Bug #105986 for File-Path: bad test count when cloning from github

Mon Jul 20 16:16:09 2015 jkeen [...] verizon.net - Ticket created

Subject:	bad test count when cloning from github
Date:	Mon, 20 Jul 2015 15:15:40 -0500 (CDT)
To:	bug-File-Path [...] rt.cpan.org
From:	Jim Keenan <jkeen [...] verizon.net>

Today I cloned the File-Path repository onto a machine which previously never held this repository. Then I built and tested as usual. I got failures in both 'make test' and 'prove -vb t/Path.t'. The short version: t/Path.t (Wstat: 65280 Tests: 167 Failed: 0) Non-zero exit status: 255 Parse errors: Bad plan. You planned 169 tests but ran 167. See attachment for full output. However, when, on that same machine, I used 'cpanm' to upgrade to File-Path-2.11, I got no test failures: $ cpanm File::Path --> Working on File::Path Fetching http://www.cpan.org/authors/id/R/RI/RICHE/File-Path-2.11.tar.gz ... OK Configuring File-Path-2.11 ... OK Building and testing File-Path-2.11 ... OK Successfully installed File-Path-2.11 (upgraded from 2.09) 1 distribution installed

Download clone_from_github_make_test_failure.txt.gz
application/octet-stream 2.7k

Message body not shown because it is not plain text.

Mon Jul 20 16:23:33 2015 jkeenan [...] cpan.org - Correspondence added

On Mon Jul 20 16:16:09 2015, jkeen@verizon.net wrote: Show quoted text

> Today I cloned the File-Path repository onto a machine which > previously never held this repository. Then I built and tested as > usual. I got failures in both 'make test' and 'prove -vb t/Path.t'. > The short version: > > > t/Path.t (Wstat: 65280 Tests: 167 Failed: 0) > Non-zero exit status: 255 > Parse errors: Bad plan. You planned 169 tests but ran 167. > > > See attachment for full output. > > > However, when, on that same machine, I used 'cpanm' to upgrade to > File-Path-2.11, I got no test failures: > > > $ cpanm File::Path > --> Working on File::Path > Fetching http://www.cpan.org/authors/id/R/RI/RICHE/File-Path- > 2.11.tar.gz ... OK > Configuring File-Path-2.11 ... OK > Building and testing File-Path-2.11 ... OK > Successfully installed File-Path-2.11 (upgraded from 2.09) > 1 distribution installed

The correct results while using 'cpanm' are explained by the fact that that program, as one would expect, goes to CPAN for a tarball. And the tarball for 2.11 holds t/Path.t with a plan of only 159 tests. I manually ran the tests and confirmed that 259 tests ran. So the problem is with HEAD on github.com (096b2fc2c277cdafe352244c5d595d23ed82276a)

Mon Jul 20 16:23:33 2015 The RT System itself - Status changed from 'new' to 'open'

Mon Jul 20 18:11:37 2015 rich [...] richelberger.com - Correspondence added

Subject:	Re: [rt.cpan.org #105986] bad test count when cloning from github
Date:	Mon, 20 Jul 2015 18:11:21 -0400
To:	bug-File-Path [...] rt.cpan.org
From:	Richard Elberger <rich [...] richelberger.com>

Hi Jim, You probably want to check out the 2.11 tag. we have made commits since then. And yes, HEAD is broken right now (that’s why I said I was fed up with the Path.t mess — it’s horrible to keep counts straight with the SKIP matrix of root tests and Test::Output). https://github.com/rpcme/File-Path/releases/tag/2.11 Show quoted text

> On Jul 20, 2015, at 4:16 PM, jkeen@verizon.net via RT <bug-File-Path@rt.cpan.org> wrote: > > Mon Jul 20 16:16:09 2015: Request 105986 was acted upon. > Transaction: Ticket created by jkeen@verizon.net > Queue: File-Path > Subject: bad test count when cloning from github > Broken in: (no value) > Severity: (no value) > Owner: Nobody > Requestors: jkeen@verizon.net > Status: new > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=105986 > > > > Today I cloned the File-Path repository onto a machine which previously never held this repository. Then I built and tested as usual. I got failures in both 'make test' and 'prove -vb t/Path.t'. The short version: > > > t/Path.t (Wstat: 65280 Tests: 167 Failed: 0) > Non-zero exit status: 255 > Parse errors: Bad plan. You planned 169 tests but ran 167. > > > See attachment for full output. > > > However, when, on that same machine, I used 'cpanm' to upgrade to File-Path-2.11, I got no test failures: > > > $ cpanm File::Path > --> Working on File::Path > Fetching http://www.cpan.org/authors/id/R/RI/RICHE/File-Path-2.11.tar.gz ... OK > Configuring File-Path-2.11 ... OK > Building and testing File-Path-2.11 ... OK > Successfully installed File-Path-2.11 (upgraded from 2.09) > 1 distribution installed > > > > <clone_from_github_make_test_failure.txt.gz>

