screencastgen

Alignment Registry

Registry for word-level alignment providers.

Source: screencastgen/providers/align/__init__.py, screencastgen/providers/align/base.py

Overview

Manages alignment providers that take audio + text and produce word-level timing data (WordTiming). Currently only WhisperX is registered.

Functions

Function	Description
`get_alignment_provider_names() -> List[str]`	List registered providers
`get_default_alignment_provider() -> str`	Returns `"whisperx"`
`get_alignment_provider_spec(name) -> AlignmentProviderSpec`	Get provider spec
`align_with_provider(provider, audio_path, text, language, device) -> List[WordTiming]`	Dispatch alignment to named provider

Dataclass: `AlignmentProviderSpec`

Defined in base.py. Metadata for an alignment provider.

Registered Providers

Name	Module	Description
`whisperx`	WhisperX Provider	WhisperX word-level alignment

Default: whisperx

Dependencies

Alignment Registry
├── WhisperX Provider  (lazy import)
├── AlignmentProviderSpec  (base.py)
└──▶ consumed by Aligner
     └──▶ consumed by Pipeline Common