pyflink.multimodal.operators.is_valid_audio#

is_valid_audio(*columns, validation='metadata', concurrency=None)[source]#

Check whether an audio input looks like valid audio.

The default validation="metadata" only reads enough container/header data to probe metadata. This is the cheap path for filtering large tables before decode. validation="decode" skips the metadata pre-probe and fully decodes the stream; it is more expensive but catches truncated or corrupt payloads that have a readable header.

Parameters

*columns – Optional audio columns. When provided, the UDF is applied directly.
validation – "metadata" for header/container validation, or "decode" for full decode validation.
concurrency – UDF concurrency. None uses the framework default.

Returns

A UDF producing BOOLEAN. Corrupt supported inputs return False. URI access errors still fail fast.

Raises

ValueError – If validation is unsupported, or if a row value has an unsupported input kind. Supported inputs are BYTES, URI STRING, and AUDIO_CLIP_REF.

Examples::

>>> # Usage 1: create a reusable UDF and apply it to a column.
>>> valid = is_valid_audio(validation="metadata")
>>> df = df.filter(valid(col("audio_bytes")))
>>>
>>> # Usage 2: pass the column directly when building the expression.
>>> df = df.filter(
...     is_valid_audio(col("audio_bytes"), validation="metadata")
... )