run_tests.py for the latest sfepy in window 8, pythonxy-2.7.6.1 and msysgit-fullinstall-1.9.0
Hi,
For run_tests.py for the latest sfepy in window 8, here is what I do:
- Installations:
1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed are listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
- run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63 failure(s) of 116 test(s)
ouyang
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))
This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"
run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py at c:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is what I do:
- Installations:
1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed are listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
- run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63 failure(s) of 116 test(s)
ouyang
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place? Does removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py at c:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is what I do:
- Installations:
1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed are listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63 failure(s) of 116 test(s)
ouyang
I retested 'base.py' with run_tests.py. Failure of run_tests.py in my window8 has nothing to do with
top_dir = "c:\Python27\Lib\site-packages\sfepy"
It failed here:
if (package_name is None) and (top_dir == path[:len(top_dir)]):
package_name = path[len(top_dir) + 1:].replace('/', '.')
path = top_dir
I added several "print" in base.py (see below) and two attached files are from the two top_dir cases:
case 1: top_dir = os.path.normpath(os.path.join(base_dir, '..'))
case 2: top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"
def import_file(filename, package_name=None, can_reload=True): """ Import a file as a module. The module is explicitly reloaded to prevent undesirable interactions. """ from sfepy import base_dir #top_dir = os.path.normpath(os.path.join(base_dir, '..')) top_dir = "c:\Python27\Lib\site-packages\sfepy" print('top_dir===>',top_dir) #print('os.path.normpath top_dir===>',top_dir) print('package_name==>', package_name)------------------------------------------------------------------------------ path = os.path.dirname(os.path.normpath(os.path.realpath(filename))) print('path===>',type(path),len(top_dir), path[:len(top_dir)],path[len(top_dir) + 1:],path) if (package_name is None) and (top_dir == path[:len(top_dir)]): package_name = path[len(top_dir) + 1:].replace('/', '.') path = top_dir print('path....',path)
See if this helps to fix the issue in window 8.
ouyang
On Wednesday, July 23, 2014 10:16:31 PM UTC+8, Robert Cimrman wrote:
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place? Does removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py at c:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is what Ido:
- Installations:
1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed
are
listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63 failure(s) of 116 test(s)
ouyang
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]):
#package_name = path[len(top_dir) + 1:].replace('/', '.')
package_name = path[len(top_dir) + 1:].replace(os.sep, '.')
path = top_dir
Using os.sep instead of '/' is more portable here. It works now. I have not tested that one which crash python yet. No chance to test os.sep for other systems.
Ouyang
On Wednesday, July 23, 2014 10:16:31 PM UTC+8, Robert Cimrman wrote:
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place? Does removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py at c:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is what Ido:
- Installations:
1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed
are
listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63 failure(s) of 116 test(s)
ouyang
On 07/24/2014 12:23 PM, Ouyang wrote:
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]): #package_name = path[len(top_dir) + 1:].replace('/', '.') package_name = path[len(top_dir) + 1:].replace(os.sep, '.') path = top_dirUsing os.sep instead of '/' is more portable here. It works now. I have not tested that one which crash python yet. No chance to test os.sep for other systems.
Hey, good catch (and a very stupid bug), thanks for debugging!
Would you like to send a patch [1] or pull request [2]?
Cheers, r. [1] http://sfepy.org/doc-devel/developer_guide.html#with-git [2] https://help.github.com/articles/using-pull-requests
Ouyang
On Wednesday, July 23, 2014 10:16:31 PM UTC+8, Robert Cimrman wrote:
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place? Does removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py atc:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is what Ido:
- Installations: 1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages installed
are
listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit failed (msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63failure(s) of 116 test(s)
ouyang
Hi Robert,
I will send a patch. Let me read Git manual first :(( .
