You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by HuiHuang <ye...@yahoo.com.cn> on 2009/06/11 11:40:08 UTC
[PATCH] commit_tests
Hi,
log message:
[[[
* subversion/tests/cmdline/commit_tests.py
Add a test case for committing from multiple WC which have no common
parent path. It is XFail now and will be changed to Pass after we have fix
issue #2381.
]]]
Huihuang
2009-06-11
yellow.flying
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361248
Re: [PATCH] commit_tests
Posted by HuiHuang <ye...@yahoo.com.cn>.
Hi Stefan,
I am sorry. I know where I make a mistake:
+# Commit multiple working copies which are [[not]] from the same repository. This
+# is relate to issue #2381
There should be no "not" in the sentence. I think this is the reason which cause your
misunderstanding. Do you think this test is necessary? If it is necessary, I will create
the patch again, otherwise, I will change this test to commit_multiple_wc_multiple_repos
as you indicated before.
Thanks~
Huihuang
------------------
yellow.flying
2009-06-13
__________________________________________________
赶快注册雅虎超大容量免费邮箱?
http://cn.mail.yahoo.com
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361774
Re: [PATCH] commit_tests
Posted by HuiHuang <ye...@yahoo.com.cn>.
Thank you Neels, there is confusion and we make it clear now. Thanks for your help.
Best Wishes~
Huihuang
------------------
yellow.flying
2009-06-14
__________________________________________________
赶快注册雅虎超大容量免费邮箱?
http://cn.mail.yahoo.com
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361896
Re: [PATCH] commit_tests
Posted by Neels Janosch Hofmeyr <ne...@elego.de>.
Stefan Sperling wrote:
[...]
> +# Also related to issue #2381, "Cannot commit multiple WC paths which
> +# lack a common parent directory". Once that issue has been resolved,
> +# we will remove the XFail from this test.
[...]
> + # Commit should succeed, even though one target is a "child" of the other,
Stefan, I think HuiHuang wants them *not* to be parent-child related. His
point is that he also wants to test the case where we commit two WC paths
which are adjacent, but whose parent dir is not a WC itself.
[...]
> Do you see the difference between doing the above and what you proposed?
Still, HuiHuang, Stefan is right here. The tests should expect the desired
behaviour, no matter if the current svn really behaves like that.
If the current svn behaves wrong, that test fails.
If we know that the test will currently fail, we mark it XFail(..)
As soon as the behaviour is fixed, we see the test pass.
Good speed for you, HuiHuang! :)
~Neels
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361885
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Sat, Jun 13, 2009 at 02:22:01PM +0100, Stefan Sperling wrote:
> In tests, we usually want the failure or success to come from
> an invocation of svntest.actions.run_and_verify_status().
Slight correction:
There are more "verify" functions than run_and_verify_status(),
of course. E.g. there is also run_and_verify_svn().
Any run_and_verify_*() function will do.
Stefan
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Sat, Jun 13, 2009 at 02:22:01PM +0100, Stefan Sperling wrote:
> So, to be clear, I meant something like:
> @@ -1384,14 +1387,16 @@
> expected_status2.tweak('A/B/lambda', status='M ')
> svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
>
> - # Commit should fail, even though one target is a "child" of the other.
> - svntest.actions.run_and_verify_svn("Unexpectedly not locked",
> - None, svntest.verify.AnyOutput,
> + # Commit should succeed, even though one target is a "child" of the other,
> + # since both working copies come from the same repository.
> + svntest.actions.run_and_verify_svn(None, None, svntest.verify.AnyOutput,
> 'commit', '-m', 'log',
> wc_dir, wc2_dir)
This example was actually totally wrong.
The function call contradicts the comment above it,
and uses None instead of [].
So this led you to copy those errors into the patch you submitted.
I will try to be more careful when providing examples in the future.
I even ended up committing this wrong code in r38030 :(
What I really meant to say is in r38031.
I hope I didn't confuse you too much,
Stefan
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Sat, Jun 13, 2009 at 01:54:25PM +0800, HuiHuang wrote:
> Hey Stefan,
> >So instead of adding a new test, we can change the commit_multiple_wc
> >test to expect the commit to succeed instead of fail, and mark the
> >test XFail. Can you make a patch for that?
>
> Do you mean that I make a patch like this?
>
> Index: commit_tests.py
> ===================================================================
> --- commit_tests.py (版本 38022)
> +++ commit_tests.py (工作副本)
> @@ -1393,6 +1393,7 @@
> # Verify status unchanged
> svntest.actions.run_and_verify_status(wc_dir, expected_status)
> svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
> + raise svntest.Failure
Raising Failure like that is almost never the right thing to do!
In tests, we usually want the failure or success to come from
an invocation of svntest.actions.run_and_verify_status().
Otherwise we are not testing Subversion. We are just writing
code which raises exceptions but doesn't test anything.
So, to be clear, I meant something like:
Index: subversion/tests/cmdline/commit_tests.py
===================================================================
--- subversion/tests/cmdline/commit_tests.py (revision 38018)
+++ subversion/tests/cmdline/commit_tests.py (working copy)
@@ -1353,8 +1353,11 @@
# Commit from multiple working copies is not yet supported. At
# present an error is generated and none of the working copies change.
-# Related to issue 959, this test here doesn't use svn:externals but the
+# Related to issue #959, this test here doesn't use svn:externals but the
# behaviour needs to be considered.
+# Also related to issue #2381, "Cannot commit multiple WC paths which
+# lack a common parent directory". Once that issue has been resolved,
+# we will remove the XFail from this test.
def commit_multiple_wc(sbox):
"attempted commit from multiple wc fails"
@@ -1384,14 +1387,16 @@
expected_status2.tweak('A/B/lambda', status='M ')
svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
- # Commit should fail, even though one target is a "child" of the other.
- svntest.actions.run_and_verify_svn("Unexpectedly not locked",
- None, svntest.verify.AnyOutput,
+ # Commit should succeed, even though one target is a "child" of the other,
+ # since both working copies come from the same repository.
+ svntest.actions.run_and_verify_svn(None, None, svntest.verify.AnyOutput,
'commit', '-m', 'log',
wc_dir, wc2_dir)
- # Verify status unchanged
+ # Verify status changed
+ expected_status.tweak('A/mu', status=' ', wc_rev=2)
svntest.actions.run_and_verify_status(wc_dir, expected_status)
+ expected_status2.tweak('A/B/lambda', status=' ', wc_rev=2)
svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
@@ -2696,7 +2701,7 @@
commit_from_long_dir,
commit_with_lock,
commit_current_dir,
- commit_multiple_wc,
+ XFail(commit_multiple_wc),
commit_nonrecursive,
failed_commit,
commit_out_of_date_deletions,
Do you see the difference between doing the above and what you proposed?
Stefan
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Sat, Jun 13, 2009 at 10:34:06AM +0800, yellow.flying wrote:
> >Wait, there is something wrong in what we are doing!
> >It looks like I didn't give you good advice :(
> >After looking at commit_tests.py more closely, I realised that there
> >already is a test that does what we need!
> >It is the "commit_multiple_wc" test below which you put your test.
> >So instead of adding a new test, we can change the commit_multiple_wc
> >test to expect the commit to succeed instead of fail, and mark the
> >test XFail. Can you make a patch for that?
>
> I think you misunderstand me. Yes, I know that there is a test already,
> but I do not think the test is enough, you may observed that in the
> test the two wcs are nested, and I want test when the two wcs(which
> are from the same repository) are not nested. So I write
> commit_multiple_wc2. These two tests will both be used in our case.
>
> Do you think it is necessary to do like this?
OK, agreed.
Sorry about the confusion, I didn't look close enough to realise
the working copies were nested.
Maybe we should rename the first test to commit_multiple_wc_nested?
Stefan
Re: [PATCH] commit multiple wc
Posted by Stefan Sperling <st...@elego.de>.
On Sun, Jun 14, 2009 at 01:09:46PM +0100, Stefan Sperling wrote:
> On Sun, Jun 14, 2009 at 08:53:19AM +0800, yellow.flying wrote:
> > @@ -1384,17 +1388,72 @@
> > expected_status2.tweak('A/B/lambda', status='M ')
> > svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
> >
> > - # Commit should fail, even though one target is a "child" of the other.
> > - svntest.actions.run_and_verify_svn("Unexpectedly not locked",
> > - None, svntest.verify.AnyOutput,
> > + # Commit should fail, even though one target is a "child" of the other,
> > + # since targets come from different wcs.
> > + svntest.actions.run_and_verify_svn(None, None, svntest.verify.AnyOutput,
> > 'commit', '-m', 'log',
> > wc_dir, wc2_dir)
>
> In this hunk, you have changed the function to expect the commit
> to succeed, but you didn't update the comment above the function!
Hrmmm... I was too quick again... :(
You actually wrote the test to expect failure, so the comment
is correct. But we want the test to check for the situation we
will have when issue #2381 is solved. See r38030.
So, please forget my comments about not changing the comments :)
Sorry about that,
Stefan
Re: [PATCH] commit multiple wc
Posted by Stefan Sperling <st...@elego.de>.
On Sun, Jun 14, 2009 at 08:53:19AM +0800, yellow.flying wrote:
> Hey Stefan,
>
> I am sorry. I can understand what you mean now. I rewrite the patch
> commit_multiple_wc, and after this patch is submitted, I will write another
> patche-- commit_multiple_wc_multiple_repos.
>
> log message:
>
> [[[
>
> * subversion/tests/cmdline/commit_tests.py
> (commit_multiple_wc_nested): Rename commit_multiple_wc as
> commit_multiple_wc_nested, and mark it as XFail. It will be changed to Pass
> after we have fix issue #2381.
> (commit_multiple_wc_not_nested): new test, This test is the same as
> commit_multiple_wc_nested except that two wcs in this test are not nested.
> ]]]
>
> I hope this time it is ok，thank you.
It is mainly OK, except a few minor details. I have committed a patch
based on your patch in r38030, and also added another test.
Please carefully look at the patch I committed to make sure you
agree with it. If you see any problem please tell me!
Also please look at the log message carefully. For example,
your log message didn't mention the changes you made to test_list.
A good rule is to remember is that you need to mention *every* hunk
of the patch in your log message (a "hunk" starts with an @@ ... @@ line),
describing the change made by the hunk.
Some comments on your patch:
> Index: commit_tests.py
> ===================================================================
> --- commit_tests.py (°æ±¾ 38025)
> +++ commit_tests.py (¹¤×÷¸±±¾)
If you can, please create patches from the working copy root,
so that full path subversion/tests/cmdline/commit_tests.py
appears in the patch. But that is not very important.
> @@ -1384,17 +1388,72 @@
> expected_status2.tweak('A/B/lambda', status='M ')
> svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
>
> - # Commit should fail, even though one target is a "child" of the other.
> - svntest.actions.run_and_verify_svn("Unexpectedly not locked",
> - None, svntest.verify.AnyOutput,
> + # Commit should fail, even though one target is a "child" of the other,
> + # since targets come from different wcs.
> + svntest.actions.run_and_verify_svn(None, None, svntest.verify.AnyOutput,
> 'commit', '-m', 'log',
> wc_dir, wc2_dir)
In this hunk, you have changed the function to expect the commit
to succeed, but you didn't update the comment above the function!
Please make sure to always keep comments in sync with code you are
changing. Otherwise, someone might look at the code and the comment
later and get confused.
Also, when you pass "None" for stdin or stderr to a run_and_verify_*
function, the function will throw an exception saying "stderr cannot
be None". This will make your test see to XFail for a different reason
than you want it to fail for. If you don't expect output, pass an empty
list [], not None. See my commit for examples.
In general, always make sure you understand why a test is failing.
This is not easy with XFAIL tests.
For example, in the commit_multiple_wc test, we expect Subversion
to complain about the parent of the working copies not being
a working copy. Running the test, you will see that this is what
really happens:
[~/svn/svn-trunk/subversion/tests/cmdline] $ ./commit_tests.py 26
subversion/svn/commit-cmd.c:137: (apr_err=155007)
subversion/libsvn_client/commit.c:1522: (apr_err=155007)
subversion/libsvn_wc/lock.c:593: (apr_err=155007)
subversion/libsvn_wc/lock.c:484: (apr_err=155007)
svn: '/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svn-test-work/working_copies/commit_tests-26' is not a working copy
Here you see svn complaining about the right thing.
If you see any different error, the test failed, but not for the correct
reason!
Traceback (most recent call last):
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/main.py", line 1141, in run
rc = self.pred.run(sandbox)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/testcase.py", line 61, in run
return self._delegate.run(sandbox)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/testcase.py", line 121, in run
return self.func(sandbox)
File "./commit_tests.py", line 1439, in commit_multiple_wc
wc1_dir, wc2_dir)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/actions.py", line 199, in run_and_verify_svn
expected_exit, *varargs)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/actions.py", line 232, in run_and_verify_svn2
exit_code, out, err = main.run_svn(want_err, *varargs)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/main.py", line 579, in run_svn
*(_with_auth(_with_config_dir(varargs))))
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/main.py", line 357, in run_command
None, *varargs)
File "/home/stsp/svn/svn-trunk/subversion/tests/cmdline/svntest/main.py", line 512, in run_command_stdin
raise Failure
Failure
XFAIL: commit_tests.py 26: commit from two working copies
So just looking at the last line does not give any information
when we are dealing with XFAIL tests. That makes it a bit hard
to work with them. It's much easier when tests are expected to PASS.
You can get even more detailed test output by passing the --verbose
option to the test script. This often helps to understand better
why a test is failing. Try running:
commit_tests.py --verbose 25
> + # Commit should fail, since targets come from different wcs.
> + svntest.actions.run_and_verify_svn(None, None, svntest.verify.AnyOutput,
> + 'commit', '-m', 'log',
> + wc1_dir, wc2_dir)
Here, you also forgot to change the comment.
> @@ -2696,7 +2755,8 @@
> commit_from_long_dir,
> commit_with_lock,
> commit_current_dir,
> - commit_multiple_wc,
> + XFail(commit_multiple_wc_nested),
> + XFail(commit_multiple_wc_not_nested),
> commit_nonrecursive,
> failed_commit,
> commit_out_of_date_deletions,
The above change was not mentioned in your log message (I already
said this above).
Overall, the patch was good!
And I hope the tests will help you writing code that actually
solves issue #2381.
Thanks :)
Stefan
[PATCH] commit multiple wc
Posted by HuiHuang <ye...@yahoo.com.cn>.
Hey Stefan,
I am sorry. I can understand what you mean now. I rewrite the patch
commit_multiple_wc, and after this patch is submitted, I will write another
patche-- commit_multiple_wc_multiple_repos.
log message:
[[[
* subversion/tests/cmdline/commit_tests.py
(commit_multiple_wc_nested): Rename commit_multiple_wc as
commit_multiple_wc_nested, and mark it as XFail. It will be changed to Pass
after we have fix issue #2381.
(commit_multiple_wc_not_nested): new test, This test is the same as
commit_multiple_wc_nested except that two wcs in this test are not nested.
]]]
I hope this time it is ok，thank you.
Huihuang
------------------
yellow.flying
2009-06-14
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361895
Re: [PATCH] commit_tests
Posted by HuiHuang <ye...@yahoo.com.cn>.
>Wait, there is something wrong in what we are doing!
>It looks like I didn't give you good advice :(
>After looking at commit_tests.py more closely, I realised that there
>already is a test that does what we need!
>It is the "commit_multiple_wc" test below which you put your test.
>So instead of adding a new test, we can change the commit_multiple_wc
>test to expect the commit to succeed instead of fail, and mark the
>test XFail. Can you make a patch for that?
I think you misunderstand me. Yes, I know that there is a test already,
but I do not think the test is enough, you may observed that in the
test the two wcs are nested, and I want test when the two wcs(which
are from the same repository) are not nested. So I write
commit_multiple_wc2. These two tests will both be used in our case.
Do you think it is necessary to do like this?
Best Regards!
Huihuang
------------------
yellow.flying
2009-06-13
-------------------------------------------------------------
__________________________________________________
赶快注册雅虎超大容量免费邮箱?
http://cn.mail.yahoo.com
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361761
Re: [PATCH] commit_tests
Posted by HuiHuang <ye...@yahoo.com.cn>.
Hey Stefan,
>So instead of adding a new test, we can change the commit_multiple_wc
>test to expect the commit to succeed instead of fail, and mark the
>test XFail. Can you make a patch for that?
Do you mean that I make a patch like this?
Index: commit_tests.py
===================================================================
--- commit_tests.py (版本 38022)
+++ commit_tests.py (工作副本)
@@ -1393,6 +1393,7 @@
# Verify status unchanged
svntest.actions.run_and_verify_status(wc_dir, expected_status)
svntest.actions.run_and_verify_status(wc2_dir, expected_status2)
+ raise svntest.Failure
def commit_nonrecursive(sbox):
@@ -2696,7 +2697,7 @@
commit_from_long_dir,
commit_with_lock,
commit_current_dir,
- commit_multiple_wc,
+ XFail(commit_multiple_wc),
commit_nonrecursive,
failed_commit,
commit_out_of_date_deletions,
Huihuang
------------------
yellow.flying
2009-06-13
-------------------------------------------------------------
��i��'�ē扫h∈&
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361773
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Fri, Jun 12, 2009 at 09:47:40AM +0800, HuiHuang wrote:
> Hey Stefan,
>
> Yes, I agree with you. Actually the test is passed in current state.
Wait, there is something wrong in what we are doing!
It looks like I didn't give you good advice :(
After looking at commit_tests.py more closely, I realised that there
already is a test that does what we need!
It is the "commit_multiple_wc" test below which you put your test.
So instead of adding a new test, we can change the commit_multiple_wc
test to expect the commit to succeed instead of fail, and mark the
test XFail. Can you make a patch for that?
The test you wrote is still useful, however.
Your test tries to test the case where the two working copies do
not come from the same repository. That is OK, we will need this
test, too. You could add this test in another patch.
Note that when you wrote your test, you made both working copies
come from the same repository!
+ # Checkout two working copies
+ wc1_dir = os.path.join(wc_dir, 'wc1')
+ wc2_dir = os.path.join(wc_dir, 'wc2')
+ url = sbox.repo_url
Here, you define one repository URL.
+ svntest.actions.run_and_verify_svn("Output on stderr where none
expected",
+ svntest.verify.AnyOutput, [],
+ 'checkout',
+ url, wc1_dir)
Here you use the URL to checkout WC1.
+ svntest.actions.run_and_verify_svn("Output on stderr where none
expected",
+ svntest.verify.AnyOutput, [],
+ 'checkout',
+ url, wc2_dir)
And here use the same URL again to checkout WC2.
That doesn't look right to me.
Instead, we need two different repositories, with two different URLs,
and checkout each WC from a different URL.
I suggest you take the commit_multiple_wc test as it is now
(it expects failure), copy it to a new test called commit_multiple_wc2
or commit_multiple_wc_multiple_repos or something like that,
and change to use a separate repository for each working copy.
That is probably easier than writing the test from scratch.
There might be other tests using more than one repository,
I don't know. You could try to find examples for how to do this.
Or maybe if you read carefully through the code in
subversion/tests/cmdline/svntest/sandbox.py
you might be able to find out how this can be done.
Stefan
Re: [PATCH] commit_tests
Posted by HuiHuang <ye...@yahoo.com.cn>.
Hey Stefan,
Yes, I agree with you. Actually the test is passed in current state.
But it is not passed as we expect and I do not know how to distinguish
this unexpected pass from the expected pass in the future, so I raise a
failure exception and let it fails now.
Maybe I should remove the failure exception and XFail(), and it is
passed now?
Thanks~
HuiHuang
2009-06-12
yellow.flying
发件人： Stefan Sperling
发送时间： 2009-06-11 21:47:50
收件人： HuiHuang
抄送： dev
主题： Re: [PATCH] commit_tests
On Thu, Jun 11, 2009 at 07:40:08PM +0800, HuiHuang wrote:
> Hi,
>
> log message:
>
> [[[
>
> * subversion/tests/cmdline/commit_tests.py
> Add a test case for committing from multiple WC which have no common
> parent path. It is XFail now and will be changed to Pass after we have fix
> issue #2381.
> ]]]
Looks good, expect that you wrote the test to expect the commit
to fail, and then raise a failure exception.
Instead, we could make the test expect the commit to succeed.
For now, the test will fail when trying to commit and raise
an exception at that point.
Then, all we need to change later is to remove the XFail().
Do you agree?
Stefan
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361490
Re: [PATCH] commit_tests
Posted by Stefan Sperling <st...@elego.de>.
On Thu, Jun 11, 2009 at 07:40:08PM +0800, HuiHuang wrote:
> Hi,
>
> log message:
>
> [[[
>
> * subversion/tests/cmdline/commit_tests.py
> Add a test case for committing from multiple WC which have no common
> parent path. It is XFail now and will be changed to Pass after we have fix
> issue #2381.
> ]]]
Looks good, expect that you wrote the test to expect the commit
to fail, and then raise a failure exception.
Instead, we could make the test expect the commit to succeed.
For now, the test will fail when trying to commit and raise
an exception at that point.
Then, all we need to change later is to remove the XFail().
Do you agree?
Stefan
RE: [PATCH] commit_tests
Posted by Bert Huijben <rh...@sharpsvn.net>.
> -----Original Message-----
> From: Branko Čibej [mailto:brane@xbc.nu]
> Sent: donderdag 11 juni 2009 17:54
> To: Bert Huijben
> Cc: 'HuiHuang'; dev@subversion.tigris.org
> Subject: Re: [PATCH] commit_tests
>
> Bert Huijben wrote:
> >
> > Hi,
> >
> >
> >
> > (Please send your messages to this list as plain text,
> >
>
> Funny, why don't you do the same :)
>
> > to allow answering inline. Thanks!).
> >
>
> How's the format prevent that?
My mail clients put a vertical HTML line before the existing text. This way, if I convert the mail to plain html I don't get the common '>' indentations. And if I would just type my response inline (without hand-editing the entire mail) you wouldn't see what I added.
Bert
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361530
Re: [PATCH] commit_tests
Posted by Branko Cibej <br...@xbc.nu>.
Bert Huijben wrote:
>
> Hi,
>
>
>
> (Please send your messages to this list as plain text,
>
Funny, why don't you do the same :)
> to allow answering inline. Thanks!).
>
How's the format prevent that?
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361340
RE: [PATCH] commit_tests
Posted by Bert Huijben <rh...@sharpsvn.net>.
Hi,
(Please send your messages to this list as plain text, to allow answering
inline. Thanks!).
This test fails with the current implementation, and with the future
implementation.
Our tests should either check the current behavior (and just pass now), or
test for the future behavior and fail somewhere before reaching that
(XFail).
In its current state it passes all the way through the raise svntest.Failure
at the end and will fail earlier when the behavior is fixed. This doesn't
help in diagnosing errors/changes, because the error state will never
change: It will always fail.
Bert
From: HuiHuang [mailto:yellow.flying@yahoo.com.cn]
Sent: donderdag 11 juni 2009 13:40
To: Stefan Sperling
Cc: dev
Subject: [PATCH] commit_tests
Hi,
log message:
[[[
* subversion/tests/cmdline/commit_tests.py
Add a test case for committing from multiple WC which have no common
parent path. It is XFail now and will be changed to Pass after we have fix
issue #2381.
]]]
Huihuang
2009-06-11
_____
yellow.flying
_____
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2361266