2015-05-19 17:23:28 -04:00
|
|
|
#
|
|
|
|
# Licensed to the Apache Software Foundation (ASF) under one or more
|
|
|
|
# contributor license agreements. See the NOTICE file distributed with
|
|
|
|
# this work for additional information regarding copyright ownership.
|
|
|
|
# The ASF licenses this file to You under the Apache License, Version 2.0
|
|
|
|
# (the "License"); you may not use this file except in compliance with
|
|
|
|
# the License. You may obtain a copy of the License at
|
|
|
|
#
|
|
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
#
|
|
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
# See the License for the specific language governing permissions and
|
|
|
|
# limitations under the License.
|
|
|
|
#
|
|
|
|
|
2015-08-27 01:19:11 -04:00
|
|
|
import sys
|
|
|
|
|
|
|
|
if sys.version >= '3':
|
|
|
|
basestring = unicode = str
|
|
|
|
|
2015-05-19 17:23:28 -04:00
|
|
|
from py4j.java_gateway import JavaClass
|
|
|
|
|
2016-04-20 13:32:01 -04:00
|
|
|
from pyspark import RDD, since, keyword_only
|
2015-10-28 17:28:38 -04:00
|
|
|
from pyspark.rdd import ignore_unicode_prefix
|
2015-05-19 17:23:28 -04:00
|
|
|
from pyspark.sql.column import _to_seq
|
|
|
|
from pyspark.sql.types import *
|
2015-11-18 11:18:54 -05:00
|
|
|
from pyspark.sql import utils
|
2015-05-19 17:23:28 -04:00
|
|
|
|
|
|
|
__all__ = ["DataFrameReader", "DataFrameWriter"]
|
|
|
|
|
|
|
|
|
2015-08-05 20:28:23 -04:00
|
|
|
def to_str(value):
|
|
|
|
"""
|
2016-04-22 12:19:36 -04:00
|
|
|
A wrapper over str(), but converts bool values to lower case strings.
|
|
|
|
If None is given, just returns None, instead of converting it to string "None".
|
2015-08-05 20:28:23 -04:00
|
|
|
"""
|
|
|
|
if isinstance(value, bool):
|
|
|
|
return str(value).lower()
|
2016-04-22 12:19:36 -04:00
|
|
|
elif value is None:
|
|
|
|
return value
|
2015-08-05 20:28:23 -04:00
|
|
|
else:
|
|
|
|
return str(value)
|
|
|
|
|
|
|
|
|
2015-05-19 17:23:28 -04:00
|
|
|
class DataFrameReader(object):
|
|
|
|
"""
|
|
|
|
Interface used to load a :class:`DataFrame` from external storage systems
|
|
|
|
(e.g. file systems, key-value stores, etc). Use :func:`SQLContext.read`
|
|
|
|
to access this.
|
|
|
|
|
|
|
|
::Note: Experimental
|
2015-05-21 02:05:54 -04:00
|
|
|
|
|
|
|
.. versionadded:: 1.4
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
|
|
|
|
|
|
|
def __init__(self, sqlContext):
|
|
|
|
self._jreader = sqlContext._ssql_ctx.read()
|
|
|
|
self._sqlContext = sqlContext
|
|
|
|
|
|
|
|
def _df(self, jdf):
|
|
|
|
from pyspark.sql.dataframe import DataFrame
|
|
|
|
return DataFrame(jdf, self._sqlContext)
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
@since(1.4)
|
|
|
|
def format(self, source):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Specifies the input data source format.
|
|
|
|
|
|
|
|
:param source: string, name of the data source, e.g. 'json', 'parquet'.
|
|
|
|
|
|
|
|
>>> df = sqlContext.read.format('json').load('python/test_support/sql/people.json')
|
|
|
|
>>> df.dtypes
|
|
|
|
[('age', 'bigint'), ('name', 'string')]
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
|
|
|
self._jreader = self._jreader.format(source)
|
|
|
|
return self
|
|
|
|
|
|
|
|
@since(1.4)
|
|
|
|
def schema(self, schema):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Specifies the input schema.
|
|
|
|
|
|
|
|
Some data sources (e.g. JSON) can infer the input schema automatically from data.
|
|
|
|
By specifying the schema here, the underlying data source can skip the schema
|
2015-06-02 11:37:18 -04:00
|
|
|
inference step, and thus speed up data loading.
|
|
|
|
|
|
|
|
:param schema: a StructType object
|
|
|
|
"""
|
|
|
|
if not isinstance(schema, StructType):
|
|
|
|
raise TypeError("schema should be StructType")
|
|
|
|
jschema = self._sqlContext._ssql_ctx.parseDataType(schema.json())
|
|
|
|
self._jreader = self._jreader.schema(jschema)
|
|
|
|
return self
|
|
|
|
|
2015-06-29 03:13:39 -04:00
|
|
|
@since(1.5)
|
|
|
|
def option(self, key, value):
|
|
|
|
"""Adds an input option for the underlying data source.
|
|
|
|
"""
|
2015-08-05 20:28:23 -04:00
|
|
|
self._jreader = self._jreader.option(key, to_str(value))
|
2015-06-29 03:13:39 -04:00
|
|
|
return self
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
@since(1.4)
|
|
|
|
def options(self, **options):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Adds input options for the underlying data source.
