In this case, I assume using your favorite editor to add two 'u' in
sql/test/mapi/Tests/python_dbapi.stable.out
should be sufficient --- or in other works, there is no "magic" involved.
i discovered some of the testcases where not correctly using TSTHOSTNAME environment variable and such. I made the connection error a bit more useful and fixed the issue with env var.
is this issue solved now? Probably I didn't invest enough time to get to know the test infrastructure web interface, but I can't find a failing python test for the last test run.
Date: 2012-07-20 08:14:33 +0200
From: @drstmane
To: clients devs <>
Version: -- development
CC: @gijzelaerr, @mlkersten
Last updated: 2013-02-19 13:18:00 +0100
Comment 17518
Date: 2012-07-20 08:14:33 +0200
From: @drstmane
Since recent Python API changes, two Python tests fail; cf.,
http://monetdb.cwi.nl/testweb/web/testgrid.php?serial=44702:17e97402292f&order=platform,arch,compiler&targets=&module=sql&tstlimit=test/mapi
... at least I assume that there is a relation between the changes and the failing tests ...
Comment 17519
Date: 2012-07-20 08:35:48 +0200
From: @drstmane
Test
has been failing since changeset d3c1ae10f807 (http://dev.monetdb.org/hg/MonetDB/rev/d3c1ae10f807);
initially like this
http://monetdb.cwi.nl/testweb/tests/44602:f8a0abfbd2d7/GNU-Fedora-x86_64-propcheck-assert/sql/mTests/test/mapi/python_test_monetdb_sql.err.diff.html
now like this
http://monetdb.cwi.nl/testweb/tests/44702:17e97402292f/GNU-Fedora-x86_64-propcheck-assert/sql/mTests/test/mapi/python_test_monetdb_sql.err.diff.html
Comment 17520
Date: 2012-07-20 08:37:04 +0200
From: @drstmane
(In reply to comment 1)
I meant to say/type
Comment 17526
Date: 2012-07-20 14:15:20 +0200
From: @gijzelaerr
The python_dbapi should be updated to match the output, since all strings outputted by the API are always unicode now.
The failing python_test_monetdb_sql is due to a incorrect import in the test, it is fixed nou in the default branch.
Comment 17527
Date: 2012-07-20 14:24:10 +0200
From: @drstmane
Thanks, Gijs!
Did you (or can you) also update the test output of python_dbapi
(since you already calosed this bug report)?
Thanks!
Stefan
Comment 17528
Date: 2012-07-20 14:27:37 +0200
From: @gijzelaerr
sorry, I have no idea how to that, I was so modest to assume you would do that :)
Comment 17529
Date: 2012-07-20 15:08:32 +0200
From: @drstmane
In this case, I assume using your favorite editor to add two 'u' in
sql/test/mapi/Tests/python_dbapi.stable.out
should be sufficient --- or in other works, there is no "magic" involved.
Just in case, more info about MonetDB's testing environment is available at
http://dev.monetdb.org/hg/MonetDB/file/def966c3b4e7/testing/README
;-)
... but of course, I can also add the two 'u' in sql/test/mapi/Tests/python_dbapi.stable.out if you prefer that ... ;-)
S.
Comment 17541
Date: 2012-07-21 07:32:09 +0200
From: @drstmane
Test python_test_monetdb_sql
still (again?) fails in default branch:
http://monetdb.cwi.nl/testweb/web/testgrid.php?serial=44732:ee63516d78f5&order=platform,arch,compiler&targets=&module=sql&tstlimit=test/mapi
http://monetdb.cwi.nl/testweb/tests/44732:ee63516d78f5/GNU-Fedora-x86_64-propcheck-assert/sql/mTests/test/mapi/python_test_monetdb_sql.err.diff.html
Comment 17542
Date: 2012-07-21 07:39:45 +0200
From: @drstmane
In fact, test python_test_monetdb_sql
fails since changeset d3c1ae10f807 "Split the timezones catalog"
(http://dev.monetdb.org/hg/MonetDB/rev/d3c1ae10f807),
i.e., not necessarily since / due to python api changes ...
Comment 17543
Date: 2012-07-21 10:37:14 +0200
From: @gijzelaerr
Hm this is weird. i'll look at this on monday.
Comment 17544
Date: 2012-07-21 10:53:05 +0200
From: @drstmane
Maybe, the errors are not related to changeset d3c1ae10f807 after all,
as the errors back than were different from those now;
cf.,
http://monetdb.cwi.nl/testweb/tests/44602:f8a0abfbd2d7/GNU-Fedora-x86_64-propcheck-assert/sql/mTests/test/mapi/python_test_monetdb_sql.err.diff.html
vs.,
http://monetdb.cwi.nl/testweb/tests/44732:ee63516d78f5/GNU-Fedora-x86_64-propcheck-assert/sql/mTests/test/mapi/python_test_monetdb_sql.err.diff.html
Comment 17557
Date: 2012-07-23 11:28:18 +0200
From: @gijzelaerr
i discovered some of the testcases where not correctly using TSTHOSTNAME environment variable and such. I made the connection error a bit more useful and fixed the issue with env var.
Comment 17561
Date: 2012-07-24 09:41:28 +0200
From: @gijzelaerr
is this issue solved now? Probably I didn't invest enough time to get to know the test infrastructure web interface, but I can't find a failing python test for the last test run.
Comment 17562
Date: 2012-07-24 09:50:18 +0200
From: @drstmane
Yes, after your recent fixes and my (hopefully correct) approval, cf.,
http://dev.monetdb.org/hg/MonetDB/rev/5f659ae8c1a1 ,
the python tests appear to work fine, now:
http://monetdb.cwi.nl/testweb/web/testgrid.php?serial=44753:7968805ec43d&order=platform,arch,compiler&module=sql&tstlimit=test/mapi&nocondense
Thanks!
ps:
feel free to try
cd .../MonetDB/sql/test/mapi/Tests && Mtest.py
See also
Mtest.py --help
;-)
Comment 18515
Date: 2013-02-19 13:18:00 +0100
From: @sjoerdmullender
Feb2013 has been released.
The text was updated successfully, but these errors were encountered: