pyspark.sql.streaming.DataStreamReader.format

DataStreamReader.format(source: str) → pyspark.sql.streaming.readwriter.DataStreamReader

Specifies the input data source format.

New in version 2.0.0.

Parameters

sourcestr

name of the data source, e.g. ‘json’, ‘parquet’.

Notes

This API is evolving.

Examples

>>> spark.readStream.format("text")
<pyspark.sql.streaming.readwriter.DataStreamReader object ...>

This API allows to configure other sources to read. The example below writes a small text file, and reads it back via Text source.

>>> import tempfile
>>> import time
>>> with tempfile.TemporaryDirectory() as d:
...     # Write a temporary text file to read it.
...     spark.createDataFrame(
...         [("hello",), ("this",)]).write.mode("overwrite").format("text").save(d)
...
...     # Start a streaming query to read the text file.
...     q = spark.readStream.format("text").load(d).writeStream.format("console").start()
...     time.sleep(3)
...     q.stop()

pyspark.sql.streaming.DataStreamReader.csv pyspark.sql.streaming.DataStreamReader.json