Mon Jul 20 19:04:20 2015 jkeenan [...] cpan.org - Correspondence added

On Mon Jul 20 18:11:37 2015, rich@richelberger.com wrote: Show quoted text

> Hi Jim, > > You probably want to check out the 2.11 tag. we have made commits > since then. And yes, HEAD is broken right now (that’s why I said I > was fed up with the Path.t mess — it’s horrible to keep counts > straight with the SKIP matrix of root tests and Test::Output). > > https://github.com/rpcme/File-Path/releases/tag/2.11 >

At home, where I have Test::Output installed, if I run 'prove -vb t/Path.t' at commit 096b2fc2c277cdafe352244c5d595d23ed82276a in the master branch, I get the following output (excerpt): ##### ok 144 # skip extra scenarios not set up, see eg/setup-extra-tests ok 145 # skip extra scenarios not set up, see eg/setup-extra-tests ok 146 # skip extra scenarios not set up, see eg/setup-extra-tests ok 147 # skip extra scenarios not set up, see eg/setup-extra-tests ok 148 - make_path with final hashref warned due to unrecognized options ok 149 - remove_tree with final hashref failed due to unrecognized options ##### On the machine where I cloned the repo today, I do not (to the best of my knowledge) have Test::Output installed. There I get: ##### ok 144 # skip extra scenarios not set up, see eg/setup-extra-tests ok 145 # skip extra scenarios not set up, see eg/setup-extra-tests ok 146 - make_path with final hashref warned due to unrecognized options ok 147 - remove_tree with final hashref failed due to unrecognized options ok 148 - received a final list of results ok 149 - test base directory gone ##### So I suspect (a) that the difference is due to the presence/absence of Test::Output and (b) that that difference will disappear once we dispose of Test::Output. Tomorrow I will do 'cpanm Test::Output' on the second machine and see what results I get. Thank you very much. Jim Keenan

Tue Jul 21 09:44:45 2015 jkeenan [...] cpan.org - Correspondence added

On Mon Jul 20 19:04:20 2015, JKEENAN wrote: Show quoted text

> On Mon Jul 20 18:11:37 2015, rich@richelberger.com wrote:

> > Hi Jim, > > > > You probably want to check out the 2.11 tag. we have made commits > > since then. And yes, HEAD is broken right now (that’s why I said I > > was fed up with the Path.t mess — it’s horrible to keep counts > > straight with the SKIP matrix of root tests and Test::Output). > > > > https://github.com/rpcme/File-Path/releases/tag/2.11 > >

> > At home, where I have Test::Output installed, if I run 'prove -vb > t/Path.t' at commit 096b2fc2c277cdafe352244c5d595d23ed82276a in the > master branch, I get the following output (excerpt): > > ##### > ok 144 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 145 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 146 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 147 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 148 - make_path with final hashref warned due to unrecognized > options > ok 149 - remove_tree with final hashref failed due to unrecognized > options > ##### > > On the machine where I cloned the repo today, I do not (to the best of > my knowledge) have Test::Output installed. There I get: > > ##### > ok 144 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 145 # skip extra scenarios not set up, see eg/setup-extra-tests > ok 146 - make_path with final hashref warned due to unrecognized > options > ok 147 - remove_tree with final hashref failed due to unrecognized > options > ok 148 - received a final list of results > ok 149 - test base directory gone > ##### > > So I suspect (a) that the difference is due to the presence/absence of > Test::Output and (b) that that difference will disappear once we > dispose of Test::Output. > > Tomorrow I will do 'cpanm Test::Output' on the second machine and see > what results I get. >

I installed Test::Output via 'cpanm' on the machine in question, then built and tested as I usually do. Branch 'master': $ git show |head -1 commit 096b2fc2c277cdafe352244c5d595d23ed82276a The tests ran correctly -- i.e., showing 169 tests running, matching the count specified in the plan for t/Path.t -- via both 'make test' and 'prove -vb t/Path.t'. Thank you very much. Jim Keenan

Tue Jul 21 09:44:48 2015 jkeenan [...] cpan.org - Status changed from 'open' to 'resolved'