| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The attached patch bring the core Perl version code (including a fairly
significant leak when run in a tight loop) up to parity with CPAN
0.9902. This deals with all open issues except:
https://rt.cpan.org/Ticket/Display.html?id=81294
which I am having a hard time modeling.
John
Signed-off-by: Chris 'BinGOs' Williams <chris@bingosnet.co.uk>
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit reverts:
commit a820780c62918236069b3b55502640c99a9e812a
Author: David Golden <dagolden@cpan.org>
AuthorDate: Tue Feb 5 22:23:37 2013 -0500
Commit: David Golden <dagolden@cpan.org>
CommitDate: Tue Feb 5 22:49:36 2013 -0500
Benchmark iteration warnings should warn, not print
because the warnings appearing on STDERR weren't being trapped by
Benchmark.t (see [perl #116831]). Revert for now, and re-address after
5.18.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
DragonflyBSD's file tests don't seem to be chflags-aware:
$ touch foo
$ if [ -w foo ]; then echo "Yes"; fi
Yes
$ chflags uchg foo
$ if [ -w foo ]; then echo "Yes"; fi
Yes
versus FreeBSD:
$ touch foo
$ if [ -w foo ]; then echo "Yes"; fi
Yes
$ chflags uchg foo
$ if [ -w foo ]; then echo "Yes"; fi
$
|
|
|
|
| |
Setting 0100, 0300 and 0500 on files is totally non-reversable
|
|
|
|
|
| |
Using git clean -dXf would remove all files from lib/Net including
tracked files.
|
|
|
|
|
|
|
|
|
|
|
|
| |
arguments"
This reverts commit f041cf0f9c6469c41de8b73d5f7b426710c3ff8b.
This is reverted in reference to [perl #113834]. Although this commit
does fix a real bug, it introduces a new one that is arguably worse.
More importantly, it is a regression from 5.16.0. We should produce a
test case that solves both bugs, then solve them both, then get that in
place. Before 5.18.0 would be nice. After 5.18.0 may have to do.
|
| |
|
|
|
|
|
| |
This only happens should Perl be compiled on the very first Unicode
release, which is extremely unlikely, but fix it anyway.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
This would speed up compilation and execution by a trivial amount, but
makes sure that no regular expression even tries to compile in Unicode,
which could cause recursive invocation of this. Before this construct
was invented, special care had to be taken to keep that from happening;
those measures remain in effect, but this gives added insurance.
|
|
|
|
|
|
|
| |
Deleting a hash slice compiles 5 fewer ops, and executes 21 fewer than
looping over the keys to delete each in turn. Whilst this is arguably a
micro-optimisation, it does not increase obfuscation and is in code loaded
by nearly every Perl program, so feels worthwhile.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
use locale - this will now die if $Config{d_setlocale} is not true.
All tests that use locale will skip if $Config{d_setlocale} is not true.
This enables us to pass tests on Android which uses ICU instead of locales.
The committer removed trailing white space
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Sometimes, dumpvar's dumpvalue routine gets a value it can't dump. The
simplest example to contrive is the one in this test: a tied hash that
can't tell you its keys. Until now, this would cause the whole dump to
abort as soon as it failed to dump one part.
With this commit, each stringify or unwind is inside an eval. Failed
stringifications or unwindings are replaced with a placeholder showing
the error.
unwind uses return to stop early, and rather than go through contortions
to wrap the eval in something that can then return 1 to test that eval
worked, I've just asserted that this code requires 5.14.0, which made $@
a much more reliable indicator of failure after eval.
|
| |
|
|
|
|
|
| |
These were all uncovered by the new Pod::Checker, not yet in core.
Fixing these will speed up debugging the new Checker.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Four recently introduced tests in locale.t fail for two locales of all
the ones that get tested in our smoke farm. I investigated the failures
and it looks to me like the problem in each case is that the locale
definition is defective.
The tests were added because of finding and fixing a bug in Perl, so I
don't want to remove them. Instead these 4 tests will be marked as TODO
if at least 95% of locales pass on any given machine.
This works for our current smokers.
|
|
|
|
|
|
|
|
|
| |
This on-by-default warning will be raised upon use of the
soon-to-be-delivered regex sets feature. The syntax used is currently
invalid, so no feature activation is necessary.
The purpose of this warning is to make sure that any user is properly
notified that this feature is experimental and subject to change
|
|
|
|
|
|
|
| |
A user-defined character name with trailing or multiple spaces in a row
is likely a typo, and hence likely won't match what the other uses of
it. These names also won't work if we extend :loose to these. This
now generates a warning.
|
|
|
|
|
|
|
|
|
| |
The documentation says this is how it should behave, but only 1 of the
three paths in the code did it, and in fact there was a test to the
contrary.
I'm only adding a test for one of the two fixed paths, as the other one
appears to require a weird file name.
|
|
|
|
|
| |
The code says that identifier syntax is required for a file name, but
the code accepted any sequence of \w's.
|
| |
|
| |
|
|
|
|
| |
There's been a release since this change was submitted.
|
|
|
|
| |
Merge-SNAFU.
|
|
|
|
|
| |
This was done by reverting parts of the offending commit from the
git bisecting and adding a test. Thanks to Kevin Dawson for the report.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
Thanks to Father Chrysostomos for fixing the bug in blead, I was now
able to fix the test in the debugger.
|
|
|
|
|
|
|
|
| |
It is caused by $obj spontaneously becoming undef for no good reason -
seems like a bug in the perl interpreter (or elsewhere). This was
noticed by Father C here:
https://rt.perl.org/rt3/Ticket/Display.html?id=114926
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|