Python 3 support #40

Closed
wbolster opened this Issue Aug 28, 2013 · 27 comments

Projects

None yet

8 participants

@wbolster
Owner
wbolster commented Aug 28, 2013 edited

This would be trivial for HappyBase, but the underlying Thrift library needs to be Python 3 compatible first.

A simple '2to3' on the source code seems to work, but this needs to be properly supported in a stable Thrift release first.

Status/todo for this issue:

  • Merge #78
  • Regenerate Thrift bindings using a newer Thrift version (maybe not needed)
  • Update docs (TODO.rst)
@wbolster
Owner

Thift seems to be Python 3 compatible nowadyas. See also #78.

@wbolster
Owner

(removed comment, now part of issue description)

@jeroenvlek

Thrift is getting there: https://issues.apache.org/jira/browse/THRIFT-1857.

I'm highly interested in this patch. Please let me know if there's anything I can do to help. For now, I'll just fork this, apply #78 and regenerate the bindings with a nightly from Thrift.

@jeroenvlek

Ugh, I didn't actually want to link these two issues, but I just wanted to indicate where I moved my comments...

Anyway, so I used Thrift 1.0.0-dev to gain Python 3 support and one of the tests fails:

======================================================================
ERROR: tests.test_api.test_atomic_counters
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/case.py", line 198, in runTest
    self.test(*self.arg)
  File "/home/jvlek/dev/lib/happybase/tests/test_api.py", line 172, in test_atomic_counters
    table.counter_set(row, column, 0)
  File "/home/jvlek/dev/lib/happybase/happybase/table.py", line 540, in counter_set
    self.put(row, {column: pack_i64(value)})
  File "/home/jvlek/dev/lib/happybase/happybase/table.py", line 436, in put
    batch.put(row, data)
  File "/home/jvlek/dev/lib/happybase/happybase/batch.py", line 132, in __exit__
    self.send()
  File "/home/jvlek/dev/lib/happybase/happybase/batch.py", line 55, in send
    self._table.connection.client.mutateRows(self._table.name, bms, {})
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/Hbase.py", line 1598, in mutateRows
    self.send_mutateRows(tableName, rowBatches, attributes)
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/Hbase.py", line 1607, in send_mutateRows
    args.write(self._oprot)
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/Hbase.py", line 7470, in write
    iter351.write(oprot)
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/ttypes.py", line 589, in write
    iter6.write(oprot)
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/ttypes.py", line 495, in write
    oprot.writeString(self.value)
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/thrift/protocol/TBinaryProtocol.py", line 122, in writeString
    encoded = bytearray(str, 'utf-8')
nose.proxy.TypeError: encoding or errors without a string argument

Which simplifies to:

>>> from struct import Struct              
>>> bytearray(Struct('>q').pack(0), 'utf-8')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: encoding or errors without a string argument

This is due to a new line in TBinaryProtocol that wasn't there before:

It seems that the 'utf-8' encoding parameter in TBinaryProtocol causes the problems, which is new in 1.0.0-dev:

## Thrift 0.9.2: lib/py/src/protocol/TBinaryProtocol.py ## 
def writeString(self, str):
  self.writeI32(len(str))
  self.trans.write(str)

## Thrift 1.0.0-dev: lib/py/src/protocol/TBinaryProtocol.py ## 
def writeString(self, str):
  encoded = bytearray(str, 'utf-8')
  self.writeI32(len(encoded))
  self.trans.write(encoded)

The integer probably should be send as a TType.I64. Does that mean a new message type, because it can't be send as part of a mutateRows_args anymore? Or is there a way to enforce a different datatype with mutations? edit: No, in ttypes.Mutation.thrift_spec a value is defined as TType.STRING

Furthermore, the tearDown fails:

