Age | Commit message (Collapse) | Author |
|
OTP-11046
|
|
* siri/cuddle-with-tests:
[sasl] Fix tests that compare execution time
[jinterface] Kill process between tests to avoid consequential error
[jinterface] Make tests handle space in path on windows
[common_test] Extend timers for slave nodes in ct_cover_SUITE
[test_server] Clean up unused code in test_server_SUITE
[reltool] Increase timetrap timeout for reltool_server_SUTIE to 10 minutes
[test_server] Fix encoding of HTML file references to work on windows
[test_server] Use simpler tc name in unicode test to make it work on windows
|
|
After removing the functionality for remote target nodes in R16A,
there was still some unused code left and some documentation that was
not updated. This is now completed.
|
|
The URI encoding when in unicode file name mode did not preserve @, :
and /, and thus some links were did not work on windows (think
c:/). This has been corrected.
|
|
|
|
The unicode update of test_server for R16A introduced a few potential
errors when logging to files. Sometimes ~tp or ~ts was used for
formatting also when writing to files that were not opened with the
{encoding,utf8} option. If then the argument contained unicode
characters above 255, the file descriptor would crash. This has been
corrected by the following modifications:
* Since the 'unexpected_io' log file is used only when the test case
HTML file is not available (e.g. between test cases), this file is
now also a HTML file and as other test_server HTML logs it is always
UTF-8 encoded
* Since it is possible to change which information is going to which
log file (with test_server_ctrl:set_levels/3), we do not have full
control over which information is written to which file. This means
that any printout could be written to the 'major' log file
(suite.log), which was earlier encoded as latin1. To avoid crashing
this file descriptor due to unicode strings, the 'major' log file is
now also encoded in UTF-8 (possible incopatibility).
* The cross_cover.info file is no longer a text file which can be read
with file:consult/1, instead it is written as a pure binary file
using term_to_binary when writing and binary_to_term when reading.
* The encoding of the file named 'last_name', which only content is
the path to the last run.<timestamp> directory, is now dependent on
the file name mode of the VM. If file names are expected to be
unicode, then the 'last_name' file is UTF-8 encoded, else it is
latin1 encoded.
Also, ~tp is changed back to ~p unless it is somehow likely that the
argument includes strings. It is not obvious that this is the correct
thing to do, but some decission had to be taken...
|
|
If escaping :, then links of type "file:///c:/ldisk..." will not work.
Escaping @ does not create any problems, but the link becomes harder
to read and most modern browsers can handle this.
|
|
When running under framework common_test, the encoding statement in
the HTML header of the file generated with erl2html2:convert would
sometimes be faulty. This has been corrected.
|
|
* Use UTF-8 encoding for all HTML files, except the HTML version of
the test suite generated with erl2html2:convert, which will have the
same encoding as the original test suite (.erl) file.
* Encode link targets in HTML files, allowing both special characters
like "/", "&", "?" etc, and latin1 or unicode characters.
* Use unicode modifier 't' with ~s and ~p when appropriate.
* Use unicode:characters_to_list and unicode:characters_to_binary for
conversion between binaries and strings instead of binary_to_list
and list_to_binary.
|
|
* siri/cross-cover/OTP-9870:
[test_server] Don't use print/3 in cross_cover_analyse
[common_test] Add documentation for cross cover analysis
[common_test] Add test case for cross cover mechanism
[test_server] Update documentation about cross cover
[test_server] Add test of code cover mechanism
[test_server,common_test] Fix cross cover mechansim
|
|
|
|
Cross cover analysis is executed outside the scope of
test_server_ctrl, i.e. test_server_ctrl is not necessarily
started. This may cause test_server_ctrl:print/3 to hang. Instead use
io:fwrite directly (as it was before).
|
|
Update the interface for cross cover analysis (collection of cover
data over multiple tests) so it can be used via common_test and
directly with test server.
The concept of 'application' in the cross cover interface is removed
and replaced with an arbitrary Tag=atom() which identifies a test run.
|
|
Line numbering of erlang files that were not correctly indented could
be wrong after coverting to html with erl2html2:convert/[2,3]. This has
been corrected.
This commit also fixes the following:
* There are now link targets for each line and not only for each 10th
line - meaning that links from test logs are now to the exact line,
and not to the last number for which N rem 10 == 0.
* there will only be one link target per function, i.e. the faulty
link targets for function clauses are removed.
* link targets for function now includes the arity (e.g. func/1 has a
link target "func-1")
And some tests are added.
|
|
* bjorn/ct/minor-corrections:
test_server_io: Correct a few comments
Handle ct:abort_current_testcase/1 when executing parallel groups
Make sure that "Cover analysing..." is written to stdout
Make sure that "Testing..." is not written if testing has finished
|
|
The support for remote target nodes has not worked in a long time.
We are unlikely to ever need remot target node support and if
we'll need it will be easier to start over from scratch.
|
|
ct:abort_current_testcase/1 (which is a wrapper for the function
with the same name in test_server_ctrl) would not work as expected
when a group with parallel test cases was executing. Essentially,
the abort_current_testcase message would be ignored until the group
has finished and would then abort the end_per_group testcase for
the group.
Since there is no unique current testcase when a parallel group is
executing, we must handle the abort_current_testcase message and
return an error tuple.
Also fix the bug that test_server_ctrl:abort_current_testcase/1
would always return 'ok'.
|
|
When writing messages to stdout, all messages are prefixed with
"Testing" followed by the name of the test suite. We don't want
to write that prefix if testing has stopped.
|
|
'user_timetrap_error' was not detected during any "conf init function"
except init_per_suite, so if e.g. init_per_group failed with
user_timetrap_error, then the test cases in the group would not be
skipped. This has been corrected.
|
|
By default, test_server will always stop cover (and thus load back the
non cover compiled original beam files) after a test with code
coverage analysis is completed. The new option allows the common_test
user to specify that cover shall not be stopped. This can be useful if
there are processes still running old code, i.e. processes that have
not done any fully qualified function call after the cover
compilation, since loading the original code then will kill those
processes. This is only recommended if the erlang node is to be
stopped after the test run, or if cover can be manually stopped.
|
|
The cross cover analysis functionality in test_server was broken since
OTP converted to running daily tests through common_test. This has
been corrected.
|
|
In test_server_ctrl:start_node and wait_for_node, the timer in
controller_call was not multiplied by the timetrap_scale_factor. This
caused unexpected behaviour since the scale factor was used furhter
down the the code, and timeout would happen in unexpected order when
cover was running.
|
|
* bjorn/ct/separate-io-server/OTP-10101:
Test ct:capture/start/stop/get
Introduce ct_group_leader_SUITE that does nasty things with group leaders
Eliminate unexpected I/O to the minor log file
Introduce test_server_io and test_server_gl
Clean up initialization of parallel test cases
Introduce is_io_buffered/0 to somewhat improve readability
test_server_ctrl: Consistently use set_io_buffering/1
|
|
If the test suite itself was included in code coverage analysis, then
test_server_ctrl would not manage to set data_dir correctly for the
test, since it relied on the result of code:which(Suite). This has
been corrected.
|
|
|
|
|
|
|
|
Note that there are some more direct use of:
get(test_server_common_io_handler)
that cannot be replaced with a call to is_io_buffered/0.
|
|
Since we will want to change the implementation of I/O buffering in
a future commit, make sure that all updates of the buffering state
is done by calling set_io_buffering/1.
|
|
* bjorn/test_server/dialyzer-fixes:
test_server_ctrl: Don't use undocumented features of io:format()
Fix obsolete assumption for the return type of os:type/0
|
|
* dgud/ct/fix-trashed-logs/OTP-10369:
[ct] Shorten testcase names
Shorten testcase names
[ct] Block output into single io call
|
|
The Format string argument for io:format/3 is not documented
to accept an iolist, so we should not depend on it. Fix the
problem by using io:put_chars/2 instead.
|
|
Add newlines to the same io function call,
so that concurrent io always starts at a new line.
|
|
If a severe error occurs in test_server (e.g. failing to write to
log files), test_server would terminate without commont_test knowing
about it. Since ct_run can now return an exit code, it is important
that common_test is aware of severe problem so that it can indicate
that an error has occurred.
|
|
|
|
If it is not possible for test_server_ctrl:start_log_file/0 to
create the log file directory or log files, then stop_extra_tools/1
(called from init_tester/10) is also likely to crash and hide the
original error reason. That will be confusing to the user.
Therefore, catch the call to stop_extra_tools/1.
|
|
|
|
* origin/peppe/common_test/verbosity_level:
Fix doc build error
Implement verbosity levels and parameter for log printout importance
Conflicts:
lib/common_test/src/ct_logs.erl
lib/common_test/src/ct_run.erl
lib/common_test/src/ct_testspec.erl
lib/common_test/src/ct_util.hrl
lib/common_test/test/Makefile
OTP-9625
OTP-10067
|
|
|
|
|
|
|
|
OTP-9896
Introduce java scripts in the web pages to enable sorting
of table elements.
|
|
|
|
|
|
Now that calls to ct_framework:init/end_per_suite take place
whenever init/end_per_suite is missing in the test suite,
these calls should trigger init/end_tc to perform configuration
(such as calling calling CTH functions, suite/0, etc).
|
|
|
|
The new values for create_priv_dir are:
auto_per_run, auto_per_tc, manual_per_tc
|
|
Introduce the optional feature to have Test Server generate
priv_dir directory names that are unique for each test case
or config function. The name of the option/flag is
'unique_priv_dir' and it can be set to value 'auto' or
'manual'. If auto, Test Server creates each priv_dir
automatically (can be expensive in case of many and/or repeated
cases). If manual, the user needs to create the priv_dir
explicitly by calling ct:make_priv_dir/0.
|
|
For each repeated test during a test run, an entry is created
in the index.html file (i.e. the overview file for the test run).
In the top level (logdir) index file, only the last test result is
listed.
E.g, given the test spec:
[{merge_tests,false},{dirs,testobj1},{dirs,testobj1}]
In the index file for the test run (under Logdir/ct_run.Node.Date.Time),
both dir tests are listed. In the top level index file (under Logdir),
only the last test is listed (one has to find all results through the
all_runs.html file).
|
|
OTP-9730
|