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
|
|
|
for k in options:
|
2015-08-05 20:28:23 -04:00
|
|
|
self._jreader = self._jreader.option(k, to_str(options[k]))
|
2015-06-02 11:37:18 -04:00
|
|
|
return self
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-05-19 17:23:28 -04:00
|
|
|
def load(self, path=None, format=None, schema=None, **options):
|
|
|
|
"""Loads data from a data source and returns it as a :class`DataFrame`.
|
|
|
|
|
2015-11-24 21:16:07 -05:00
|
|
|
:param path: optional string or a list of string for file-system backed data sources.
|
2015-05-19 17:23:28 -04:00
|
|
|
:param format: optional string for format of the data source. Default to 'parquet'.
|
|
|
|
:param schema: optional :class:`StructType` for the input schema.
|
|
|
|
:param options: all other string options
|
2015-06-03 03:23:34 -04:00
|
|
|
|
2015-08-05 20:28:23 -04:00
|
|
|
>>> df = sqlContext.read.load('python/test_support/sql/parquet_partitioned', opt1=True,
|
|
|
|
... opt2=1, opt3='str')
|
2015-06-03 03:23:34 -04:00
|
|
|
>>> df.dtypes
|
|
|
|
[('name', 'string'), ('year', 'int'), ('month', 'int'), ('day', 'int')]
|
2015-11-24 21:16:07 -05:00
|
|
|
|
2015-10-17 17:56:24 -04:00
|
|
|
>>> df = sqlContext.read.format('json').load(['python/test_support/sql/people.json',
|
|
|
|
... 'python/test_support/sql/people1.json'])
|
|
|
|
>>> df.dtypes
|
|
|
|
[('age', 'bigint'), ('aka', 'string'), ('name', 'string')]
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
|
|
|
if format is not None:
|
2015-06-02 11:37:18 -04:00
|
|
|
self.format(format)
|
2015-05-19 17:23:28 -04:00
|
|
|
if schema is not None:
|
2015-06-02 11:37:18 -04:00
|
|
|
self.schema(schema)
|
|
|
|
self.options(**options)
|
2015-05-19 17:23:28 -04:00
|
|
|
if path is not None:
|
2016-01-04 21:02:38 -05:00
|
|
|
if type(path) != list:
|
|
|
|
path = [path]
|
|
|
|
return self._df(self._jreader.load(self._sqlContext._sc._jvm.PythonUtils.toSeq(path)))
|
2015-05-19 17:23:28 -04:00
|
|
|
else:
|
2015-06-02 11:37:18 -04:00
|
|
|
return self._df(self._jreader.load())
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2016-04-20 13:32:01 -04:00
|
|
|
@since(2.0)
|
|
|
|
def stream(self, path=None, format=None, schema=None, **options):
|
|
|
|
"""Loads a data stream from a data source and returns it as a :class`DataFrame`.
|
|
|
|
|
|
|
|
:param path: optional string for file-system backed data sources.
|
|
|
|
:param format: optional string for format of the data source. Default to 'parquet'.
|
|
|
|
:param schema: optional :class:`StructType` for the input schema.
|
|
|
|
:param options: all other string options
|
|
|
|
|
|
|
|
>>> df = sqlContext.read.format('text').stream('python/test_support/sql/streaming')
|
|
|
|
>>> df.isStreaming
|
|
|
|
True
|
|
|
|
"""
|
|
|
|
if format is not None:
|
|
|
|
self.format(format)
|
|
|
|
if schema is not None:
|
|
|
|
self.schema(schema)
|
|
|
|
self.options(**options)
|
|
|
|
if path is not None:
|
|
|
|
if type(path) != str or len(path.strip()) == 0:
|
|
|
|
raise ValueError("If the path is provided for stream, it needs to be a " +
|
|
|
|
"non-empty string. List of paths are not supported.")
|
|
|
|
return self._df(self._jreader.stream(path))
|
|
|
|
else:
|
|
|
|
return self._df(self._jreader.stream())
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-05-19 17:23:28 -04:00
|
|
|
def json(self, path, schema=None):
|
|
|
|
"""
|
2015-08-27 01:19:11 -04:00
|
|
|
Loads a JSON file (one object per line) or an RDD of Strings storing JSON objects
|
|
|
|
(one object per record) and returns the result as a :class`DataFrame`.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
|
|
|
If the ``schema`` parameter is not specified, this function goes
|
|
|
|
through the input once to determine the input schema.
|
|
|
|
|
2015-08-27 01:19:11 -04:00
|
|
|
:param path: string represents path to the JSON dataset,
|
|
|
|
or RDD of Strings storing JSON objects.
|
2015-05-19 17:23:28 -04:00
|
|
|
:param schema: an optional :class:`StructType` for the input schema.
|
|
|
|
|
2015-11-16 03:06:14 -05:00
|
|
|
You can set the following JSON-specific options to deal with non-standard JSON files:
|
|
|
|
* ``primitivesAsString`` (default ``false``): infers all primitive values as a string \
|
|
|
|
type
|
2016-04-03 02:12:04 -04:00
|
|
|
* `prefersDecimal` (default `false`): infers all floating-point values as a decimal \
|
|
|
|
type. If the values do not fit in decimal, then it infers them as doubles.
|
2015-11-16 03:06:14 -05:00
|
|
|
* ``allowComments`` (default ``false``): ignores Java/C++ style comment in JSON records
|
|
|
|
* ``allowUnquotedFieldNames`` (default ``false``): allows unquoted JSON field names
|
|
|
|
* ``allowSingleQuotes`` (default ``true``): allows single quotes in addition to double \
|
|
|
|
quotes
|
|
|
|
* ``allowNumericLeadingZeros`` (default ``false``): allows leading zeros in numbers \
|
|
|
|
(e.g. 00012)
|
2016-01-03 20:01:19 -05:00
|
|
|
* ``allowBackslashEscapingAnyCharacter`` (default ``false``): allows accepting quoting \
|
|
|
|
of all character using backslash quoting mechanism
|
2016-03-21 03:42:35 -04:00
|
|
|
* ``mode`` (default ``PERMISSIVE``): allows a mode for dealing with corrupt records \
|
|
|
|
during parsing.
|
|
|
|
* ``PERMISSIVE`` : sets other fields to ``null`` when it meets a corrupted \
|
|
|
|
record and puts the malformed string into a new field configured by \
|
2016-03-22 08:30:48 -04:00
|
|
|
``columnNameOfCorruptRecord``. When a schema is set by user, it sets \
|
2016-03-21 03:42:35 -04:00
|
|
|
``null`` for extra fields.
|
|
|
|
* ``DROPMALFORMED`` : ignores the whole corrupted records.
|
|
|
|
* ``FAILFAST`` : throws an exception when it meets corrupted records.
|
2016-03-22 08:30:48 -04:00
|
|
|
* ``columnNameOfCorruptRecord`` (default ``_corrupt_record``): allows renaming the \
|
|
|
|
new field having malformed string created by ``PERMISSIVE`` mode. \
|
|
|
|
This overrides ``spark.sql.columnNameOfCorruptRecord``.
|
2015-11-16 03:06:14 -05:00
|
|
|
|
2015-08-27 01:19:11 -04:00
|
|
|
>>> df1 = sqlContext.read.json('python/test_support/sql/people.json')
|
|
|
|
>>> df1.dtypes
|
|
|
|
[('age', 'bigint'), ('name', 'string')]
|
|
|
|
>>> rdd = sc.textFile('python/test_support/sql/people.json')
|
|
|
|
>>> df2 = sqlContext.read.json(rdd)
|
|
|
|
>>> df2.dtypes
|
2015-06-03 03:23:34 -04:00
|
|
|
[('age', 'bigint'), ('name', 'string')]
|
|
|
|
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2015-06-02 11:37:18 -04:00
|
|
|
if schema is not None:
|
|
|
|
self.schema(schema)
|
2015-08-27 01:19:11 -04:00
|
|
|
if isinstance(path, basestring):
|
|
|
|
return self._df(self._jreader.json(path))
|
2015-11-24 21:16:07 -05:00
|
|
|
elif type(path) == list:
|
|
|
|
return self._df(self._jreader.json(self._sqlContext._sc._jvm.PythonUtils.toSeq(path)))
|
2015-08-27 01:19:11 -04:00
|
|
|
elif isinstance(path, RDD):
|
2016-01-04 21:02:38 -05:00
|
|
|
def func(iterator):
|
|
|
|
for x in iterator:
|
|
|
|
if not isinstance(x, basestring):
|
|
|
|
x = unicode(x)
|
|
|
|
if isinstance(x, unicode):
|
|
|
|
x = x.encode("utf-8")
|
|
|
|
yield x
|
|
|
|
keyed = path.mapPartitions(func)
|
|
|
|
keyed._bypass_serializer = True
|
|
|
|
jrdd = keyed._jrdd.map(self._sqlContext._jvm.BytesToString())
|
|
|
|
return self._df(self._jreader.json(jrdd))
|
2015-08-27 01:19:11 -04:00
|
|
|
else:
|
|
|
|
raise TypeError("path can be only string or RDD")
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-05-19 17:23:28 -04:00
|
|
|
def table(self, tableName):
|
|
|
|
"""Returns the specified table as a :class:`DataFrame`.
|
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
:param tableName: string, name of the table.
|
|
|
|
|
|
|
|
>>> df = sqlContext.read.parquet('python/test_support/sql/parquet_partitioned')
|
|
|
|
>>> df.registerTempTable('tmpTable')
|
|
|
|
>>> sqlContext.read.table('tmpTable').dtypes
|
|
|
|
[('name', 'string'), ('year', 'int'), ('month', 'int'), ('day', 'int')]
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
|
|
|
return self._df(self._jreader.table(tableName))
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-07-21 03:08:44 -04:00
|
|
|
def parquet(self, *paths):
|
2015-05-19 17:23:28 -04:00
|
|
|
"""Loads a Parquet file, returning the result as a :class:`DataFrame`.
|
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
>>> df = sqlContext.read.parquet('python/test_support/sql/parquet_partitioned')
|
|
|
|
>>> df.dtypes
|
|
|
|
[('name', 'string'), ('year', 'int'), ('month', 'int'), ('day', 'int')]
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2015-07-21 03:08:44 -04:00
|
|
|
return self._df(self._jreader.parquet(_to_seq(self._sqlContext._sc, paths)))
|
|
|
|
|
2015-10-28 17:28:38 -04:00
|
|
|
@ignore_unicode_prefix
|
|
|
|
@since(1.6)
|
2015-11-24 21:16:07 -05:00
|
|
|
def text(self, paths):
|
2015-12-17 12:19:46 -05:00
|
|
|
"""Loads a text file and returns a [[DataFrame]] with a single string column named "value".
|
2015-10-28 17:28:38 -04:00
|
|
|
|
|
|
|
Each line in the text file is a new row in the resulting DataFrame.
|
|
|
|
|
2015-11-24 21:16:07 -05:00
|
|
|
:param paths: string, or list of strings, for input path(s).
|
|
|
|
|
2015-10-28 17:28:38 -04:00
|
|
|
>>> df = sqlContext.read.text('python/test_support/sql/text-test.txt')
|
|
|
|
>>> df.collect()
|
2015-11-06 20:22:30 -05:00
|
|
|
[Row(value=u'hello'), Row(value=u'this')]
|
2015-10-28 17:28:38 -04:00
|
|
|
"""
|
2015-11-24 21:16:07 -05:00
|
|
|
if isinstance(paths, basestring):
|
|
|
|
paths = [paths]
|
|
|
|
return self._df(self._jreader.text(self._sqlContext._sc._jvm.PythonUtils.toSeq(paths)))
|
2015-10-28 17:28:38 -04:00
|
|
|
|
2016-02-29 12:44:29 -05:00
|
|
|
@since(2.0)
|
|
|
|
def csv(self, paths):
|
|
|
|
"""Loads a CSV file and returns the result as a [[DataFrame]].
|
|
|
|
|
|
|
|
This function goes through the input once to determine the input schema. To avoid going
|
|
|
|
through the entire data once, specify the schema explicitly using [[schema]].
|
|
|
|
|
|
|
|
:param paths: string, or list of strings, for input path(s).
|
|
|
|
|
|
|
|
>>> df = sqlContext.read.csv('python/test_support/sql/ages.csv')
|
|
|
|
>>> df.dtypes
|
|
|
|
[('C0', 'string'), ('C1', 'string')]
|
|
|
|
"""
|
|
|
|
if isinstance(paths, basestring):
|
|
|
|
paths = [paths]
|
|
|
|
return self._df(self._jreader.csv(self._sqlContext._sc._jvm.PythonUtils.toSeq(paths)))
|
|
|
|
|
2015-07-21 03:08:44 -04:00
|
|
|
@since(1.5)
|
|
|
|
def orc(self, path):
|
2015-10-28 17:28:38 -04:00
|
|
|
"""Loads an ORC file, returning the result as a :class:`DataFrame`.
|
2015-07-21 03:08:44 -04:00
|
|
|
|
|
|
|
::Note: Currently ORC support is only available together with
|
|
|
|
:class:`HiveContext`.
|
|
|
|
|
|
|
|
>>> df = hiveContext.read.orc('python/test_support/sql/orc_partitioned')
|
|
|
|
>>> df.dtypes
|
|
|
|
[('a', 'bigint'), ('b', 'int'), ('c', 'int')]
|
|
|
|
"""
|
|
|
|
return self._df(self._jreader.orc(path))
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-05-19 17:23:28 -04:00
|
|
|
def jdbc(self, url, table, column=None, lowerBound=None, upperBound=None, numPartitions=None,
|
2015-08-14 15:46:05 -04:00
|
|
|
predicates=None, properties=None):
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
|
|
|
Construct a :class:`DataFrame` representing the database table accessible
|
|
|
|
via JDBC URL `url` named `table` and connection `properties`.
|
|
|
|
|
|
|
|
The `column` parameter could be used to partition the table, then it will
|
|
|
|
be retrieved in parallel based on the parameters passed to this function.
|
|
|
|
|
|
|
|
The `predicates` parameter gives a list expressions suitable for inclusion
|
|
|
|
in WHERE clauses; each one defines one partition of the :class:`DataFrame`.
|
|
|
|
|
|
|
|
::Note: Don't create too many partitions in parallel on a large cluster;
|
|
|
|
otherwise Spark might crash your external database systems.
|
|
|
|
|
|
|
|
:param url: a JDBC URL
|
|
|
|
:param table: name of table
|
|
|
|
:param column: the column used to partition
|
|
|
|
:param lowerBound: the lower bound of partition column
|
|
|
|
:param upperBound: the upper bound of the partition column
|
|
|
|
:param numPartitions: the number of partitions
|
|
|
|
:param predicates: a list of expressions
|
|
|
|
:param properties: JDBC database connection arguments, a list of arbitrary string
|
|
|
|
tag/value. Normally at least a "user" and "password" property
|
|
|
|
should be included.
|
|
|
|
:return: a DataFrame
|
|
|
|
"""
|
2015-08-14 15:46:05 -04:00
|
|
|
if properties is None:
|
|
|
|
properties = dict()
|
2015-05-19 17:23:28 -04:00
|
|
|
jprop = JavaClass("java.util.Properties", self._sqlContext._sc._gateway._gateway_client)()
|
|
|
|
for k in properties:
|
|
|
|
jprop.setProperty(k, properties[k])
|
|
|
|
if column is not None:
|
|
|
|
if numPartitions is None:
|
|
|
|
numPartitions = self._sqlContext._sc.defaultParallelism
|
|
|
|
return self._df(self._jreader.jdbc(url, table, column, int(lowerBound), int(upperBound),
|
|
|
|
int(numPartitions), jprop))
|
|
|
|
if predicates is not None:
|
2015-11-18 11:18:54 -05:00
|
|
|
gateway = self._sqlContext._sc._gateway
|
|
|
|
jpredicates = utils.toJArray(gateway, gateway.jvm.java.lang.String, predicates)
|
|
|
|
return self._df(self._jreader.jdbc(url, table, jpredicates, jprop))
|
2015-05-19 17:23:28 -04:00
|
|
|
return self._df(self._jreader.jdbc(url, table, jprop))
|
|
|
|
|
|
|
|
|
|
|
|
class DataFrameWriter(object):
|
|
|
|
"""
|
|
|
|
Interface used to write a [[DataFrame]] to external storage systems
|
|
|
|
(e.g. file systems, key-value stores, etc). Use :func:`DataFrame.write`
|
|
|
|
to access this.
