tony@oha.UUCP (Tony Olekshy) (06/29/90)
When I run each of the following through perl 3.0.1.5 patch level 18:
print "Test 1\n"; eval 'die "me"'; print "> $@";
print "Test 2\n"; eval 'die ""'; print "> $@";
print "Test 3\n"; eval 'die'; print "> $@";
I get the following output (and I say there should be a > from each case):
Test 1
> me at (eval) line 1.
Test 2
Test 3
Am I: 1) missing something, 2) misunderstanding something, or 3) triggering
a bug?
--
Yours etc., Tony Olekshy. Internet: tony%oha@CS.UAlberta.CA
BITNET: tony%oha@UALTAMTS.BITNET
uucp: alberta!oha!tonylwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (06/29/90)
In article <438@oha.UUCP> tony@oha.UUCP (Tony Olekshy) writes:
: When I run each of the following through perl 3.0.1.5 patch level 18:
:
: print "Test 1\n"; eval 'die "me"'; print "> $@";
: print "Test 2\n"; eval 'die ""'; print "> $@";
: print "Test 3\n"; eval 'die'; print "> $@";
:
: I get the following output (and I say there should be a > from each case):
:
: Test 1
: > me at (eval) line 1.
: Test 2
: Test 3
:
: Am I: 1) missing something, 2) misunderstanding something, or 3) triggering
: a bug?
Mmm, let's just say the "die" routine was coded by a stupid programmer
before eval even existed, and he didn't anticipate that die should always
be trapped by eval. The programmer has been severely reprimanded and promises
never to do it again. The problem will be fixed in the next patch. If you
give a null argument to die it will now be as if you said
die "Died";
And an enclosing eval will trap it.
Larrylogan@rockville.dg.com (James L. Logan) (07/04/90)
In article <8537@jpl-devvax.JPL.NASA.GOV> lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) writes: # # Mmm, let's just say the "die" routine was coded by a stupid programmer # before eval even existed, and he didn't anticipate that die should always # be trapped by eval. The programmer has been severely reprimanded and promises # never to do it again. The problem will be fixed in the next patch. If you # give a null argument to die it will now be as if you said I have a question along the same lines: why is it that "die;" by itself does not print anything? When I throw together a quick, one-time hack, I don't need the verbosity of die "$MyName: cannot do whatever: $!"; and I don't want the script to die without saying anything at all. I just want: script died at - line 20 Is there any reason this shouldn't be the default? -Jim -- James Logan UUCP: uunet!inpnms!logan Data General Telecommunications Inet: logan@rockville.dg.com 2098 Gaither Road Phone: (301) 590-3198 Rockville, MD 20850
lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (07/04/90)
In article <603@inpnms.ROCKVILLE.DG.COM> logan@rockville.dg.com (James L. Logan) writes: : In article <8537@jpl-devvax.JPL.NASA.GOV> : lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) writes: : # : # Mmm, let's just say the "die" routine was coded by a stupid programmer : # before eval even existed, and he didn't anticipate that die should always : # be trapped by eval. The programmer has been severely reprimanded and promises : # never to do it again. The problem will be fixed in the next patch. If you : # give a null argument to die it will now be as if you said : : I have a question along the same lines: why is it that "die;" by : itself does not print anything? When I throw together a quick, : one-time hack, I don't need the verbosity of : : die "$MyName: cannot do whatever: $!"; : : and I don't want the script to die without saying anything at : all. I just want: : : script died at - line 20 : : Is there any reason this shouldn't be the default? That's more or less what I just said was going to happen. Larry