======================================================================
ERROR: test suite for <module 'tests.test_api' from '/home/jvlek/dev/lib/happybase/tests/test_api.py'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/suite.py", line 229, in run
    self.tearDown()
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/suite.py", line 352, in tearDown
    self.teardownContext(ancestor)
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/suite.py", line 368, in teardownContext
    try_run(context, names)
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/util.py", line 471, in try_run
    return func()
  File "/home/jvlek/dev/lib/happybase/tests/test_api.py", line 76, in teardown_module
    connection.delete_table(TEST_TABLE_NAME, disable=True)
  File "/home/jvlek/dev/lib/happybase/happybase/connection.py", line 320, in delete_table
    if disable and self.is_table_enabled(name):
  File "/home/jvlek/dev/lib/happybase/happybase/connection.py", line 351, in is_table_enabled
    return self.client.isTableEnabled(name)
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/Hbase.py", line 727, in isTableEnabled
    return self.recv_isTableEnabled()
  File "/home/jvlek/dev/lib/happybase/happybase/hbase/Hbase.py", line 744, in recv_isTableEnabled
    raise x
thrift.Thrift.TApplicationException: Negative length: -2147418111

The tear down failure also caused an NPE at the server's side:

2015-04-21 17:10:21,018 ERROR [thrift-worker-0] thrift.ProcessFunction: Internal error processing isTableEnabled
java.lang.NullPointerException
        at org.apache.hadoop.hbase.util.Bytes.getBytes(Bytes.java:968)
        at org.apache.hadoop.hbase.thrift.ThriftServerRunner$HBaseHandler.isTableEnabled(ThriftServerRunner.java:675)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:606)
        at org.apache.hadoop.hbase.thrift.HbaseHandlerMetricsProxy.invoke(HbaseHandlerMetricsProxy.java:67)
        at com.sun.proxy.$Proxy6.isTableEnabled(Unknown Source)
        at org.apache.hadoop.hbase.thrift.generated.Hbase$Processor$isTableEnabled.getResult(Hbase.java:3697)
        at org.apache.hadoop.hbase.thrift.generated.Hbase$Processor$isTableEnabled.getResult(Hbase.java:3681)
        at org.apache.thrift.ProcessFunction.process(ProcessFunction.java:39)
        at org.apache.thrift.TBaseProcessor.process(TBaseProcessor.java:39)
        at org.apache.hadoop.hbase.thrift.TBoundedThreadPoolServer$ClientConnnection.run(TBoundedThreadPoolServer.java:289)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
        at java.lang.Thread.run(Thread.java:745)

I will look into the tearDown() exception now.

@wbolster
Owner

This code you mentioned handles byte strings on Python 2:

## Thrift 0.9.2: lib/py/src/protocol/TBinaryProtocol.py ## 
def writeString(self, str):
  self.writeI32(len(str))
  self.trans.write(str)

The updated code you mentioned is plain wrong if it is still intended to handle byte strings:

## Thrift 1.0.0-dev: lib/py/src/protocol/TBinaryProtocol.py ## 
def writeString(self, str):
  encoded = bytearray(str, 'utf-8')
  self.writeI32(len(encoded))
  self.trans.write(encoded)

This code will handle unicode text only, but I think writeString() is supposed to write byte strings. Additionally, this code (ab)uses a bytearray to encode a unicode string into bytes, while a simple text.encode('utf-8') would have sufficed.

Anyway, the problem is the buggy Thrift code: it needs to make up its mind about unicode text versus bytes. It should handle byte strings.

@jeroenvlek

I see now that it is not in Thrift's master, but it was introduced by one of the patches for Python 3 support. I will notify that thread.

@jeroenvlek jeroenvlek referenced this issue in apache/thrift Apr 22, 2015
Closed

Thrift-1857: Python 3 support, redux #213

@jeroenvlek
Traceback (most recent call last):
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/case.py", line 198, in runTest
    self.test(*self.arg)
  File "/home/jvlek/dev/lib/happybase/tests/test_api.py", line 360, in test_scan
    list(table.scan(batch_size=None))
  File "/home/jvlek/dev/lib/happybase/happybase/table.py", line 294, in scan
    if batch_size < 1:
TypeError: unorderable types: NoneType() < int()

In Python 3, comparison with None raises a TypeError. Which in the context of table.scan() makes sense, because batch_size should be an integer >= 1. The TypeError conveys more precise information, without having to change any code. So I suggest updating test_scan() as follows:

## tests_api.test_scan() ##
with assert_raises(TypeError):
  list(table.scan(batch_size=None))

with assert_raises(ValueError):
  list(table.scan(batch_size=0))
@jeroenvlek
Traceback (most recent call last):
  File "/home/jvlek/.virtualenvs/happybase/lib/python3.4/site-packages/nose/case.py", line 198, in runTest
    self.test(*self.arg)
  File "/home/jvlek/dev/lib/happybase/tests/test_util.py", line 34, in check
    s = s_hex.decode('hex')
AttributeError: 'str' object has no attribute 'decode'

This one seems like an architectural decision. Changing the decode() call for codec.decode(s_hex, 'hex'), led me to a "TypeError: ord() expected string of length 1, but int found".

util.str_increment() says its intent is to increment and truncate a byte string, so it would be enough to remove the call to ord(), since s[i] would already result in an integer. However, then other tests start to fail, because, for example, the row_prefix for table.scan is not a byte string.

Now, I guess you don't want to convert all strings to byte strings, yet the str_increment is written with byte strings in mind, and assumes a hexadecimal space. However, HBase's row keys are UTF-8 encoded, so I reckon there's no need to wrap around after '\xFF'.

Please find my proposed fix below. All the tests pass, but I don't know if there are any side-effects I'm overlooking, or whether you have a completely different direction in mind.

## happybase/util.py ##
MAX_UNICODE = chr(1114111)

def str_increment(s):
    """Increment and truncate a string (for sorting purposes)

    This functions returns the shortest string that sorts after the given
    string when compared using regular string comparison semantics.

    This function increments the last character that is smaller than MAX_UNICODE, and
    drops everything after it. If the string only contains MAX_UNICODE characters,
    `None` is returned.
    """
    for i in range(len(s) - 1, -1, -1):
        if s[i] != MAX_UNICODE:
            return s[:i] + chr(ord(s[i]) + 1)

    return None

## tests/test_util.py ##
def test_str_increment():
    def check(s, expected):
        v = util.str_increment(s)
        assert_equal(expected, v)
        assert_less(s, v)

    test_values = [
        ('00', '01'),
        ('01', '02'),
        ('fe', 'ff'),
        ('1234', '1235'),
        ('12fe', '12ff'),
        ('12ff', '12fg'),
        ('424242' + util.MAX_UNICODE, '424243'),
        ('4242' + 2 * util.MAX_UNICODE, '4243'),
    ]

    assert util.str_increment(util.MAX_UNICODE * 4) is None

    for s, expected in test_values:
        yield check, s, expected
@wbolster
Owner

HBase's row keys are UTF-8 encoded

That's not true. Row keys are byte strings, and a lot of data models depend on that.

@wbolster
Owner

so it would be enough to remove the call to ord(), since s[i] would already result in an integer

That will break on Python 2, since b'foo'[0] will be b'f' on Python 2, and 102 on Python 3. Some compat code is needed here I think. In my Plyvel project I have had the same issue for the same routine.

@jeroenvlek

That's not true. Row keys are byte strings, and a lot of data models depend on that.

Ah yes, I should have read the follow up comment on this one: http://permalink.gmane.org/gmane.comp.java.hadoop.hbase.user/18006

Although I guess for generating end keys, it still suffices, albeit potentially slightly less efficient (i.e. sending longer end keys than necessary to the server).

That will break on Python 2

Another reason to leave it in (I left the call to ord() in the code).

However, the call to chr() is not backwards compatible, since the equivalent under Python 2 would be unichr(). How did you solve it in your Plyvel project? A quick search on "str_increment" didn't yield anything.

@wbolster
Owner

Few quick follow-up notes:

  • No, it must do the right thing on the byte level, not on the UTF-8 encoding level. This code is used to determine the stop key when using a row prefix, so it must be on the byte level.
  • unichr() should be avoided; it operates on unicode strings. we're talking bytes here.
  • In Plyvel the routine is called bytes_increment(), it uses a bytearray under the hood. (That one has PyPy compat issues after compiling it to C code, but that doesn't apply here.)
@jeroenvlek

This encoding compatibility stuff is a tricky business, and I still feel that it involves some architectural decisions, but I'm effectively only using happybase since this week, so I'm either wrong about it involving architectural decisions, or I'm too inexperienced with the library to make them ;)

Below my solution, which suits my needs for now, and which I'll happily replace with happybase > 0.9.