|
|
|
|
|
|
|
|
::Note: Experimental
|
2015-05-21 02:05:54 -04:00
|
|
|
|
|
|
|
.. versionadded:: 1.4
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
|
|
|
def __init__(self, df):
|
|
|
|
self._df = df
|
|
|
|
self._sqlContext = df.sql_ctx
|
|
|
|
self._jwrite = df._jdf.write()
|
|
|
|
|
2016-04-20 13:32:01 -04:00
|
|
|
def _cq(self, jcq):
|
|
|
|
from pyspark.sql.streaming import ContinuousQuery
|
|
|
|
return ContinuousQuery(jcq, self._sqlContext)
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
@since(1.4)
|
|
|
|
def mode(self, saveMode):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Specifies the behavior when data or table already exists.
|
|
|
|
|
|
|
|
Options include:
|
2015-06-02 11:37:18 -04:00
|
|
|
|
|
|
|
* `append`: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* `overwrite`: Overwrite existing data.
|
|
|
|
* `error`: Throw an exception if data already exists.
|
|
|
|
* `ignore`: Silently ignore this operation if data already exists.
|
2015-06-03 03:23:34 -04:00
|
|
|
|
|
|
|
>>> df.write.mode('append').parquet(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
2015-06-22 16:51:23 -04:00
|
|
|
# At the JVM side, the default value of mode is already set to "error".
|
|
|
|
# So, if the given saveMode is None, we will not call JVM-side's mode method.
|
|
|
|
if saveMode is not None:
|
|
|
|
self._jwrite = self._jwrite.mode(saveMode)
|
2015-06-02 11:37:18 -04:00
|
|
|
return self
|
|
|
|
|
|
|
|
@since(1.4)
|
|
|
|
def format(self, source):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Specifies the underlying output data source.
|
|
|
|
|
|
|
|
:param source: string, name of the data source, e.g. 'json', 'parquet'.
|
|
|
|
|
|
|
|
>>> df.write.format('json').save(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
|
|
|
self._jwrite = self._jwrite.format(source)
|
|
|
|
return self
|
|
|
|
|
2015-06-29 03:13:39 -04:00
|
|
|
@since(1.5)
|
|
|
|
def option(self, key, value):
|
|
|
|
"""Adds an output option for the underlying data source.
|
|
|
|
"""
|
2016-04-22 12:19:36 -04:00
|
|
|
self._jwrite = self._jwrite.option(key, to_str(value))
|
2015-06-29 03:13:39 -04:00
|
|
|
return self
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
@since(1.4)
|
|
|
|
def options(self, **options):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Adds output options for the underlying data source.
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
|
|
|
for k in options:
|
2016-04-22 12:19:36 -04:00
|
|
|
self._jwrite = self._jwrite.option(k, to_str(options[k]))
|
2015-06-02 11:37:18 -04:00
|
|
|
return self
|
|
|
|
|
|
|
|
@since(1.4)
|
|
|
|
def partitionBy(self, *cols):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Partitions the output by the given columns on the file system.
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
If specified, the output is laid out on the file system similar
|
|
|
|
to Hive's partitioning scheme.
|
|
|
|
|
|
|
|
:param cols: name of columns
|
2015-06-03 03:23:34 -04:00
|
|
|
|
|
|
|
>>> df.write.partitionBy('year', 'month').parquet(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-06-02 11:37:18 -04:00
|
|
|
"""
|
|
|
|
if len(cols) == 1 and isinstance(cols[0], (list, tuple)):
|
|
|
|
cols = cols[0]
|
2015-06-29 03:22:44 -04:00
|
|
|
self._jwrite = self._jwrite.partitionBy(_to_seq(self._sqlContext._sc, cols))
|
2015-06-02 11:37:18 -04:00
|
|
|
return self
|
|
|
|
|
2016-04-20 13:32:01 -04:00
|
|
|
@since(2.0)
|
|
|
|
def queryName(self, queryName):
|
|
|
|
"""Specifies the name of the :class:`ContinuousQuery` that can be started with
|
|
|
|
:func:`startStream`. This name must be unique among all the currently active queries
|
|
|
|
in the associated SQLContext.
|
|
|
|
|
|
|
|
:param queryName: unique name for the query
|
|
|
|
|
|
|
|
>>> writer = sdf.write.queryName('streaming_query')
|
|
|
|
"""
|
|
|
|
if not queryName or type(queryName) != str or len(queryName.strip()) == 0:
|
|
|
|
raise ValueError('The queryName must be a non-empty string. Got: %s' % queryName)
|
|
|
|
self._jwrite = self._jwrite.queryName(queryName)
|
|
|
|
return self
|
|
|
|
|
|
|
|
@keyword_only
|
|
|
|
@since(2.0)
|
|
|
|
def trigger(self, processingTime=None):
|
|
|
|
"""Set the trigger for the stream query. If this is not set it will run the query as fast
|
|
|
|
as possible, which is equivalent to setting the trigger to ``processingTime='0 seconds'``.
|
|
|
|
|
|
|
|
:param processingTime: a processing time interval as a string, e.g. '5 seconds', '1 minute'.
|
|
|
|
|
|
|
|
>>> # trigger the query for execution every 5 seconds
|
|
|
|
>>> writer = sdf.write.trigger(processingTime='5 seconds')
|
|
|
|
"""
|
|
|
|
from pyspark.sql.streaming import ProcessingTime
|
|
|
|
trigger = None
|
|
|
|
if processingTime is not None:
|
|
|
|
if type(processingTime) != str or len(processingTime.strip()) == 0:
|
|
|
|
raise ValueError('The processing time must be a non empty string. Got: %s' %
|
|
|
|
processingTime)
|
|
|
|
trigger = ProcessingTime(processingTime)
|
|
|
|
if trigger is None:
|
|
|
|
raise ValueError('A trigger was not provided. Supported triggers: processingTime.')