Ouyang
On Friday, July 25, 2014 3:00:04 PM UTC+8, Robert Cimrman wrote:
On 07/24/2014 12:23 PM, Ouyang wrote:
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]): #package_name = path[len(top_dir) + 1:].replace('/', '.') package_name = path[len(top_dir) + 1:].replace(os.sep, '.') path = top_dirUsing os.sep instead of '/' is more portable here. It works now. I have not tested that one which crash python yet. No chance to test os.sep for other systems.
Hey, good catch (and a very stupid bug), thanks for debugging!
Would you like to send a patch [1] or pull request [2]?
Cheers, r. [1] http://sfepy.org/doc-devel/developer_guide.html#with-git [2] https://help.github.com/articles/using-pull-requests
Ouyang
On Wednesday, July 23, 2014 10:16:31 PM UTC+8, Robert Cimrman wrote:
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place? Does removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py atc:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is whatI
do:
- Installations: 1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages
installed
are
listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit
failed
(msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py
which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous msysgit installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63failure(s) of 116 test(s)
ouyang
Hi Robert,
The attached two files are the patch files. Let me know if it is not in correct format.
ouyang
On Saturday, July 26, 2014 3:44:29 PM UTC+8, Ouyang wrote:
Hi Robert,
I will send a patch. Let me read Git manual first :(( .
Ouyang
On Friday, July 25, 2014 3:00:04 PM UTC+8, Robert Cimrman wrote:
On 07/24/2014 12:23 PM, Ouyang wrote:
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]): #package_name = path[len(top_dir) + 1:].replace('/', '.') package_name = path[len(top_dir) + 1:].replace(os.sep, '.') path = top_dirUsing os.sep instead of '/' is more portable here. It works now. I
have
not tested that one which crash python yet. No chance to test os.sep for other systems.
Hey, good catch (and a very stupid bug), thanks for debugging!
Would you like to send a patch [1] or pull request [2]?
Cheers, r. [1] http://sfepy.org/doc-devel/developer_guide.html#with-git [2] https://help.github.com/articles/using-pull-requests
Ouyang
On Wednesday, July 23, 2014 10:16:31 PM UTC+8, Robert Cimrman wrote:
Thanks for the info.
So you have both sfepy installed system-wide, and run it in-place?
Does
removing the system-wide files change something?
r.
On 07/23/2014 03:59 PM, Ouyang wrote:
On Wednesday, July 23, 2014 8:53:33 PM UTC+8, Ouyang wrote:
top_dir = os.path.normpath(os.path.join(base_dir, '..'))This statement assign "c:\sfepy" to top_dir instead of "c:\Python27\Lib\site-packages\sfepy".
After I hack the base.py in c:\sfepy\sfepy\base\base.py (NOT c:\Python27\Lib\site-packages\sfepy\base\base.py)
top_dir = "c:\\Python27\\Lib\\site-packages\\sfepy"run_tests.py passes except stokes_slip_bc.py which crash python.
Sfepy should call base.py atc:\Python27\Lib\site-packages\sfepy\base\ at very beginning.
Ouyang
Hi,
For run_tests.py for the latest sfepy in window 8, here is whatI
do:
- Installations: 1.1 Upgrade python(x,y) from 2.7.6.0 to 2.7.6.1 (packages
installed
are
listed in the attached file using "yolk -l".) 1.2 Reinstall msysgit: Since installation of 1.9.4 "NETINSTALL" verion of msysGit
failed
(msysGit-netinstall-1.9.4-preview20140611.exe), I installed 1.9.0 "FULLINSTALL" version of msysGit (msysGit-fullinstall-1.9.0-preview20140217.exe at https://github.com/msysgit/msysgit/releases).
run_tests.py:
2.1 Old version of sfepy: 2.1.1 Start msysGit shell with msys.bat 2.1.2 cd to sfepy root directory 2.1.3 python setup.py build_ext --inplace --compiler=mingw32 (I ignored the error info: "msvcr90.dll not found"). 2.1.4 ./run_tests.py 2.1.5 This time, all tests are ok except stokes_slip_bc.py
which
causes python crash. It was ok before with python(x,y)-2.7.6.0 and previous
msysgit
installation.
