-----Original Message----- From: Chris Marshall Sent: Sunday, May 07, 2017 6:39 AM To: pdl-devel@lists.sourceforge.net Subject: [Pdl-devel] Inline failures for PDL-2.017_01...more bitrot
> The PDL "bitrot fix" developers release is testing well except for 3 > reports where there is a failure in t/inline-with.t and > t/inline-comment-test.t: > > http://www.cpantesters.org/cpan/report/9555ca7a-2d05-11e7-b551-93771d927244 > http://www.cpantesters.org/cpan/report/97798df0-2d05-11e7-b551-93771d927244 > http://www.cpantesters.org/cpan/report/99341d22-2d05-11e7-b551-93771d927244 The warning emanates from Inline/SMITH.pm, and the error from JSON.pm (which I don't have iinstalled). I find that if Inline::SMITH is installed then an attempt to load it is made during the running of t/inline-comment-test.t. On my perl-5.25.12: t/inline-comment-test.t ......... Unescaped left brace in regex is illegal here in regex; marked by <-- HERE in m/function(\s+)([a-z0-9_]+)(\s*){{ <-- HERE (.*?)}}/ at C:/_32/blead-5.25.12/site/lib/Inline/SMITH.pm line 63. Compilation failed in require at (eval 21) line 1. t/inline-comment-test.t ......... ok t/inline-with.t ................. ok Note that in 5.25.12 (and hence 5.26) the unescaped left brace is now a fatal error. This means that the module won't load - yet the test still passes for me (which is not all that surprising given that Inline::SMITH is not actually going to be called upon to do anything). I'm inclined to point the finger at JSON.pm since it seems to be the one that throws the error and causes the tests to fail. Why/how is it getting involved ? .... and what's it doing ? Is it responding to the warning emitted by Inline::SMITH, or is that just coincidental ? For sure, it's pretty stupid that an attempt to load Inline::SMITH is being made, and I don't know just which Inline module is responsible for that stupidity. But given the rarity of Inline::SMITH (which hasn't been updated in 15 years) in the wild, and given that its presence is not in itself proven to be a problem for PDL or Inline::C, I'm inclined to think "why bother". Instead, concentrate on why JSON.pm wants to kill the show, and get that issue addressed. Regarding t/inline-with.t, I could see no sign of any attempt to load Inline::SMITH (and I'm not about to install JSON.pm just to try to re-create it ;-) C:\sisyphusion\PDL-2.017_01>perl -Mblib t/inline-with.t ok 1 - use Inline; ok 2 - bind no error # Inline Version: 0.80 ok 3 - compiled # PDL: Float D [5,5,5] ok 4 - dims correct 1..4 BTW, Slaven has already reported the bug in Inline::SMITH 8 months ago (see https://rt.cpan.org/Public/Bug/Display.html?id=117769). No action and no response. Cheers, Rob ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ pdl-devel mailing list pdl-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pdl-devel