|
|
|
|
self._jwrite = self._jwrite.trigger(trigger._to_java_trigger(self._sqlContext))
|
|
|
|
return self
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-06-29 03:22:44 -04:00
|
|
|
def save(self, path=None, format=None, mode=None, partitionBy=None, **options):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Saves the contents of the :class:`DataFrame` to a data source.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
|
|
|
The data source is specified by the ``format`` and a set of ``options``.
|
|
|
|
If ``format`` is not specified, the default data source configured by
|
|
|
|
``spark.sql.sources.default`` will be used.
|
|
|
|
|
|
|
|
:param path: the path in a Hadoop supported file system
|
|
|
|
:param format: the format used to save
|
2015-06-03 03:23:34 -04:00
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
|
|
|
|
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
2015-06-22 16:51:23 -04:00
|
|
|
:param partitionBy: names of partitioning columns
|
2015-05-19 17:23:28 -04:00
|
|
|
:param options: all other string options
|
2015-06-03 03:23:34 -04:00
|
|
|
|
|
|
|
>>> df.write.mode('append').parquet(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2015-06-29 03:22:44 -04:00
|
|
|
self.mode(mode).options(**options)
|
|
|
|
if partitionBy is not None:
|
|
|
|
self.partitionBy(partitionBy)
|
2015-05-19 17:23:28 -04:00
|
|
|
if format is not None:
|
2015-06-02 11:37:18 -04:00
|
|
|
self.format(format)
|
2015-05-19 17:23:28 -04:00
|
|
|
if path is None:
|
2015-06-02 11:37:18 -04:00
|
|
|
self._jwrite.save()
|
2015-05-19 17:23:28 -04:00
|
|
|
else:
|
2015-06-02 11:37:18 -04:00
|
|
|
self._jwrite.save(path)
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2016-04-20 13:32:01 -04:00
|
|
|
@ignore_unicode_prefix
|
|
|
|
@since(2.0)
|
|
|
|
def startStream(self, path=None, format=None, partitionBy=None, queryName=None, **options):
|
|
|
|
"""Streams the contents of the :class:`DataFrame` to a data source.
|
|
|
|
|
|
|
|
The data source is specified by the ``format`` and a set of ``options``.
|
|
|
|
If ``format`` is not specified, the default data source configured by
|
|
|
|
``spark.sql.sources.default`` will be used.
|
|
|
|
|
|
|
|
:param path: the path in a Hadoop supported file system
|
|
|
|
:param format: the format used to save
|
|
|
|
|
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
|
|
|
:param partitionBy: names of partitioning columns
|
|
|
|
:param queryName: unique name for the query
|
|
|
|
:param options: All other string options. You may want to provide a `checkpointLocation`
|
|
|
|
for most streams, however it is not required for a `memory` stream.
|
|
|
|
|
|
|
|
>>> cq = sdf.write.format('memory').queryName('this_query').startStream()
|
|
|
|
>>> cq.isActive
|
|
|
|
True
|
|
|
|
>>> cq.name
|
|
|
|
u'this_query'
|
|
|
|
>>> cq.stop()
|
|
|
|
>>> cq.isActive
|
|
|
|
False
|
|
|
|
>>> cq = sdf.write.trigger(processingTime='5 seconds').startStream(
|
|
|
|
... queryName='that_query', format='memory')
|
|
|
|
>>> cq.name
|
|
|
|
u'that_query'
|
|
|
|
>>> cq.isActive
|
|
|
|
True
|
|
|
|
>>> cq.stop()
|
|
|
|
"""
|
|
|
|
self.options(**options)
|
|
|
|
if partitionBy is not None:
|
|
|
|
self.partitionBy(partitionBy)
|
|
|
|
if format is not None:
|
|
|
|
self.format(format)
|
|
|
|
if queryName is not None:
|
|
|
|
self.queryName(queryName)
|
|
|
|
if path is None:
|
|
|
|
return self._cq(self._jwrite.startStream())
|
|
|
|
else:
|
|
|
|
return self._cq(self._jwrite.startStream(path))
|
|
|
|
|
2015-06-02 11:37:18 -04:00
|
|
|
@since(1.4)
|
2015-05-23 12:07:14 -04:00
|
|
|
def insertInto(self, tableName, overwrite=False):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Inserts the content of the :class:`DataFrame` to the specified table.
|
|
|
|
|
2015-05-23 12:07:14 -04:00
|
|
|
It requires that the schema of the class:`DataFrame` is the same as the
|
|
|
|
schema of the table.
|
|
|
|
|
|
|
|
Optionally overwriting any existing data.
|
|
|
|
"""
|
|
|
|
self._jwrite.mode("overwrite" if overwrite else "append").insertInto(tableName)
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-06-29 03:22:44 -04:00
|
|
|
def saveAsTable(self, name, format=None, mode=None, partitionBy=None, **options):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Saves the content of the :class:`DataFrame` as the specified table.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-05-23 12:07:14 -04:00
|
|
|
In the case the table already exists, behavior of this function depends on the
|
|
|
|
save mode, specified by the `mode` function (default to throwing an exception).
|
|
|
|
When `mode` is `Overwrite`, the schema of the [[DataFrame]] does not need to be
|
|
|
|
the same as that of the existing table.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
|
|
|
* `append`: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* `overwrite`: Overwrite existing data.
|
|
|
|
* `error`: Throw an exception if data already exists.
|
|
|
|
* `ignore`: Silently ignore this operation if data already exists.
