[GRASS-dev] All Python addons are missing

GRASS GIS GRASS GIS developer
1

Sorry to report but grass.script cannot be imported when building addons, e.g.:

http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

and recently posted log for 71/trunk binary is 404:

http://grass.osgeo.org/grass71/binary/linux/snapshot/build.log

Thanks,

Vaclav

2

Hi,

2015-06-14 1:56 GMT+02:00 Vaclav Petras <wenzeslaus@gmail.com>:

Sorry to report but grass.script cannot be imported when building addons,
e.g.:

http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

thanks for reporting, problem was broken svn repo:

svn: Failed to add file 'setup.py': an unversioned file of the same
name already exists

I fixed that.

and recently posted log for 71/trunk binary is 404:

http://grass.osgeo.org/grass71/binary/linux/snapshot/build.log

No idea here, Martin

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

3

On Jun 14, 2015 10:15 AM, “Martin Landa” <landa.martin@gmail.com> wrote:

Hi,

2015-06-14 1:56 GMT+02:00 Vaclav Petras <wenzeslaus@gmail.com>:

Sorry to report but grass.script cannot be imported when building addons,
e.g.:

http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

thanks for reporting, problem was broken svn repo:

svn: Failed to add file ‘setup.py’: an unversioned file of the same
name already exists

I fixed that.

I had seen it on phone but could not fix it in the past 12hs.

and recently posted log for 71/trunk binary is 404:

http://grass.osgeo.org/grass71/binary/linux/snapshot/build.log

No idea here, Martin

As I mentioned I generated manually that file. Then removed it to avoid confusion with the next build. Will modify my cronjobs later today to autogenerate the file later today.

Markus

–
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

4

Hi,

2015-06-14 10:18 GMT+02:00 Markus Neteler <neteler@osgeo.org>:

thanks for reporting, problem was broken svn repo:

svn: Failed to add file 'setup.py': an unversioned file of the same
name already exists

I fixed that.

to avoid confusion, addons are not built on osgeo server, but on the
server at the CTU. Martin

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

5

Hi,

2015-06-14 1:56 GMT+02:00 Vaclav Petras <wenzeslaus@gmail.com>:

http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

btw, I am getting strange error on server when compiling relbr70:

GRASS GIS 7.0.1svn r65457 compilation log
--------------------------------------------------
Started compilation: Sun Jun 14 10:16:48 CEST 2015
--
Errors in:
/opt/src/grass70_release/lib/python/gunittest
--
In case of errors please change into the directory with error and run 'make'.
If you get multiple errors, you need to deal with them in the order they
appear in the error log. If you get an error building a library, you will
also get errors from anything which uses the library.
--
Finished compilation: Sun Jun 14 10:16:57 CEST 2015
make: *** [default] Error 1

but this directory doesn't exist

cd lib/python/gunittest
-bash: cd: lib/python/gunittest: No such file or directory

Martin

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

6

On Sun, Jun 14, 2015 at 4:49 AM, Martin Landa <landa.martin@gmail.com>
wrote:

Hi,

2015-06-14 1:56 GMT+02:00 Vaclav Petras <wenzeslaus@gmail.com>:
> http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

btw, I am getting strange error on server when compiling relbr70:

GRASS GIS 7.0.1svn r65457 compilation log
--------------------------------------------------
Started compilation: Sun Jun 14 10:16:48 CEST 2015
--
Errors in:
/opt/src/grass70_release/lib/python/gunittest
--
In case of errors please change into the directory with error and run
'make'.
If you get multiple errors, you need to deal with them in the order they
appear in the error log. If you get an error building a library, you will
also get errors from anything which uses the library.
--
Finished compilation: Sun Jun 14 10:16:57 CEST 2015
make: *** [default] Error 1

but this directory doesn't exist

cd lib/python/gunittest
-bash: cd: lib/python/gunittest: No such file or directory

The build for gunittest dir was requested in the Makefile. This was caused
by accidental backport in r65439. Pietro, svn st and svn diff are your
friends :slight_smile:

https://trac.osgeo.org/grass/browser/grass/branches/releasebranch_7_0/lib/python/Makefile?annotate=blame&rev=65439
https://trac.osgeo.org/grass/changeset/65439
https://trac.osgeo.org/grass/changeset/65461

Martin

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

7

On Sun, Jun 14, 2015 at 10:21 AM, Martin Landa <landa.martin@gmail.com> wrote:

2015-06-14 10:18 GMT+02:00 Markus Neteler <neteler@osgeo.org>:
>> thanks for reporting, problem was broken svn repo:
>>
>> svn: Failed to add file 'setup.py': an unversioned file of the same
>> name already exists
>>
>> I fixed that.

to avoid confusion, addons are not built on osgeo server, but on the
server at the CTU. Martin

Right. However, the same issue happened on the main server, now fixed.

I scheduled for later today (in 1-2 hs) the rebuild of the
binaries/manual with hopefully build.log file output. Hope the cronjob
will behave with the log redirection.

Markus

8

On Sun, Jun 14, 2015 at 4:15 AM, Martin Landa <landa.martin@gmail.com>
wrote:

2015-06-14 1:56 GMT+02:00 Vaclav Petras <wenzeslaus@gmail.com>:
> Sorry to report but grass.script cannot be imported when building addons,
> e.g.:
>
> http://grass.osgeo.org/addons/grass7/logs/d.vect.thematic2.log

thanks for reporting, problem was broken svn repo:

svn: Failed to add file 'setup.py': an unversioned file of the same
name already exists

I fixed that.

Markus had the same issue. This was discussed some time ago, I think in
relation to changes in module names (g.mlist/g.list) and perhaps builds for
automated tests. If you want to avoid problems run `make distclean` before
`svn up`. For automated builds this is necessary. Here is what I use:

# if we do distclean but svn up fails, then distclean fails next time
# (thus || true is necessary with set -e)
make distclean $GENERAL_MAKE_FLAGS &>$OUTPUT_LOGFILE || true
svn up &>>$OUTPUT_LOGFILE
./script_configure.sh &>>$OUTPUT_LOGFILE
make $GENERAL_MAKE_FLAGS &>>$OUTPUT_LOGFILE

9

On Sun, Jun 14, 2015 at 12:40 PM, Vaclav Petras <wenzeslaus@gmail.com> wrote:

...
If you want to avoid problems run `make distclean` before
`svn up`. For automated builds this is necessary.

Sure, I have "make distclean" in for years. Yet not sufficient if the
failing file is not defined as clean target. This rarely happens but
this time it did.

Markus

10

On Sun, Jun 14, 2015 at 6:43 AM, Markus Neteler <neteler@osgeo.org> wrote:

On Sun, Jun 14, 2015 at 12:40 PM, Vaclav Petras <wenzeslaus@gmail.com>
wrote:
>...
> If you want to avoid problems run `make distclean` before
> `svn up`. For automated builds this is necessary.

Sure, I have "make distclean" in for years. Yet not sufficient if the
failing file is not defined as clean target. This rarely happens but
this time it did.

I apparently learn it from you [1]. However, the build for automated tests
did not fail with the setup.py change, so how is that possible? setup.py
was supposed to be in the OBJ file [2] if I understand the Makefile
correctly.

[1] http://lists.osgeo.org/pipermail/grass-dev/2014-September/070986.html
[1 nabble]
http://osgeo-org.1560.x6.nabble.com/SVN-trunk-issue-with-g-list-and-g-remove-and-solution-td5164890.html
[2]
https://trac.osgeo.org/grass/browser/grass/trunk/lib/python/script/Makefile?rev=62097

Markus

11

On Sun, Jun 14, 2015 at 7:01 AM, Vaclav Petras <wenzeslaus@gmail.com> wrote:

On Sun, Jun 14, 2015 at 6:43 AM, Markus Neteler <neteler@osgeo.org> wrote:

On Sun, Jun 14, 2015 at 12:40 PM, Vaclav Petras <wenzeslaus@gmail.com>
wrote:
>...
> If you want to avoid problems run `make distclean` before
> `svn up`. For automated builds this is necessary.

Sure, I have "make distclean" in for years. Yet not sufficient if the
failing file is not defined as clean target. This rarely happens but
this time it did.

I apparently learn it from you [1]. However, the build for automated tests
did not fail with the setup.py change, so how is that possible? setup.py
was supposed to be in the OBJ file [2] if I understand the Makefile
correctly.

As I'm thinking about it I would say that I have seen setup.py in the
script dir, not script/.OBJ..., so the Makefile might have been wrong
(which explains why it fails for you, but it does not explain why it worked
for me, and this doesn't ensure that it won't happen again).

[1] http://lists.osgeo.org/pipermail/grass-dev/2014-September/070986.html
[1 nabble]
http://osgeo-org.1560.x6.nabble.com/SVN-trunk-issue-with-g-list-and-g-remove-and-solution-td5164890.html
[2]
https://trac.osgeo.org/grass/browser/grass/trunk/lib/python/script/Makefile?rev=62097

Markus

12

On Sun, Jun 14, 2015 at 1:56 AM, Vaclav Petras <wenzeslaus@gmail.com> wrote:

and recently posted log for 71/trunk binary is 404:

http://grass.osgeo.org/grass71/binary/linux/snapshot/build.log

Now back, added to cronjob:

http://grass.osgeo.org/grass70/binary/linux/snapshot/build.log
http://grass.osgeo.org/grass71/binary/linux/snapshot/build.log

Markus