1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
|
== Windows Build ==
Download and install python from www.python.org. Get both 32-bit and 64-bit versions. Install the 64-bit version in C:\Python27_64 and the 32-bit version in C:\Python27
Download and install msysgit (the windows git user is winbuild and there's no passphrase when generating the ssh keys).
Download and install PyQt4 binary installers (includes sip, Qt and QScintilla)
Download and install active tcl/tk 8.5 both 32-bit and 64-bit versions (make sure to install them in different directories C:\Tcl and C:\Tcl_64
download boost library from www.boost.org and uncompress it to a local folder,
I had problems building vtk with recent versions of boost, the version I know it works for sure is 1.46.1.
So extract the zip file to a folder, for example ~/src/boost_1_46_1 vtk will use the same boost for both 32-bit and 64-bit versions
Download and install cmake
Get vtk source
uncompress it to (e.g, ~\src\vtk\vtk-5.10.1\
and rename the VTK folder to src
create build and build64 folders in ~\src\vtk\vtk-5.10.1
vtk 32-bit:
run cmake-gui pointing the src folder to src and build folder to build and press configure
select Visual Studio 9 2008
mark the following variables:
BUILD_SHARED_LIBS
VTK_WRAP_PYTHON
on the top check Advanced
now mark
VTK_USE_BOOST
VTK_USE_TEXT_ANALYSIS
and press configure
Set
Boost_INCLUDE_DIR to ~/src/boost_1_46_1 (or where you uncompressed boost)
make sure it is using the correct versions of tcl and python (they should point to 32-bit installed versions)
press configure again and finally press generate
now browse to the build directory and open VTK.sln in Visual Studio
On the top bar select Release and then build solution (F7)
If all goes well, you can install vtk python bindings:
open a command prompt in the build/Wrapping/Python folder and run the following command:
C:\Python27\python.exe setup.py BUILD_TYPE=Release install
I've created two environment variables called PATH32 and PATH64 configured with
the environment for each version of VisTrails:
PATH32 = C:\Tcl\bin;C:\Python27\Lib\site-packages\PyQt4;C:\Python27;C:\Python27\Scripts;C:\Users\vistrails\src\vtk\vtk-5.10.1\build\bin\Release;C:\Users\vistrails\code\vistrails\dist\windows\Input
PATH64 = C:\Tcl_64\bin;C:\Python27_64\Lib\site-packages\PyQt4;C:\Python27_64;C:\Python27_64\Scripts;C:\Users\vistrails\src\vtk\vtk-5.10.1\build64\bin\Release;C:\Users\vistrails\code\vistrails\dist\windows\Input
Then, when I need to use the right environment, I just update the PATH
variable with the right one:
PATH = %PATH64%;%PATH%
vtk 64-bit:
run cmake-gui again pointing the src folder to src and build folder to build64 and press configure
select Visual Studio 9 2008 Win64
mark the following variables:
BUILD_SHARED_LIBS
VTK_WRAP_PYTHON
on the top check Advanced
now mark
VTK_USE_BOOST
VTK_USE_TEXT_ANALYSIS
and press configure
Set
Boost_INCLUDE_DIR to ~/src/boost_1_46_1 (or where you uncompressed boost)
Boost_LIBRARY_DIRS to ~/src/boost_1_46_1/libs
make sure it is using the correct versions of tcl and python (they should point to 64-bit installed versions) press configure again and finally press generate now browse to the build64 directory and open VTK.sln in Visual Studio On the top bar select Release and then build solution (F7)
If all goes well, switch to the 64 bit Path and you can install vtk python bindings:
open a command prompt in the build64/Wrapping/Python folder and run the following command:
C:\Python27_64\python.exe setup.py BUILD_TYPE=Release install
Now let's build each configuration at time, starting with the 32-bit version:
== 32-bit version ==
Download and install numpy from numpy.scipy.org (there is a win32 super pack for python 2.7 binary version on source forge)
Download and install scipy from scipy.org (there is a win32 super pack for python 2.7 binary version on source forge)
Download and install matplotlib from matplotlib.org (there is a win32 for python 2.7 version on source forge)
Download and install mysqldb python from source forge, there's a win32 binary installer for python 2.7.
Download and unzip hdf5-1.8.4 from http://www.hdfgroup.org/ftp/HDF5/releases/hdf5-1.8.4/bin/windows/hdf5-1.8.4-32bit-VS2008-IVF101.zip
Download fftw (http://www.fftw.org/install/windows.html) from ftp://ftp.fftw.org/pub/fftw/fftw-3.3.3-dll32.zip and extract the zip file to the folder ~/src/fftw-3.3.3-dll32
Download and install setuptools (easy_install)
Download and install the patched version of suds
Copy stockwell files from x86_bin folder to C:\Python27\Lib\site-packages
st.pyd and sine.pyd
== 64-bit version ==
Go to http://www.lfd.uci.edu/~gohlke/pythonlibs/ and download and
install the 64-bit versions of (they are all binaries):
numpy (get the MKL version)
scipy
matplotlib
MySQL-python
psycopg
Download and unzip hdf5-1.8.4 from http://www.hdfgroup.org/ftp/HDF5/releases/hdf5-1.8.4/bin/windows/hdf5-1.8.4-64bit-VS2008-IVF101.zip
Download fftw (http://www.fftw.org/install/windows.html) from
ftp://ftp.fftw.org/pub/fftw/fftw-3.3.3-dll64.zip and extract the zip
file to the folder ~/src/fftw-3.3.3-dll64 Follow the instructions on
the web page to generate the .lib files
Download and install setuptools (easy_install)
Download and install the patched version of suds
Copy stockwell files from x64_bin folder to C:\Python27_64\Lib\site-packages
st.pyd and sine.pyd
#Under construction
|
