spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Brian Bruggeman (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (SPARK-19872) UnicodeDecodeError in Pyspark on sc.textFile read with repartition
Date Wed, 08 Mar 2017 21:48:38 GMT

    [ https://issues.apache.org/jira/browse/SPARK-19872?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15902023#comment-15902023
] 

Brian Bruggeman edited comment on SPARK-19872 at 3/8/17 9:48 PM:
-----------------------------------------------------------------

Using the Spark 2.1.0 serializers.py and the Spark 2.0.2 rdd.py, the code runs.  Therefore
there must be a bug in the Spark 2.1.0 rdd.py file.


I attempted the naive approach of reverting both {code}def _load_from_socket{code} and {code}def
collect{code}, but I didn't see any appreciable difference within the output.


was (Author: bbruggeman):
Using the Spark 2.1.0 serializers.py and the Spark 2.0.2 rdd.py, the code runs.  Therefore
there must be a bug in the Spark 2.1.0 rdd.py file.


> UnicodeDecodeError in Pyspark on sc.textFile read with repartition
> ------------------------------------------------------------------
>
>                 Key: SPARK-19872
>                 URL: https://issues.apache.org/jira/browse/SPARK-19872
>             Project: Spark
>          Issue Type: Bug
>          Components: PySpark
>    Affects Versions: 2.1.0
>         Environment: Mac and EC2
>            Reporter: Brian Bruggeman
>
> I'm receiving the following traceback:
> {code}
> >>> sc.textFile('test.txt').repartition(10).collect()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/rdd.py", line 810,
in collect
>     return list(_load_from_socket(port, self._jrdd_deserializer))
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/rdd.py", line 140,
in _load_from_socket
>     for item in serializer.load_stream(rf):
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/serializers.py",
line 539, in load_stream
>     yield self.loads(stream)
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/serializers.py",
line 534, in loads
>     return s.decode("utf-8") if self.use_unicode else s
>   File "/Users/brianbruggeman/.envs/dg/lib/python2.7/encodings/utf_8.py", line 16, in
decode
>     return codecs.utf_8_decode(input, errors, True)
> UnicodeDecodeError: 'utf8' codec can't decode byte 0x80 in position 0: invalid start
byte
> {code}
> I created a textfile (text.txt) with standard linux newlines:
> {code}
> a
> b
> d
> e
> f
> g
> h
> i
> j
> k
> l
> {code}
> I think ran pyspark:
> {code}
> $ pyspark
> Python 2.7.13 (default, Dec 18 2016, 07:03:39)
> [GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.42.1)] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
> Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties
> Setting default log level to "WARN".
> To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).
> 17/03/08 13:59:27 WARN NativeCodeLoader: Unable to load native-hadoop library for your
platform... using builtin-java classes where applicable
> 17/03/08 13:59:32 WARN ObjectStore: Failed to get database global_temp, returning NoSuchObjectException
> Welcome to
>       ____              __
>      / __/__  ___ _____/ /__
>     _\ \/ _ \/ _ `/ __/  '_/
>    /__ / .__/\_,_/_/ /_/\_\   version 2.1.0
>       /_/
> Using Python version 2.7.13 (default, Dec 18 2016 07:03:39)
> SparkSession available as 'spark'.
> >>> sc.textFile('test.txt').collect()
> [u'a', u'b', u'c', u'd', u'e', u'f', u'g', u'h', u'i', u'j', u'k', u'l']
> >>> sc.textFile('test.txt', use_unicode=False).collect()
> ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l']
> >>> sc.textFile('test.txt', use_unicode=False).repartition(10).collect()
> ['\x80\x02]q\x01(U\x01aU\x01bU\x01cU\x01dU\x01eU\x01fU\x01ge.', '\x80\x02]q\x01(U\x01hU\x01iU\x01jU\x01kU\x01le.']
> >>> sc.textFile('test.txt').repartition(10).collect()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/rdd.py", line 810,
in collect
>     return list(_load_from_socket(port, self._jrdd_deserializer))
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/rdd.py", line 140,
in _load_from_socket
>     for item in serializer.load_stream(rf):
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/serializers.py",
line 539, in load_stream
>     yield self.loads(stream)
>   File "/usr/local/Cellar/apache-spark/2.1.0/libexec/python/pyspark/serializers.py",
line 534, in loads
>     return s.decode("utf-8") if self.use_unicode else s
>   File "/Users/brianbruggeman/.envs/dg/lib/python2.7/encodings/utf_8.py", line 16, in
decode
>     return codecs.utf_8_decode(input, errors, True)
> UnicodeDecodeError: 'utf8' codec can't decode byte 0x80 in position 0: invalid start
byte
> {code}
> This really looks like a bug in the `serializers.py` code.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message