A little background: I am working on some python modules that other developers on our team will use. A common theme of each module is that one or more messages will be published to Kafka. We intend at this time to use the Confluent Kafka client. We are pretty new to python development in our organization -- we have traditionally been a .NET shop.
The complication: while the code that we create will run on Linux (rhel 7), most of the developers will do their work on Windows.
So we need the librdkafka C library compiled on each developer machine (which has dependencies of its own, one of which is OpenSSL). Then a pip install of confluent-kafka should just work, which means a pip install of our package will work. Theoretically.
To start I did the install on my Linux laptop (Arch). I knew I already had OpenSSL and the other zip lib dependencies available, so this process was painless:
git clone librdkafka repo
configure, make and install per the README
pip install confluent-kafka
done
The install of librdkafka went into /usr/local:
/usr/local/lib/librdkafka.a
/usr/local/lib/librdkafka++.a
/usr/local/lib/librdkafka.so -> librdkafka.so.l
/usr/local/lib/librdkafka++.so -> librdkafka++.so.l
/usr/local/lib/librdkafka.so.l
/usr/local/lib/librdkafka++.so.l
/usr/local/lib/pkgconfig/rdkafka.pc
/usr/local/lib/pkgconfig/rdkafka++.pc
/usr/local/include/librdkafka/rdkafkacpp.h
/usr/local/include/librdkafka/rdkafka.h
Now the painful part, making it work on Windows:
install precompiled OpenSSL
git clone librdkafka repo
open in VS2015
install libz via NuGet
build solution
install to where???
This is where I'm stuck. What would a standard install on a Windows 7/8/10 machine look like?
I have the following from the build output, but no idea what should go where in order to make the pip install confluent-kafka "just work":
/librdkafka/win32/Release/librdkafka.dll
/librdkafka/win32/Release/librdkafka.exp
/librdkafka/win32/Release/librdkafka.lib
/librdkafka/win32/Release/librdkafkacpp.dll
/librdkafka/win32/Release/librdkafkacpp.exp
/librdkafka/win32/Release/librdkafkacpp.lib
/librdkafka/win32/Release/zlib.dll
<and the .h files back in the src>
Any recommendations on an install location?
I'm not sure where the ideal place to install on Windows would be, but I ran the following test with some success.
I copied my output and headers to C:\test\lib and C:\test\include, then ran a pip install with the following options:
pip install --global-option=build_ext --global-option="-LC:\test\lib" --global-option="-IC:\test\include" confluent-kafka
Unfortunately, this doesn't quite work because the confluent-kafka setup does not support Windows at this time: https://github.com/confluentinc/confluent-kafka-python/issues/52#issuecomment-252098462
It's an old question but seems still no easy answer yet. Also Confluent seems too busy to work on Windows supporting...
I had the same headache couple weeks ago and after some research, I managed to make it work for me on Windows. I logged my finding and uploaded a pre-compiled library to my Git, please check and see if it helps. :D
https://github.com/MichaelZhangCA/confluent-kafka-python
My environment is Python 3.6 64 bit version but ideally, it should also work for 32 bit if you follow the same approach.
I assume you have successfully followed instructions from MichaelZhangCA (https://github.com/MichaelZhangCA/confluent-kafka-python/) from previous post.
If you did so, these probably were the last two commands executed:
::Install confluent-kafka
cd C:\confluent-kafka-python\confluent-kafka-python-0.11.4
python setup.py install
If that is correct, those DLLs were created under C:\confluent-kafka-python\librdkafka-reference\release\.
All you have to do is to copy them to a diretory already in Widnows' PATH.
For example, I use Anaconda 5.2 For Windows, with python 3.6. My Anaconda Prompt has an empty directory in PATH, so I copied there those DLL's:
::Anaconda Prompt - copy DLLs to a directory already in PATH
mkdir %CONDA_PREFIX%\Library\usr\bin
copy C:\confluent-kafka-python\librdkafka-reference\release %CONDA_PREFIX%\Library\usr\bin
If you don't use Anaconda, just copy those DLL's to any other directory in Windows' PATH. You may also leave them in C:\confluent-kafka-python\librdkafka-reference\release, and add this directory to PATH.
Related
I have a windows 7 machine with python 2.7 and I am trying to install pyzmq following these steps. I built libzmq got the binaries and copied them from libzmq\bin\Win32\Debug\v140\dynamic\ to libzmq\lib\ so the next step will work(compiler will have access to /lib and /includes from the same parent folder). But, on this step:
$ python setup.py configure --zmq=../libzmq
I installed pyzmq and libzmq on the same parent folder, as in the installation description related to pyzmq, libzmq is here: ../libzmq
But when I need to configure the pyzmq, I get this error:
I have VS Community 2015 installed and everything seems fine.