|
|
|
|
|
|
|
|
:param name: the table name
|
|
|
|
:param format: the format used to save
|
|
|
|
:param mode: one of `append`, `overwrite`, `error`, `ignore` (default: error)
|
2015-06-22 16:51:23 -04:00
|
|
|
:param partitionBy: names of partitioning columns
|
2015-05-19 17:23:28 -04:00
|
|
|
:param options: all other string options
|
|
|
|
"""
|
2015-06-29 03:22:44 -04:00
|
|
|
self.mode(mode).options(**options)
|
|
|
|
if partitionBy is not None:
|
|
|
|
self.partitionBy(partitionBy)
|
2015-05-19 17:23:28 -04:00
|
|
|
if format is not None:
|
2015-06-02 11:37:18 -04:00
|
|
|
self.format(format)
|
2015-06-03 03:23:34 -04:00
|
|
|
self._jwrite.saveAsTable(name)
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2016-03-03 13:30:55 -05:00
|
|
|
def json(self, path, mode=None, compression=None):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Saves the content of the :class:`DataFrame` in JSON format at the specified path.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
:param path: the path in any Hadoop supported file system
|
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
2016-03-03 13:30:55 -05:00
|
|
|
:param compression: compression codec to use when saving to file. This can be one of the
|
|
|
|
known case-insensitive shorten names (none, bzip2, gzip, lz4,
|
|
|
|
snappy and deflate).
|
2016-02-29 12:44:29 -05:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
>>> df.write.json(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2016-03-03 13:30:55 -05:00
|
|
|
self.mode(mode)
|
|
|
|
if compression is not None:
|
|
|
|
self.option("compression", compression)
|
|
|
|
self._jwrite.json(path)
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2016-03-03 13:30:55 -05:00
|
|
|
def parquet(self, path, mode=None, partitionBy=None, compression=None):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Saves the content of the :class:`DataFrame` in Parquet format at the specified path.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
:param path: the path in any Hadoop supported file system
|
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
2015-06-22 16:51:23 -04:00
|
|
|
:param partitionBy: names of partitioning columns
|
2016-03-03 13:30:55 -05:00
|
|
|
:param compression: compression codec to use when saving to file. This can be one of the
|
|
|
|
known case-insensitive shorten names (none, snappy, gzip, and lzo).
|
|
|
|
This will overwrite ``spark.sql.parquet.compression.codec``.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
>>> df.write.parquet(os.path.join(tempfile.mkdtemp(), 'data'))
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2015-06-29 03:22:44 -04:00
|
|
|
self.mode(mode)
|
|
|
|
if partitionBy is not None:
|
|
|
|
self.partitionBy(partitionBy)
|
2016-03-03 13:30:55 -05:00
|
|
|
if compression is not None:
|
|
|
|
self.option("compression", compression)
|
2015-06-22 16:51:23 -04:00
|
|
|
self._jwrite.parquet(path)
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-10-28 17:28:38 -04:00
|
|
|
@since(1.6)
|
2016-03-03 13:30:55 -05:00
|
|
|
def text(self, path, compression=None):
|
2015-10-28 17:28:38 -04:00
|
|
|
"""Saves the content of the DataFrame in a text file at the specified path.
|
|
|
|
|
2016-02-29 12:44:29 -05:00
|
|
|
:param path: the path in any Hadoop supported file system
|
2016-03-03 13:30:55 -05:00
|
|
|
:param compression: compression codec to use when saving to file. This can be one of the
|
|
|
|
known case-insensitive shorten names (none, bzip2, gzip, lz4,
|
|
|
|
snappy and deflate).
|
2016-02-29 12:44:29 -05:00
|
|
|
|
2015-10-28 17:28:38 -04:00
|
|
|
The DataFrame must have only one column that is of string type.
|
|
|
|
Each row becomes a new line in the output file.
|
|
|
|
"""
|
2016-03-03 13:30:55 -05:00
|
|
|
if compression is not None:
|
|
|
|
self.option("compression", compression)
|
2015-10-28 17:28:38 -04:00
|
|
|
self._jwrite.text(path)
|
|
|
|
|
2016-02-29 12:44:29 -05:00
|
|
|
@since(2.0)
|
2016-03-03 13:30:55 -05:00
|
|
|
def csv(self, path, mode=None, compression=None):
|
2016-02-29 12:44:29 -05:00
|
|
|
"""Saves the content of the [[DataFrame]] in CSV format at the specified path.
|
|
|
|
|
|
|
|
:param path: the path in any Hadoop supported file system
|
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
|
|
|
|
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
|
|
|
|
2016-03-03 13:30:55 -05:00
|
|
|
:param compression: compression codec to use when saving to file. This can be one of the
|
|
|
|
known case-insensitive shorten names (none, bzip2, gzip, lz4,
|
|
|
|
snappy and deflate).
|
2016-02-29 12:44:29 -05:00
|
|
|
|
|
|
|
>>> df.write.csv(os.path.join(tempfile.mkdtemp(), 'data'))
|
|
|
|
"""
|
2016-03-03 13:30:55 -05:00
|
|
|
self.mode(mode)
|
|
|
|
if compression is not None:
|
|
|
|
self.option("compression", compression)
|
|
|
|
self._jwrite.csv(path)
|
2016-02-29 12:44:29 -05:00
|
|
|
|
2015-10-28 17:28:38 -04:00
|
|
|
@since(1.5)
|
2016-03-03 13:30:55 -05:00
|
|
|
def orc(self, path, mode=None, partitionBy=None, compression=None):
|
2015-07-21 03:08:44 -04:00
|
|
|
"""Saves the content of the :class:`DataFrame` in ORC format at the specified path.
|
|
|
|
|
|
|
|
::Note: Currently ORC support is only available together with
|
|
|
|
:class:`HiveContext`.
