jiuyuan
/
test_ok1
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
399 Commits 2 Branches 0 Tags 34 MiB
Commit Graph

164 Commits

Author SHA1 Message Date
hpk 7973f472d7 [svn r38505] clarify that config.getvalue will first lookup 
in options.

--HG--
branch : trunk
 2007-02-11 19:43:34 +01:00
hpk 14ad83da06 [svn r38496] don't print the test-mode, that needs more thought how 
to do it now

--HG--
branch : trunk
 2007-02-11 19:01:19 +01:00
hpk bda58e9862 [svn r38484] make config.option override any conftest provided value. 
--HG--
branch : trunk
 2007-02-11 18:08:13 +01:00
hpk 5c000e57f1 [svn r38474] make io capturing configurable per e.g. 
conf_iocapture = "sys"

by default it has "fd" for performing fd-based capturing

--HG--
branch : trunk
 2007-02-11 17:25:01 +01:00
hpk 22d98ac3a1 [svn r38460] removing a very-likely-obsolete test 
and otherwise fixing files and tests
to do py.test.skip()s only during setup.

--HG--
branch : trunk
 2007-02-11 14:46:46 +01:00
hpk e068a463a1 [svn r38459] two functions with the same name, use the latter one. 
--HG--
branch : trunk
 2007-02-11 14:45:49 +01:00
guido ac5c05b688 [svn r38438] Rolling back r38422 and part of r38397 because it made that a lot of stuff was 
printed to stdout/err.

--HG--
branch : trunk
 2007-02-11 02:54:22 +01:00
fijal 7dd0491764 [svn r38422] Argh. Forgotten to check that in 
--HG--
branch : trunk
 2007-02-10 22:56:23 +01:00
hpk 1095cec0dc [svn r38398] for setup/teardown pairs the correct idiom is 
setup()
    try:
        ...
    finally:
        teardown()

and not to do the setup() within the try,
because if that raises, teardown() will
likely go wrong as well etc.

--HG--
branch : trunk
 2007-02-10 18:00:58 +01:00
fijal 1739ba4f55 [svn r38397] Kill PidInfo and make boxing optional for distributed testing. 
This should make dist testing on windows possible.

--HG--
branch : trunk
 2007-02-10 17:50:47 +01:00
hpk e280dfe1f0 [svn r38392] looponfailing tricked me into committing too early 
(syntax error, bah)

--HG--
branch : trunk
 2007-02-10 16:07:30 +01:00
hpk bc009c4aef [svn r38391] check more exactly if we are working 
from a default rsync_root (checking
for rsync_root == topdir is not safe
as the topdir may be explicitely contained
in the rsync_root list)

--HG--
branch : trunk
 2007-02-10 16:05:32 +01:00
hpk d3e2c68e03 [svn r38390] refining the logic: for inplace-localhosts ("optimized localhost") 
we do not add any relative path.

--HG--
branch : trunk
 2007-02-10 15:51:25 +01:00
hpk fa94900836 [svn r38389] completing the picture: now if you don't have rsync_roots 
specified, the config.topdir is transfered but it is
transferred to the "remotepath.join(topdir.basename)"
(not actual code) to avoid random such rsyncs to
destroy/affect remote filesystem state.

--HG--
branch : trunk
 2007-02-10 15:45:41 +01:00
hpk e11e7472bc [svn r38378] snapshot commit for: each rsync root is now copied to 
remote-topdir.join(root.basename)

--HG--
branch : trunk
 2007-02-10 14:21:26 +01:00
hpk e04e08718f [svn r38352] made localhost inplace handling safer (and more 
redundant, there is an additional flag now,
and host.gw_remotepath is None for localhost-inplace
hosts)

--HG--
branch : trunk
 2007-02-10 09:52:22 +01:00
fijal 0d6d17dc96 [svn r38302] Kill _dist_import_pypy 
--HG--
branch : trunk
 2007-02-09 19:33:00 +01:00
fijal aa4896fa04 [svn r38249] It seems that there was somehow different idea in mind, but I actually 
do not get it.

Right now there is no way to override default session other than
from py.__.test import config; config.TerminalSession = 'name_to_import'
which is scary. Holger please take a look.

--HG--
branch : trunk
 2007-02-09 14:24:08 +01:00
hpk 828ef252d0 [svn r38219] looponfailing and exec are not so experimental 
--HG--
branch : trunk
 2007-02-09 00:38:07 +01:00
hpk b175e90f26 [svn r38218] i give up for now, somehow the local change 
in the FileBox affects --dist testing in ways i
can't fully understand at the moment (i would
think it only affects --boxed test runs). So
we are back to share tempdirs when running boxed
(causing failures elswhere).  We care later, i guess.

--HG--
branch : trunk
 2007-02-09 00:25:32 +01:00
hpk 07310d0799 [svn r38214] be a bit nicer: don't pollute /tmp with a new tempdir 
for each boxed test run.  Also fix the counting
of boxes (how did it ever work like this, anyway)

--HG--
branch : trunk
 2007-02-08 23:24:44 +01:00
hpk 0f37c632ac [svn r38212] have each boxed test run use its own tempdir 
--HG--
branch : trunk
 2007-02-08 23:13:59 +01:00
hpk ab6dded07a [svn r38207] revert 38182 and 38184 changes to ensuretemp() 
because getpid() is different per thread
(and if e.g. run via execnet in threads ...
it creates new tempdirectories all the time)
let's consider this sometime else

--HG--
branch : trunk
 2007-02-08 21:31:50 +01:00
hpk 3ea6a768e0 [svn r38198] almost cosmetic 
--HG--
branch : trunk
 2007-02-08 20:15:05 +01:00
hpk 479aa2e03f [svn r38197] slight reporting fix 
--HG--
branch : trunk
 2007-02-08 20:10:58 +01:00
hpk 4976889c53 [svn r38196] reducing the callback-indirections for rsyncing 
and improving the reporting for localhosts non-rsyncs

--HG--
branch : trunk
 2007-02-08 20:02:28 +01:00
hpk 6e293f593a [svn r38192] streamlining localhost optimization handling, 
and simplifying the test a bit.

--HG--
branch : trunk
 2007-02-08 19:35:18 +01:00
hpk 411157756b [svn r38184] avoid globals and add a comment 
--HG--
branch : trunk
 2007-02-08 18:39:03 +01:00
guido e89d5e5880 [svn r38182] Made that ensuretemp() creates a new dir after forks. 
--HG--
branch : trunk
 2007-02-08 18:25:32 +01:00
hpk 7fccd77b8a [svn r38176] make delete a per-host option (internally it is anyway) 
--HG--
branch : trunk
 2007-02-08 17:04:58 +01:00
fijal c18ab8fd7e [svn r38175] Fix 
--HG--
branch : trunk
 2007-02-08 17:02:15 +01:00
fijal 27cf3997e3 [svn r38173] * Minor semantics change, now host and host: are the same (we don't 
want to rsync to home dir usually)
* Make a flag rsync_flag in hostinfo which tells whether to rsync
  or no, semantics are that in case of localhost we do not rsync
  and in case of localhost: we do.

--HG--
branch : trunk
 2007-02-08 16:56:33 +01:00
fijal 0e8510648f [svn r38164] Move source as an rsync initialisation argument rather than 
as a send parameter

--HG--
branch : trunk
 2007-02-08 16:31:38 +01:00
hpk 938680d50f [svn r38125] run unboxed by default, split/improve tests for it 
--HG--
branch : trunk
 2007-02-07 22:56:14 +01:00
fijal ee04997eb2 [svn r38124] No, this is needed for C-c to work. 
--HG--
branch : trunk
 2007-02-07 22:35:14 +01:00
fijal cbfca77f39 [svn r38114] Kill fatals() 
to be considered in future

--HG--
branch : trunk
 2007-02-07 21:14:06 +01:00
fijal 1e6016ccc2 [svn r38113] Make fatal() attribute which makes boxed tests catch SystemExit, while 
in-process let it slide.

--HG--
branch : trunk
 2007-02-07 21:04:01 +01:00
fijal 632cc727e4 [svn r38112] This time don't catch SystemExit. Increases stability 
--HG--
branch : trunk
 2007-02-07 20:53:03 +01:00
hpk f07e482ddb [svn r38109] bailing out the optimize_localhost approach 
to consider it differently

--HG--
branch : trunk
 2007-02-07 20:36:41 +01:00
fijal 1d6a72d936 [svn r38107] * Have optimise_localhost flag and a test for it (but not turning it on 
by default, this will be done later)
* two other occurences of possibly-failing channel.send

--HG--
branch : trunk
 2007-02-07 20:24:38 +01:00
hpk d791086561 [svn r38106] making tests independent of --dist implications 
--HG--
branch : trunk
 2007-02-07 20:16:28 +01:00
fijal 8258b40fd6 [svn r38105] Make errors bright red and bold, easier to spot. 
--HG--
branch : trunk
 2007-02-07 20:14:10 +01:00
fijal 5b188f1167 [svn r38103] Hum. This test was just broken. Fix it and add additional checks. 
--HG--
branch : trunk
 2007-02-07 20:02:21 +01:00
fijal c5df4c335c [svn r38102] Skip this test not to interfere with py.test.config.boxed 
--HG--
branch : trunk
 2007-02-07 19:56:48 +01:00
fijal 3bb0c2b433 [svn r38101] oops, this will be implied anyway 
--HG--
branch : trunk
 2007-02-07 19:56:10 +01:00
fijal d7331146da [svn r38100] Slave is always boxed as well (different level of skips) 
--HG--
branch : trunk
 2007-02-07 19:55:39 +01:00
fijal ee1214fcfe [svn r38099] * RSession cannot work with nocapture, LSession can 
* RSession is always boxed

--HG--
branch : trunk
 2007-02-07 19:54:51 +01:00
fijal ecd605e3b2 [svn r38098] A test and typo discovered by that test. 
--HG--
branch : trunk
 2007-02-07 19:44:49 +01:00
hpk 77b8a010c5 [svn r38097] streamline boxed test configuration: 
"config.option.boxed"  indicates now boxed tests
and RSession.fixoptions takes care to honour
dist_boxed accordingly.  So you can do

    if not py.test.config.boxed:
        py.test.skip(...)

i also fixed the documentation.

--HG--
branch : trunk
 2007-02-07 19:43:23 +01:00
fijal ae9ffd2c19 [svn r38096] typo 
--HG--
branch : trunk
 2007-02-07 19:40:16 +01:00