## happybase/util.py ##
def str_increment(s):
    """Increment and truncate a byte string (for sorting purposes)

    This functions returns the shortest string that sorts after the given
    string when compared using regular string comparison semantics.

    This function increments the last byte that is smaller than ``0xFF``, and
    drops everything after it. If the string only contains ``0xFF`` bytes,
    `None` is returned.
    """

    try: 
        ba = bytearray(s, "utf-8")
    except TypeError:
        ba = bytearray(s)

    for i in range(len(ba) - 1, -1, -1):
        if ba[i] != 255:
            ba[i] += 1
            return ba[:(i+1)]

    return None

## tests/test_util.py ##
def test_str_increment():
    def check(s_hex, expected):
        s = codecs.decode(s_hex, 'hex')
        v = util.str_increment(s)
        v_hex = codecs.encode(v, 'hex')
        assert_equal(expected, v_hex)
        assert_less(s, v)

    test_values = [
        (b'00', b'01'),
        (b'01', b'02'),
        (b'fe', b'ff'),
        (b'1234', b'1235'),
        (b'12fe', b'12ff'),
        (b'12ff', b'13'),
        (b'424242ff', b'424243'),
        (b'4242ffff', b'4243'),
    ]

    assert util.str_increment(b'\xff\xff\xff') is None

    for s, expected in test_values:
        yield check, s, expected
@csabaszilveszter

Hi,

...pip install happybase doesn't seem to work...

(test33)[deploy@sfleu1webap11 test33]$ pip install happybase
Downloading/unpacking happybase
  Downloading happybase-0.9.tar.gz (62kB): 62kB downloaded
  Running setup.py egg_info for package happybase
    Traceback (most recent call last):
      File "<string>", line 16, in <module>
      File "/home/deploy/test33/build/happybase/setup.py", line 5, in <module>
        execfile('happybase/_version.py')
    NameError: name 'execfile' is not defined
    Complete output from command python setup.py egg_info:
    Traceback (most recent call last):

  File "<string>", line 16, in <module>

  File "/home/deploy/test33/build/happybase/setup.py", line 5, in <module>

    execfile('happybase/_version.py')

NameError: name 'execfile' is not defined

----------------------------------------
Cleaning up...
Command python setup.py egg_info failed with error code 1 in /home/deploy/test33/build/happybase
Storing complete log in /home/deploy/.pip/pip.log

pip.log:

(test33)[deploy@sfleu1webap11 test33]$ cat /home/deploy/.pip/pip.log
------------------------------------------------------------
/home/deploy/test33/bin/pip run on Thu Feb  4 12:10:22 2016
Downloading/unpacking happybase

  Getting page https://pypi.python.org/simple/happybase/
  URLs to search for versions for happybase:
  * https://pypi.python.org/simple/happybase/
  Analyzing links from page https://pypi.python.org/simple/happybase/
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.1.tar.gz#md5=376d9f4d22a1d0d475064e3289de6a92 (from https://pypi.python.org/simple/happybase/), version: 0.1
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.2.tar.gz#md5=b1204c9713612028846629a47d876116 (from https://pypi.python.org/simple/happybase/), version: 0.2
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.3.tar.gz#md5=3b6ea750d8fb83d63fb4f128d11f3e4f (from https://pypi.python.org/simple/happybase/), version: 0.3
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.4.tar.gz#md5=ea87560a8d3946a3fcde6beb3400e395 (from https://pypi.python.org/simple/happybase/), version: 0.4
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.5.tar.gz#md5=ae7de65b85270d54bb05d566ff3b0496 (from https://pypi.python.org/simple/happybase/), version: 0.5
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.6.tar.gz#md5=28905e2a334d6d7b42495f2c5ca46add (from https://pypi.python.org/simple/happybase/), version: 0.6
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.7.tar.gz#md5=655fd08ff83a7132e607176bcee3d0b0 (from https://pypi.python.org/simple/happybase/), version: 0.7
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.8.tar.gz#md5=dc322b9ea8a31121e75ab1d1cd50538c (from https://pypi.python.org/simple/happybase/), version: 0.8
    Found link https://pypi.python.org/packages/source/h/happybase/happybase-0.9.tar.gz#md5=0d9ca6e22a7edb03ccd75e99d0acf1ea (from https://pypi.python.org/simple/happybase/), version: 0.9
  Using version 0.9 (newest of versions: 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1)
  Downloading from URL https://pypi.python.org/packages/source/h/happybase/happybase-0.9.tar.gz#md5=0d9ca6e22a7edb03ccd75e99d0acf1ea (from https://pypi.python.org/simple/happybase/)
  Running setup.py egg_info for package happybase