|
|
|
|
|
|
|
|
:param path: the path in any Hadoop supported file system
|
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
|
|
|
|
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
|
|
|
:param partitionBy: names of partitioning columns
|
2016-03-03 13:30:55 -05:00
|
|
|
:param compression: compression codec to use when saving to file. This can be one of the
|
|
|
|
known case-insensitive shorten names (none, snappy, zlib, and lzo).
|
|
|
|
This will overwrite ``orc.compress``.
|
2015-07-21 03:08:44 -04:00
|
|
|
|
|
|
|
>>> orc_df = hiveContext.read.orc('python/test_support/sql/orc_partitioned')
|
|
|
|
>>> orc_df.write.orc(os.path.join(tempfile.mkdtemp(), 'data'))
|
|
|
|
"""
|
|
|
|
self.mode(mode)
|
|
|
|
if partitionBy is not None:
|
|
|
|
self.partitionBy(partitionBy)
|
2016-03-03 13:30:55 -05:00
|
|
|
if compression is not None:
|
|
|
|
self.option("compression", compression)
|
2015-07-21 03:08:44 -04:00
|
|
|
self._jwrite.orc(path)
|
|
|
|
|
2015-05-21 02:05:54 -04:00
|
|
|
@since(1.4)
|
2015-08-14 15:46:05 -04:00
|
|
|
def jdbc(self, url, table, mode=None, properties=None):
|
2015-06-03 03:23:34 -04:00
|
|
|
"""Saves the content of the :class:`DataFrame` to a external database table via JDBC.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
.. note:: Don't create too many partitions in parallel on a large cluster;\
|
|
|
|
otherwise Spark might crash your external database systems.
|
2015-05-19 17:23:28 -04:00
|
|
|
|
2015-06-03 03:23:34 -04:00
|
|
|
:param url: a JDBC URL of the form ``jdbc:subprotocol:subname``
|
2015-05-19 17:23:28 -04:00
|
|
|
:param table: Name of the table in the external database.
|
2015-06-03 03:23:34 -04:00
|
|
|
:param mode: specifies the behavior of the save operation when data already exists.
|
|
|
|
|
|
|
|
* ``append``: Append contents of this :class:`DataFrame` to existing data.
|
|
|
|
* ``overwrite``: Overwrite existing data.
|
|
|
|
* ``ignore``: Silently ignore this operation if data already exists.
|
|
|
|
* ``error`` (default case): Throw an exception if data already exists.
|
2015-05-19 17:23:28 -04:00
|
|
|
:param properties: JDBC database connection arguments, a list of
|
2015-06-03 03:23:34 -04:00
|
|
|
arbitrary string tag/value. Normally at least a
|
|
|
|
"user" and "password" property should be included.
|
2015-05-19 17:23:28 -04:00
|
|
|
"""
|
2015-08-14 15:46:05 -04:00
|
|
|
if properties is None:
|
|
|
|
properties = dict()
|
2015-05-19 17:23:28 -04:00
|
|
|
jprop = JavaClass("java.util.Properties", self._sqlContext._sc._gateway._gateway_client)()
|
|
|
|
for k in properties:
|
|
|
|
jprop.setProperty(k, properties[k])
|
|
|
|
self._jwrite.mode(mode).jdbc(url, table, jprop)
|
|
|
|
|
|
|
|
|
|
|
|
def _test():
|
|
|
|
import doctest
|
2015-06-03 03:23:34 -04:00
|
|
|
import os
|
|
|
|
import tempfile
|
2015-05-19 17:23:28 -04:00
|
|
|
from pyspark.context import SparkContext
|
2015-07-21 03:08:44 -04:00
|
|
|
from pyspark.sql import Row, SQLContext, HiveContext
|
2015-05-19 17:23:28 -04:00
|
|
|
import pyspark.sql.readwriter
|
2015-06-03 03:23:34 -04:00
|
|
|
|
|
|
|
os.chdir(os.environ["SPARK_HOME"])
|
|
|
|
|
2015-05-19 17:23:28 -04:00
|
|
|
globs = pyspark.sql.readwriter.__dict__.copy()
|
|
|
|
sc = SparkContext('local[4]', 'PythonTest')
|
2015-06-03 03:23:34 -04:00
|
|
|
|
|
|
|
globs['tempfile'] = tempfile
|
|
|
|
globs['os'] = os
|
2015-05-19 17:23:28 -04:00
|
|
|
globs['sc'] = sc
|
|
|
|
globs['sqlContext'] = SQLContext(sc)
|
2016-04-28 13:55:48 -04:00
|
|
|
globs['hiveContext'] = HiveContext._createForTesting(sc)
|
2015-06-03 03:23:34 -04:00
|
|
|
globs['df'] = globs['sqlContext'].read.parquet('python/test_support/sql/parquet_partitioned')
|
2016-04-20 13:32:01 -04:00
|
|
|
globs['sdf'] =\
|
|
|
|
globs['sqlContext'].read.format('text').stream('python/test_support/sql/streaming')
|
2015-06-03 03:23:34 -04:00
|
|
|
|
2015-05-19 17:23:28 -04:00
|
|
|
(failure_count, test_count) = doctest.testmod(
|
|
|
|
pyspark.sql.readwriter, globs=globs,
|
|
|
|
optionflags=doctest.ELLIPSIS | doctest.NORMALIZE_WHITESPACE | doctest.REPORT_NDIFF)
|
|
|
|
globs['sc'].stop()
|
|
|
|
if failure_count:
|
|
|
|
exit(-1)
|
|
|
|
|
|
|
|
|
|
|
|
if __name__ == "__main__":
|
|
|
|
_test()
|