2.2 Latest verstion of sfepy: 2.2.1 python setup.py clean 2.2.2 python setup.py build_ext --inplace --compiler=mingw32 2.2.3 ./run_tests.py 2.2.5 No luck again: 79 test file(s) executed in 10.12 s, 63failure(s) of 116 test(s)
ouyang
Hi Ouyang
On 07/26/2014 03:27 PM, Ouyang wrote:
Hi Robert,
The attached two files are the patch files. Let me know if it is not in correct format.
Yes, they were in the correct format (*), thanks! Try [1] to see if all is correct, I will then push the changes to the main master branch.
Thanks! r. (*) The first file was my earlier commit already in the repository. I also tweaked a little bit the commit message (present tense etc.). [1] https://github.com/rc/sfepy - also the docs problem should be fixed there.
ouyang
On Saturday, July 26, 2014 3:44:29 PM UTC+8, Ouyang wrote:
Hi Robert,
I will send a patch. Let me read Git manual first :(( .
Ouyang
On Friday, July 25, 2014 3:00:04 PM UTC+8, Robert Cimrman wrote:
On 07/24/2014 12:23 PM, Ouyang wrote:
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]): #package_name = path[len(top_dir) + 1:].replace('/', '.') package_name = path[len(top_dir) + 1:].replace(os.sep, '.') path = top_dirUsing os.sep instead of '/' is more portable here. It works now. I
have
not tested that one which crash python yet. No chance to test os.sep for other systems.
Hey, good catch (and a very stupid bug), thanks for debugging!
Would you like to send a patch [1] or pull request [2]?
Cheers, r. [1] http://sfepy.org/doc-devel/developer_guide.html#with-git [2] https://help.github.com/articles/using-pull-requests
Hi Robert,
- I have not installed SciPy 0.13.3 yet so run_tests.py fails at
stokes_slip_bc.py. 2. "make latex" seems to go through all the steps. But pdf pages is less than the pages of the copy on site. There are some errors while compiling of "make latex" (see the two files attached).
Ouyang
On Sunday, July 27, 2014 3:54:34 AM UTC+8, Robert Cimrman wrote:
Hi Ouyang
On 07/26/2014 03:27 PM, Ouyang wrote:
Hi Robert,
The attached two files are the patch files. Let me know if it is not in correct format.
Yes, they were in the correct format (*), thanks! Try [1] to see if all is correct, I will then push the changes to the main master branch.
Thanks! r. (*) The first file was my earlier commit already in the repository. I also tweaked a little bit the commit message (present tense etc.). [1] https://github.com/rc/sfepy - also the docs problem should be fixed there.
ouyang
On Saturday, July 26, 2014 3:44:29 PM UTC+8, Ouyang wrote:
Hi Robert,
I will send a patch. Let me read Git manual first :(( .
Ouyang
On Friday, July 25, 2014 3:00:04 PM UTC+8, Robert Cimrman wrote:
On 07/24/2014 12:23 PM, Ouyang wrote:
Hi Robert,
I think I found the small bug in base.py in
if (package_name is None) and (top_dir == path[:len(top_dir)]): #package_name = path[len(top_dir) + 1:].replace('/', '.') package_name = path[len(top_dir) + 1:].replace(os.sep, '.') path = top_dirUsing os.sep instead of '/' is more portable here. It works now. I
have
not tested that one which crash python yet. No chance to test os.sep for other systems.
Hey, good catch (and a very stupid bug), thanks for debugging!
Would you like to send a patch [1] or pull request [2]?
Cheers, r. [1] http://sfepy.org/doc-devel/developer_guide.html#with-git [2] https://help.github.com/articles/using-pull-requests
On 07/28/2014 02:31 AM, Ouyang wrote:
Hi Robert,
- I have not installed SciPy 0.13.3 yet so run_tests.py fails at
stokes_slip_bc.py.