    Traceback (most recent call last):

      File "<string>", line 16, in <module>

      File "/home/deploy/test33/build/happybase/setup.py", line 5, in <module>

        execfile('happybase/_version.py')

    NameError: name 'execfile' is not defined

    Complete output from command python setup.py egg_info:

    Traceback (most recent call last):

  File "<string>", line 16, in <module>

  File "/home/deploy/test33/build/happybase/setup.py", line 5, in <module>

    execfile('happybase/_version.py')

NameError: name 'execfile' is not defined

----------------------------------------

Cleaning up...

  Removing temporary dir /home/deploy/test33/build...
Command python setup.py egg_info failed with error code 1 in /home/deploy/test33/build/happybase

Exception information:
Traceback (most recent call last):
  File "/home/deploy/test33/lib/python3.3/site-packages/pip/basecommand.py", line 134, in main
    status = self.run(options, args)
  File "/home/deploy/test33/lib/python3.3/site-packages/pip/commands/install.py", line 236, in run
    requirement_set.prepare_files(finder, force_root_egg_info=self.bundle, bundle=self.bundle)
  File "/home/deploy/test33/lib/python3.3/site-packages/pip/req.py", line 1134, in prepare_files
    req_to_install.run_egg_info()
  File "/home/deploy/test33/lib/python3.3/site-packages/pip/req.py", line 259, in run_egg_info
    command_desc='python setup.py egg_info')
  File "/home/deploy/test33/lib/python3.3/site-packages/pip/util.py", line 670, in call_subprocess
    % (command_desc, proc.returncode, cwd))
pip.exceptions.InstallationError: Command python setup.py egg_info failed with error code 1 in /home/deploy/test33/build/happybase

My environment:

(test33)[deploy@sfleu1webap11 test33]$ uname -a && cat /etc/issue && which python && python --version
Linux sfleu1webap11.eu1.ecsna.com 2.6.32-573.12.1.el6.x86_64 #1 SMP Tue Dec 15 21:19:08 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
CentOS release 6.7 (Final)
Kernel \r on an \m

~/test33/bin/python
Python 3.3.2`

Am I doing something wrong or is this a bug?

@wbolster
Owner

python 3 is unfortunately not yet supported, see related tickets and known issue list for details

— wouter

sent from my phone. please forgive my tpyos.

On February 4, 2016 7:06:52 PM GMT+01:00, Csaba Szilveszter notifications@github.com wrote:

Hi,

I might have put this in a separate issue, but...
...pip install happybase doesn't seem to work...

`(test33)[deploy@sfleu1webap11 test33]$ pip install docopt
python-jenkins happybase pywebhdfs parse
Downloading/unpacking docopt
Downloading docopt-0.6.2.tar.gz
Running setup.py egg_info for package docopt

Downloading/unpacking python-jenkins
Downloading python-jenkins-0.4.12.tar.gz (43kB): 43kB downloaded
Running setup.py egg_info for package python-jenkins

Installed /home/deploy/test33/build/python-jenkins/pbr-1.8.1-py3.3.egg
[pbr] Processing SOURCES.txt
warning: LocalManifestMaker: standard file '-c' not found

warning: no previously-included files found matching '.gitignore'
warning: no previously-included files found matching '.gitreview'

warning: no previously-included files matching '.pyc' found anywhere
in distribution
warning: no previously-included files found matching '.gitignore'
warning: no previously-included files found matching '.gitreview'
warning: no previously-included files matching '
.pyc' found anywhere
in distribution
Downloading/unpacking happybase
Downloading happybase-0.9.tar.gz (62kB): 62kB downloaded
Running setup.py egg_info for package happybase
Traceback (most recent call last):
File "", line 16, in
File "/home/deploy/test33/build/happybase/setup.py", line 5, in

execfile('happybase/_version.py')
NameError: name 'execfile' is not defined
Complete output from command python setup.py egg_info:
Traceback (most recent call last):

