSV

Additional sources and output formats will be added with time.

Usage

The most common approach is to use a mapping file.

name: example
file_type: yaml
file_path: source/customer.yml
key: customer
output:
  - table: customers
    output_type: database
fields:
  - name: id
    source_name: customer_id
    field_type: str
    nullable: false
  - name: status
    source_name: account_status
    field_type: str
    nullable: false

This simple mapping file expects to be able to load a source file called customer.yml with the following content:

customer:
    - customer_id: 123
      account_status: active
    - customer_id: 456
      account_status: on_hold

The pipeline wll read the input file and populate a database that you can then query.

The final part is the pancham configuration.