[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Test suite errors for r-cran-fs 1.5.0+dfsg-1



Hi Jim,

thanks a lot for your fast reply.

On Thu, Sep 03, 2020 at 09:29:52AM -0400, Jim Hester wrote:
> As far as the test-file.R#43 failure one way I could see it failing is if
> your test machine does _not_ have the tibble package installed. The tests
> assume that all Suggested dependencies are installed.

That's a very good hint.  I've added all Suggests with the exception of
"spelling" (since this is not packaged yet) to the environment where the
test is executed.  This leaves me now with:

── 1. Failure: file_access: checks for file read and write ability (@test-access
file_access("foo/bar", "read") not equal to c(`foo/bar` = FALSE).
1 element mismatch

── 2. Failure: file_access: checks for file read and write ability (@test-access
file_access("foo/bar", "write") not equal to c(`foo/bar` = FALSE).
1 element mismatch

── 3. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)  ───────────────
`dir_ls(".", recurse = TRUE)` did not throw an error.

── 4. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)  ───────────────
`dir_ls(fail = FALSE, recurse = TRUE)` did not produce any warnings.

── 5. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)  ───────────────
`dir_ls("foo2", fail = FALSE, recurse = TRUE)` did not produce any warnings.

── 6. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)  ─────────────
`dir_map(".", fun = identity, recurse = TRUE)` did not throw an error.

── 7. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)  ─────────────
`dir_map(fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

── 8. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)  ─────────────
`dir_map("foo2", fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

── 9. Failure: dir_walk: warns if fail == FALSE (@test-list.R#161)  ────────────
`dir_walk(".", fun = identity, recurse = TRUE)` did not throw an error.

── 10. Failure: dir_walk: warns if fail == FALSE (@test-list.R#161)  ───────────
`dir_walk(fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

── 11. Failure: dir_walk: warns if fail == FALSE (@test-list.R#161)  ───────────
`dir_walk("foo2", fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

── 12. Failure: dir_info: warns if fail == FALSE (@test-list.R#192)  ───────────
`dir_info(".", fun = identity, recurse = TRUE)` did not throw an error.

── 13. Failure: dir_info: warns if fail == FALSE (@test-list.R#192)  ───────────
`dir_info(fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

── 14. Failure: dir_info: warns if fail == FALSE (@test-list.R#192)  ───────────
`dir_info("foo2", fail = FALSE, fun = identity, recurse = TRUE)` did not produce any warnings.

══ testthat results  ═══════════════════════════════════════════════════════════
[ OK: 3437 | SKIPPED: 3 | WARNINGS: 0 | FAILED: 14 ]
1. Failure: file_access: checks for file read and write ability (@test-access.R#14) 
2. Failure: file_access: checks for file read and write ability (@test-access.R#15) 
3. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78) 
4. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78) 
5. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78) 
6. Failure: dir_map: warns if fail == FALSE (@test-list.R#111) 
7. Failure: dir_map: warns if fail == FALSE (@test-list.R#111) 
8. Failure: dir_map: warns if fail == FALSE (@test-list.R#111) 
9. Failure: dir_walk: warns if fail == FALSE (@test-list.R#161) 
1. ...
 
> I do not know why the other tests are failing.
> 
> fs 1.5.0 is passing on CRAN's debian based check machines without any test
> failures (
> https://www.r-project.org/nosvn/R.check/r-devel-linux-x86_64-debian-gcc/fs-00check.html),
> so if it is failing in the debian build machine there must be something
> non-standard going on with the environment.

Well, non-standard means we install only those packages that are
explicitly specified as Test-Depends.  As I said I added all packages
listed in Suggests with the exception of spelling.  Do you think
that the missing spelling might be responsible for the issues above?

Or is there possibly another Test-Depends that by default is installed
on CRAN Debian machines but not necessarily in the list of Dependencies?

Kind regards

      Andreas.
 
> On Thu, Sep 3, 2020 at 4:25 AM Andreas Tille <andreas@an3as.eu> wrote:
> 
> > Hi Jim,
> >
> > I tried to upgrade the Debian package of fs.  Unfortunately it has a
> > couple of errors in the test suite:
> >
> >
> > tthat results  ═══════════════════════════════════════════════════════════
> > [ OK: 3436 | SKIPPED: 3 | WARNINGS: 0 | FAILED: 15 ]
> > 1. Failure: file_access: checks for file read and write ability
> > (@test-access.R#14)
> > 2. Failure: file_access: checks for file read and write ability
> > (@test-access.R#15)
> > 3. Failure: file_info: can be subset as a tibble (@test-file.R#43)
> > 4. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)
> > 5. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)
> > 6. Failure: dir_ls: warns if fail == FALSE (@test-list.R#78)
> > 7. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)
> > 8. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)
> > 9. Failure: dir_map: warns if fail == FALSE (@test-list.R#111)
> > 1. ...
> >
> > Error: testthat unit tests failed
> > Execution halted
> >
> >
> > Any idea what might be wrong here?
> >
> > I discussed this on the debian-r mailing list first and got two answers so
> > far:
> >
> >    https://lists.debian.org/debian-r/2020/09/msg00001.html
> >    https://lists.debian.org/debian-r/2020/09/msg00002.html
> >
> > I do not speak R myself so I do not fully understand these answers but
> > may be you can make some sense out of it.
> >
> > Kind regards
> >
> >        Andreas.
> >
> > PS: It would be great if you could answer in public to our list
> >     debian-r@lists.debian.org
> >
> > --
> > http://fam-tille.de
> >

-- 
http://fam-tille.de


Reply to: