On 1/8/2016 12:49 PM, Andy Bradford wrote:
Thus said Ross Berteig on Thu, 07 Jan 2016 23:55:12 -0800:
Exactly. And I think it is a good idea for the test suite to run from
a folder outside of all checkouts for a lot of reasons.
Perhaps we should update the test: make target to create a temporary
directory, cd to it and run tests from there...
Probably a good idea, but it might need some thought about where it
should do that. Although I suppose the answer $TEMP or $TMP would be
traditional, and is almost certainly not pointing inside the tree. Or
the Makefile could have a variable that defaults to $TEMP but can be
overridden if needed.
Does that work for you (anyone?) right now from trunk?
Depends on what you mean by ``work'' I suppose, but it does for me.
There are some failures:
Perhaps I should have been clearer by what I meant by "work". The test
suite does work for me in the sense that the tests run and report results.
But I have 38 tests reporting failure. IMHO, a regression test suite
should 100% pass for released software, and I intended "work" in that
sense. The tests don't all pass, but they are running.
***** Final result: 38 errors out of 27800 tests
***** Failures: ....
Now I'm a lot more comfortable that I don't have a systematic issue. I
see precisely the same 38 failures here.
And that fails for me because a passel of cases in file1.test and
th1.test appear to require an open checkout, but don't provide one by
calling repo_init.
If they require repo_init and aren't calling it, then I guess they need
to be updated.
I will try to see if the obvious fixes in file1.test and th1.test do the
trick for most of those.
Someone who already deeply understands the three-way-merge and its test
cases in merge2.test needs to look closely at what is reported as
merge-utf-27-23 and merge-utf-27-32. That would be the 27th pass of
random edits to text based on the content of test/utf.test, and the
results are puzzlingly wrong.
--
Ross Berteig r...@cheshireeng.com
Cheshire Engineering Corp. http://www.CheshireEng.com/
_______________________________________________
fossil-dev mailing list
fossil-dev@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/fossil-dev