No problem - you do not need to do that - it has already been confirmed by Matt, see the other thread.
- "make latex" seems to go through all the steps. But pdf pages is less
than the pages of the copy on site. There are some errors while compiling of "make latex" (see the two files attached).
That is OK. The missing pages are the examples (see my other post). I will push the changes now.
BTW. for some reason the PDF I can generate has two same "Python Module Index" sections, and and "Index" section. What sphinx version do you have?
Thanks for your help!
r.
Ouyang
On Sunday, July 27, 2014 3:54:34 AM UTC+8, Robert Cimrman wrote:
Hi Ouyang
On 07/26/2014 03:27 PM, Ouyang wrote:
Hi Robert,
The attached two files are the patch files. Let me know if it is not in correct format.
Yes, they were in the correct format (*), thanks! Try [1] to see if all is correct, I will then push the changes to the main master branch.
Thanks! r. (*) The first file was my earlier commit already in the repository. I also tweaked a little bit the commit message (present tense etc.). [1] https://github.com/rc/sfepy - also the docs problem should be fixed there.
sphinx version on my machine is 1.2.2.
ouyang
On Monday, July 28, 2014 4:13:22 PM UTC+8, Robert Cimrman wrote:
On 07/28/2014 02:31 AM, Ouyang wrote:
Hi Robert,
- I have not installed SciPy 0.13.3 yet so run_tests.py fails at
stokes_slip_bc.py.
No problem - you do not need to do that - it has already been confirmed by Matt, see the other thread.
- "make latex" seems to go through all the steps. But pdf pages is
less
than the pages of the copy on site. There are some errors while
compiling
of "make latex" (see the two files attached).
That is OK. The missing pages are the examples (see my other post). I will push the changes now.
BTW. for some reason the PDF I can generate has two same "Python Module Index" sections, and and "Index" section. What sphinx version do you have?
Thanks for your help!
r.
Ouyang
On Sunday, July 27, 2014 3:54:34 AM UTC+8, Robert Cimrman wrote:
Hi Ouyang
On 07/26/2014 03:27 PM, Ouyang wrote:
Hi Robert,
The attached two files are the patch files. Let me know if it is not
in
correct format.
Yes, they were in the correct format (*), thanks! Try [1] to see if all
is
correct, I will then push the changes to the main master branch.
Thanks! r. (*) The first file was my earlier commit already in the repository. I
also
tweaked a little bit the commit message (present tense etc.). [1] https://github.com/rc/sfepy - also the docs problem should be
fixed
there.
sphinx version on my machine is 1.2.2.
ouyang
On Monday, July 28, 2014 4:13:22 PM UTC+8, Robert Cimrman wrote:
On 07/28/2014 02:31 AM, Ouyang wrote:
Hi Robert,
- I have not installed SciPy 0.13.3 yet so run_tests.py fails at
stokes_slip_bc.py.
No problem - you do not need to do that - it has already been confirmed by Matt, see the other thread.
- "make latex" seems to go through all the steps. But pdf pages is
less
than the pages of the copy on site. There are some errors while
compiling
of "make latex" (see the two files attached).
That is OK. The missing pages are the examples (see my other post). I will push the changes now.
BTW. for some reason the PDF I can generate has two same "Python Module Index" sections, and and "Index" section. What sphinx version do you have?
Thanks for your help!
r.
Ouyang
On Sunday, July 27, 2014 3:54:34 AM UTC+8, Robert Cimrman wrote:
Hi Ouyang
On 07/26/2014 03:27 PM, Ouyang wrote:
Hi Robert,
The attached two files are the patch files. Let me know if it is not
in
correct format.
Yes, they were in the correct format (*), thanks! Try [1] to see if all
is
correct, I will then push the changes to the main master branch.
Thanks! r. (*) The first file was my earlier commit already in the repository. I
also
tweaked a little bit the commit message (present tense etc.). [1] https://github.com/rc/sfepy - also the docs problem should be
fixed
there.
-
Ouyang -
Robert Cimrman