Description:
Retrieve data from a partition-based Parquet file.
Syntax:
f.parallel([col,…]; [partitionFilter/...]; [colFilter]; [n])
Note:
External library function (See External Library Guide).
The function retrieves data from a partition-based Parquet file with segment-based parallel processing; data in each partition should be ordered.
Parameter:
|
f |
A file object. |
|
col |
To-be-retrieved fields; return all fields by default. |
|
partitionFilter |
A partition filtering condition, which has a k=v structure, such as year=2024 and year=2023/month=10. |
|
colFilter |
A field filtering condition, which uses a comparison operator such as >,>=,<,<=,=,!=, not, in and like; this parameter becomes invalid when @v option works. |
|
n |
A positive integer representing the number of records to be retrieved; return all records when this parameter is absent. This parameter becomes invalid when @c option works. |
Option:
|
@c |
Return a cursor. |
Return value:
Table sequence
Example:
|
|
|
|
|
|
=file("F:/tmp/mytest.parquet") |
Open a local Parquet file. |
|
|
=A1.parallel() |
Retrieve data from A1’s file and return all fields. |
|
|
=A1.parallel@c() |
Return a cursor. |
|
|
|
|
|
|
|
Retrieve the specified fields from the partition where year=2023 and month=10 and only return the first 10 records meeting the filtering condition. |