File "", line 16, in

File "/home/deploy/test33/build/happybase/setup.py", line 5, in

execfile('happybase/_version.py')

NameError: name 'execfile' is not defined


Cleaning up...
Command python setup.py egg_info failed with error code 1 in
/home/deploy/test33/build/happybase
Storing complete log in /home/deploy/.pip/pip.log`

My environment:
`(test33)[deploy@sfleu1webap11 test33]$ uname -a && cat /etc/issue &&
which python && python --version
Linux sfleu1webap11.eu1.ecsna.com 2.6.32-573.12.1.el6.x86_64 #1 SMP
Tue Dec 15 21:19:08 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
CentOS release 6.7 (Final)
Kernel \r on an \m

~/test33/bin/python
Python 3.3.2`

Am I doing something wrong or is this a bug?


Reply to this email directly or view it on GitHub:
#40 (comment)

@surfacepatterns

I'm a little confused as to the status of related tickets and known issues. Can you summarize what needs to be done in order to make happybase work with Python 3? If the amount of work is reasonable and I can get permission from my employer, I'm likely willing to do what needs to be done.

@surfacepatterns

I created a pull request for Python 3 support. Note that while I don't think accepting the pull request is a good idea yet (see details in the pull request), it might still be useful for other developers.

@surfacepatterns

I closed the pull request, as most of the Hbase documentation conflicts with the comment in the Hbase.thrift file that suggests that keys, values, etc. should be UTF-8. I'll make another pull request which accepts bytes() instances in Python 3.

@surfacepatterns

I've filed a new pull request with code that does the right thing (mainly, expecting bytes() instances for table names, columns, cell values, etc.).

@kayzhou
kayzhou commented Mar 18, 2016

For Python3,There is still an error. I used pip to install the happybase.

➜  ~ pip3 install happybase
Collecting happybase
  Using cached happybase-0.9.tar.gz
    Complete output from command python setup.py egg_info:
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "/private/var/folders/tz/dhnmphd97tl_90vhb80c3bgw0000gn/T/pip-build-2ekx3nnz/happybase/setup.py", line 5, in <module>
        execfile('happybase/_version.py')
    NameError: name 'execfile' is not defined

    ----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /private/var/folders/tz/dhnmphd97tl_90vhb80c3bgw0000gn/T/pip-build-2ekx3nnz/happybase/
@surfacepatterns

@kayzhou: Python 3 compatibility has not yet made its way into the master branch.

@ssaamm
ssaamm commented Apr 22, 2016

Is there a recommendation for users who want to use happybase in Python 3?

@scttcper
scttcper commented Apr 22, 2016 edited

@ssaamm
use the python3 branch.
requirements.txt git+https://github.com/wbolster/happybase.git@python3

@csabaszilveszter

Niiiiiiice.... thx

On Fri, Apr 22, 2016 at 8:46 PM, Scott Cooper notifications@github.com
wrote:

@ssaamm https://github.com/ssaamm
use the python3 branch.
requirements.txt


You are receiving this because you commented.
Reply to this email directly or view it on GitHub
#40 (comment)

@ManikandanUV

Is there an ETA on when python 3 support will hit the master?

@wbolster
Owner
wbolster commented May 20, 2016 edited

warning. rant ahead. this is not personal.

there's a pr that needs some more work. you could have read my comments over there.

i don't give an eta since i maintain this project in my SPARE time even though i don't use it myself anymore, while pretty much ALL users are corporate users, and only very few of them have ever helped out in any significant way, e.g. by contributing code or providing financial support.

even though i keep my project alive since i care about it, i don't work for free on demand by others, so i only spend time on it when i feel like it.

i also refuse to feel guilty about this. this is a cultural problem with open source projects used in corporate environments. only taking while never giving back is not sustainable.

@wbolster
Owner
wbolster commented Aug 1, 2016

i merged #116 into master

@wbolster wbolster closed this Aug 1, 2016
This was referenced Aug 1, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment