Reference

Use this section for API and format lookup.

For runnable end-to-end samples, see the checked-in examples/ directory and the examples guide. From the repo root, run them with wow run examples/<pipeline>.yaml --base-dir examples.

Sections

Sources

Start here when you need to understand pipeline inputs.

The sources reference covers:

• Source(...) constructor arguments
• inline schema and CSV reader options
• source-type inference and current v0 constraints
• schema inference behavior

Go to: Sources

Transforms

Start here when you need to understand row and column operations inside a pipeline.

The transforms reference covers:

• cast, select, derive, filter, drop, string, validate, and join
• per-transform signatures
• Python and YAML examples
• argument-by-argument explanations
• the shared Expression DSL

Go to: Transforms

Transform Object

Start here when you need to understand the public Transform(...) constructor used in Python.

This page covers:

• Transform(op, params=None, output_schema_override=None)
• how Transform relates to the per-operation transform catalog
• schema override behavior in locked pipelines

Go to: Transform Object

Pipeline

Start here when you need to understand how WowData workflows are composed and executed.

The pipeline reference covers:

• Pipeline(...) construction
• then(...), run(), and preflight()
• schema inference and schema locking
• YAML and IR serialization methods
• PipelineContext

Go to: Pipeline

YAML / IR Format

Start here when you need to understand WowData's serialized pipeline format.

This page covers:

• top-level YAML structure
• source, transform, and sink descriptor shapes
• normalization rules applied during load
• the relationship between YAML and internal IR dictionaries

Go to: YAML / IR Format

CLI

Start here when you need to run WowData from the shell.

The CLI reference covers:

• run
• validate
• schema
• lock-schema
• common flags and exit codes

Go to: CLI

Sinks

Start here when you need to understand pipeline outputs.

The sinks reference covers:

• Sink(...) constructor arguments
• output-type inference and current v0 constraints
• writer options and directory requirements
• sink write behavior and failure modes

Go to: Sinks

Errors

Start here when you need to interpret WowData error codes.

The errors reference covers:

• common user-facing error codes
• the major source, transform, pipeline, and sink failure families

Go to: Errors

Repository Copy

A compact repository-level version is also available at REFERENCE.md.