HyukjinKwon commented on code in PR #46139:
URL: https://github.com/apache/spark/pull/46139#discussion_r1573469889
##########
python/docs/source/user_guide/sql/python_data_source.rst:
##########
@@ -84,6 +87,46 @@ Define the reader logic to generate synthetic data. Use the
`faker` library to p
row.append(value)
yield tuple(row)
+** Implement the Stream Reader**
+
+.. code-block:: python
+
+ class RangePartition(InputPartition):
+ def __init__(self, start, end):
+ self.start = start
+ self.end = end
+
+ class FakeStreamReader(DataSourceStreamReader):
+ current = 0
+
+ def initialOffset(self):
+ return {"offset": 0}
+
+ def latestOffset(self):
+ self.current += 2
+ return {"offset": self.current}
+
+ def partitions(self, start, end):
+ return [RangePartition(start["offset"], end["offset"])]
+
+ def commit(self, end):
+ pass
+
+ def read(self, partition):
+ start, end = partition.start, partition.end
+ for i in range(start, end):
+ yield (i, str(i))
+
+This is a dummy streaming data reader that generate 2 rows in every
microbatch. The streamReader instance has a integer offset that increase by 2
in every microbatch.
+
+initialOffset() should return the initial start offset of the reader.
Review Comment:
```suggestion
``initialOffset()`` should return the initial start offset of the reader.
```
or
```suggestion
:meth:`pyspark.sql.datasource.DataSourceStreamReader.initialOffset`
```
or
```suggestion
.. currentmodule:: pyspark.sql.datasource
:meth:`DataSourceStreamReader.initialOffset`
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