ZMQ usually provides sln files that will build from Microsoft Visual Studio. You will have to dig around a little to find it. You are better off trying to get that to work then going directly from the python setup.py that you are currently attempting. Note that you also need libsodium to be built and installed. Thankfully, they also provide Microsoft sln files.
In any case, you are probably better using one of the Python wheels like those from Christoph http://www.lfd.uci.edu/~gohlke/pythonlibs/
Just do
pip install <wheel file>
and you should be good.
Good luck!
I have 'Windows 8'. I installed 'python35-32' successfully. Then I download 'matplotlib 1.5.1' . But it can not installed . I get an error that required package can not be built.
Try this:
Open cmd with admin privilege. (right click and Run as administrator)
Enter this command pip install matplotlib, and wait for the installation to complete.
Alternative:
If you want to keep the wheels for offline installation later, you can do the following:
Open cmd, and enter this command
pip install --download /path/to/some/dir/ matplotlib. This will download matplotlib and all other required wheels in the directory /path/to/some/dir/.
To install, enter this command: pip install --no-index --find-links /path/to/some/dir/ matplotlib
Here /path/to/some/dir/ is the directory where the file gets downloaded. You can use . instead of /path/to/some/dir/ to download in the current directory.
Edit:
Take a look at matplotlib installation guide for windows. The line most relevant to your case is probably this:
For Python 3.5 the Visual C++ Redistributable for Visual Studio 2015 needs to be installed.
So just install it by following the link and then follow the above procedure. That should do the trick.
Note: They also suggest that the best option is using one of the pre-packaged python distributions that already provide matplotlib built-in, like Anaconda, miniconda and canopy. You may want take into consideration too.
Edits as per your comment
As I have already said, anaconda is a pre-packaged python distribution that includes hundreds of python packages. To quote the official python page, anaconda is "a full Python distribution for data management, analysis and visualization of large data sets". So you can do pretty much everything with anaconda that you can do with the traditional python (aka CPython), plus you have a great number of popular python packages at your disposal.
To get yourself started with anaconda, you can check this out: Test drive Anaconda.
I've downloaded python libraries (tar.gz) from external links. How to use these libraries, I mean in which directory under the python should I unzip such libraries.
How to use downloaded libraries with python when we are using:
windows 7
Ubuntu
Also how to install these libraries without using pip in Linux? Do we have to set path in windows7 to use the libraries?
Guys better to post comments then to give negative points. With your suggestion I can modify my post. I dont think you to be smart guy by rating negative points in any comments.
For Linux based systems (e.g., Ubuntu):
$ tar xzf package.tar.gz
$ cd package
$ python setup.py install
Also, look input pip https://pypi.python.org/pypi/pip
On Windows and Linux, all you need to do is unzip the archive somewhere, cd into and run the following command as administrator (usually by running cmd.exe as Administrator on Windows or by using sudo on Linux):
python setup.py install
If the package does not have a setup.py, it can get tricky. Most of the time the package just needs to be copied into your Python's site-packages folder.
You will find some packages that need to be built. On Linux, you can usually do this pretty easily. On Windows, you will need the appropriate compiler, usually a free version of Visual Studio will do, although you have to make sure you get the correct version of Visual Studio. I've heard reports that you can also use MingW.
I'm working with PyInstaller under Python 2.6, which is only partially supported due to the mess MS have created with their manifest nonense which now affects Python since it is now MSVC8 compiled.
The problem is that the manifest embedding support relies on the pywin32 extensions in order to build which is a pain because without including the host's site-packages folder when I create the virtualenv (kinda defeats the point in a build environment) I cannot find a way to install the required extensions so they are accessible to PyInstaller.
Has anyone found a solution to this issue?
I found http://old.nabble.com/Windows:-virtualenv-and-pywin32--td27658201.html (now a dead link) which offered the following solution:
Browse http://sourceforge.net/projects/pywin32/files/ for the URL of the exe you want
Activate your virtualenv
Run easy_install http://PATH.TO/EXE/DOWNLOAD
This works with modern versions of setuptools (circa February 2014, reported by tovmeod in the comments).
If you are using an old version of setuptools (or distribute it merged back into setuptools), you may get this error message:
error: c:\users\blah\appdata\local\temp\easy_install-ibkzv7\pywin32-214.win32-py2.6.exe is not a valid distutils Windows .exe
In which case:
Download the exe yourself
Activate your virtualenv
Run easy_install DOWNLOADED_FILE.exe
I rather hopefully tried "pip install" rather than "easy_install", but this didn't work, and likely never will (citation needed).
Finally, I found but haven't tested a solution at http://www.mail-archive.com/python-list#python.org/msg272040.html which is:
Solved this by copying the pywin32.pth file into my virtualenv site-packages
and editing the file to point to the path.
If the other options don't work for you, maybe this will?
For Python 2.7 or 3.x use pypiwin32.
pip install pypiwin32
OK, well since I had to find a way forward I improvised. I've internally created a git repository with a hacked-together version of pywin32 that will install within a virtualenv using the standard setup.py script. It took a lot of fiddling to make it work right but I managed to get it to load and the dependent code now works as I need it to. If people feel this would be of benefit to the community please post a comment: if I get enough I'll try and put something up on my github account.
This may have been improved since previous answer, since I've successfully installed pywin32 on sandbox on several machines without any specific "hacks" :
$ virtualenv sandbox
$ sandbox\scripts\activate
(sandbox) $ git clone https://github.com/Travis-Sun/pywin32.git
(sandbox) $ cd pywin32
(sandbox) $ python setup.py install
Tested with following environment :
windows 7
git
python 2.7.10 with virtualenv
VS2008. It may also work (but I've not tested yet) with
http://www.microsoft.com/en-us/download/details.aspx?id=44266
Edit: Scratch this for now, appears to be some problems with the installation still...
I got rather tired of the whole situation, and just created a set of converted wheels ("wheel convert <.exe>"). I'll try and keep them maintained for the most recent build, but do shout if there are any issues.
https://tr00st.co.uk/python/wheel/pywin32/
Installation can be done easily using pip and pointing to the package matching your version and architecture. For example, for Python 3.5/amd64:
pip install https://tr00st.co.uk/python/wheel/pywin32/pywin32-219-cp35-none-win_amd64.whl
Caveat: The --upgrade process currently fails, as the uninstall procedure is unable to clean up after itself (Access Denied when cleaning up win32api.pyd) - this is only when removing the temporary directory, which can be manually deleted. Easiest way around this is to uninstall and reinstall instead of upgrading, then manually delete the temporary folder.
The simplest way to deal with python package installations, so far, to me, has been to check out the source from the source control system and then add a symbolic link in the python dist-packages folder.
Clearly since source control provides the complete control to downgrade, upgrade to any branch, tag, it works very well.
Is there a way using one of the Package installers (easy_install or pip or other), one can achieve the same.
easy_install obtains the tar.gz and install them using the setup.py install which installs in the dist-packages folder in python2.6. Is there a way to configure it, or pip to use the source version control system (SVN/GIT/Hg/Bzr) instead.
Using pip this is quite easy. For instance:
pip install -e hg+http://bitbucket.org/andrewgodwin/south/#egg=South
Pip will automatically clone the source repo and run "setup.py develop" for you to install it into your environment (which hopefully is a virtualenv). Git, Subversion, Bazaar and Mercurial are all supported.
You can also then run "pip freeze" and it will output a list of your currently-installed packages with their exact versions (including, for develop-installs, the exact revision from the VCS). You can put this straight into a requirements file and later run
pip install -r requirements.txt
to install that same set of packages at the exact same versions.
If you download or check out the source distribution of a package — the one that has its "setup.py" inside of it — then if the package is based on the "setuptools" (which also power easy_install), you can move into that directory and say:
$ python setup.py develop
and it will create the right symlinks in dist-packages so that the .py files in the source distribution are the ones that get imported, rather than copies installed separately (which is what "setup.py install" would do — create separate copies that don't change immediately when you edit the source code to try a change).
As the other response indicates, you should try reading the "setuptools" documentation to learn more. "setup.py develop" is a really useful feature! Try using it in combination with a virtualenv, and you can "setup.py develop" painlessly and without messing up your system-wide Python with packages you are only developing on temporarily:
http://pypi.python.org/pypi/virtualenv
easy_install has support for downloading specific versions. For example:
easy_install python-dateutil==1.4.0
Will install v1.4, while the latest version 1.4.1 would be picked if no version was specified.
There is also support for svn checkouts, but using that doesn't give you much benefits from your manual version. See the manual for more information above.
Being able to switch to specific branches is rarely useful unless you are developing the packages in question, and then it's typically not a good idea to install them in site-packages anyway.
easy_install accepts a URL for the source tree too. Works at least when the sources are in Subversion.