John Porter <[EMAIL PROTECTED]> writes:
> Huh? They need to be "globals", because the "with" mechanism
> is going to have to be able to write them into a symbol table.
> You can't do this with lexicals.
Yes, you can. For an approach like this, the hash needs to be a
pseudo-hash anyway.
-- Joha
Damian Conway wrote:
> So C is going to have to do some pretty freaky magic to work out
> it should call that sub as part of the C. And call it with a
> specifically ordered argument list.
Yes, I never said it would work, just that it looked nicer :)
> However, your suggestion *did* s
> > This seems less of a leap of logic/faith:
> > %record = loadrecord($studentID);
> > with %record {
> > print < > Dear ^name:
> > Your tuition is now due. Please send in a payment of at least
> > ^minumum.
> > SPAM
> > };
"I do not thin' t
Dave Storrs wrote:
> On 17 Aug 2000, Johan Vromans wrote:
> > Dave Storrs <[EMAIL PROTECTED]> writes:
> >
> > > 2) "express" should guarantee that, before it creates a variable
> > > names $FOO, it first calls "local" on any existing $FOO
> >
> > Why, if the variable is lexical (see 3)?
>
> D
This seems good...the HOF stuff seems to have pretty well accepted, and
the thumbtack notation with it.
Dave
On Thu, 17 Aug 2000, Clayton Scott wrote:
> > > On Wed, 16 Aug 2000, David L. Nicol wrote:
> > > %record = loadrecord($studentID);
> > > with %record
Dave Storrs wrote:
>
> On Wed, 16 Aug 2000, David L. Nicol wrote:
>
> > Anyone for generalizing "select" to a more general "with" keyword
> > which would operate on a limited hash and be syntactic sugar for
> > replacing all appearances of unqualified variables that match
> > the hash's keys wit
Dave Storrs <[EMAIL PROTECTED]> writes:
> 2) "express" should guarantee that, before it creates a variable
> names $FOO, it first calls "local" on any existing $FOO
Why, if the variable is lexical (see 3)?
-- Johan
Chaim Frenkel wrote:
>
> NW> P.S. If you're not on -io, this implicitly means you DON'T CARE and are
> NW> willing to accept whatever we come up with. So, everyone that's
> NW> interested please get on -io. Thanks again.
>
> That's a bit strong. All we are doing is filtering the garbage for Larr
> "NW" == Nathan Wiger <[EMAIL PROTECTED]> writes:
NW> P.S. If you're not on -io, this implicitly means you DON'T CARE and are
NW> willing to accept whatever we come up with. So, everyone that's
NW> interested please get on -io. Thanks again.
That's a bit strong. All we are doing is filterin
The poll can't have been exhaustive.
I like these magic variables that depend on currently selected
fie handles, they remind me of Pascal's C construction
for entering the name space of a record structure.
Anyone for generalizing "select" to a more general "with" keyword
which would operate
Chaim Frenkel wrote:
>
> Unless one wants to have a $DEFAULT filehandle and get rid of single
> arg select.
Great minds think alike. :-)
I'm in the process of codifying an RFC that will be titled something
like:
"Replace default filehandle / select with $OUTPUT fileobject"
(chose $OUTPUT b
I'm not sure if you are disagreeing with me or not.
The context was the statment that $STDOUT is the _default_ filehandle.
I was pointing out that by _overriding_ the instantaneous meaning of
$STDOUT to the default fail handle, one would lose the immediate
access to the previous value.
I.e. $STD
On Tue, Aug 15, 2000 at 06:53:30PM -0400, Chaim Frenkel wrote:
> What if you want to print to a default file handle and also to STDOUT?
>
> select(OTHERFH);
> print "This goest to OTHERFH\n";
> print STDOOUT "This went to STDOUT\n";
print $_ "Here I come to save the day!\n" for
> "s" == skud <[EMAIL PROTECTED]> writes:
s> Command line options aren't really a language issue, however I'm not
s> sure that there is a better list for them. Anyone on bootstrap got any
s> good ideas?
shove it over onto -internals.
They will have to implement it anyway.
The only other
Command line options aren't really a language issue, however I'm not
sure that there is a better list for them. Anyone on bootstrap got any
good ideas?
K.
--
Kirrily Robert -- <[EMAIL PROTECTED]> -- http://netizen.com.au/
Open Source development, consulting and solutions
Level 10, 500 Collins
At 12:57 PM 8/15/00 -0700, Nathan Wiger wrote:
>This is a succinct summary of the basic conclusions thus far:
>
>1. a default filehandle IS needed sometimes, but only
> for stuff like print
>
>2. $|, $\, $/, etc will probably go away entirely in
> favor of object methods such a
> "NW" == Nathan Wiger <[EMAIL PROTECTED]> writes:
NW>2. $|, $\, $/, etc will probably go away entirely in
NW> favor of object methods such as $handle->autoflush
It think they will still be needed as lexical variables used as an
initializer for the corresponding per-filehandle valu
> "JSD" == Jonathan Scott Duff <[EMAIL PROTECTED]> writes:
JSD> On Tue, Aug 15, 2000 at 12:57:46PM -0700, Nathan Wiger wrote:
>> This is a succinct summary of the basic conclusions thus far:
>>
>> 1. a default filehandle IS needed sometimes, but only
>> for stuff like print
JSD> Well, I thi
Jonathan Scott Duff wrote:
>
> On Tue, Aug 15, 2000 at 12:57:46PM -0700, Nathan Wiger wrote:
> > This is a succinct summary of the basic conclusions thus far:
> >
> >1. a default filehandle IS needed sometimes, but only
> > for stuff like print
>
> Well, I think that C should always pr
On Tue, Aug 15, 2000 at 03:28:33PM -0500, Jonathan Scott Duff wrote:
> On Tue, Aug 15, 2000 at 12:57:46PM -0700, Nathan Wiger wrote:
> >1. a default filehandle IS needed sometimes, but only
> > for stuff like print
>
> Well, I think that C should always print to $PERL::STDOUT (or
> what
On Tue, Aug 15, 2000 at 12:57:46PM -0700, Nathan Wiger wrote:
> This is a succinct summary of the basic conclusions thus far:
>
>1. a default filehandle IS needed sometimes, but only
> for stuff like print
Well, I think that C should always print to $PERL::STDOUT (or
whatever we call i
Peter Scott wrote:
>
> I don't want to be in the -io discussion; I just want to know the
> conclusions that might affect -language. It seems silly to discuss
> command-line options for setting $| here if there isn't going to be a $|.
Ok, read this thread (4 messages long):
http://www.mail-arch
At 11:35 AM 8/15/00 -0700, Nathan Wiger wrote:
>Peter Scott wrote:
> >
> > Can someone knowledgeable on this issue speak to it?
> >
> > Will $| $/ $\ et al be retired, or be valid for some default filehandle?
> >
> > Will there still be default filehandles?
>
>There's been a big debate on the -io
Peter Scott wrote:
>
> Can someone knowledgeable on this issue speak to it?
>
> Will $| $/ $\ et al be retired, or be valid for some default filehandle?
>
> Will there still be default filehandles?
There's been a big debate on the -io list about this. No clear decisions
yet, but looks like the
At 12:31 PM 8/15/00 -0400, Casey R. Tweten wrote:
>Frankly, they're both ugly. I personally like the *idea*, I'm not
>advocating my solution. It would be wonderful if we could assign values
>to Perl's special variables when we're doing a one line quick
>hack.
With a bit of luck, the special var
On Tue, Aug 15, 2000 at 10:03:55AM -0700, Nathan Wiger wrote:
> Jonathan Scott Duff wrote:
> >
> > Well, now it's my turn to suggest something ;-> How about we give
> > perl the ability to look for a .perlrc file? (Yes, I know the reasons
> > against, but everything is up for grabs now right?
Jonathan Scott Duff wrote:
>
> Well, now it's my turn to suggest something ;-> How about we give
> perl the ability to look for a .perlrc file? (Yes, I know the reasons
> against, but everything is up for grabs now right? :-)
If we do this, it should be off by default. csk/ksh make you turn i
On Tue, Aug 15, 2000 at 12:31:00PM -0400, Casey R. Tweten wrote:
> Frankly, they're both ugly. I personally like the *idea*, I'm not
> advocating my solution. It would be wonderful if we could assign values
> to Perl's special variables when we're doing a one line quick
> hack. Often, I will mo
Today around 11:01am, Jonathan Scott Duff hammered out this masterpiece:
: On Tue, Aug 15, 2000 at 09:14:20AM -0400, Casey R. Tweten wrote:
: > This, by the way (even as a test) was agravating to me because in order to
: > get decent output I really had to do this:
: >
: > cat /etc/passwd | perl
Today around 11:52am, Philip Newton hammered out this masterpiece:
: On Tue, 15 Aug 2000, Casey R. Tweten wrote:
:
: > cat /etc/passwd | perl -nfe 'print((split/:/)[0])'
: >
: > -f is just like $|=1 or, for example, $fh->autoflush(1);
:
: When did -f come into being? Or was this just a propose
Ed Mills wrote:
>
> Sounds like a good idea. I propose a commandline arg that gives info about
> modules in @INC also. I emailed this to the list earlier but it never seemed
> to make it.
>
> Something like
>
> perl -M 'Digest'
>
> might return:
>
> /usr/local/lib/perl5/site_perl/:
>-r-
On Tue, Aug 15, 2000 at 09:14:20AM -0400, Casey R. Tweten wrote:
> This, by the way (even as a test) was agravating to me because in order to
> get decent output I really had to do this:
>
> cat /etc/passwd | perl -nfe '$\="\n";print((split/:/)[0])'
cat /etc/passwd | perl -lnfe 'print((s
On Tue, 15 Aug 2000, Casey R. Tweten wrote:
> cat /etc/passwd | perl -nfe 'print((split/:/)[0])'
>
> -f is just like $|=1 or, for example, $fh->autoflush(1);
When did -f come into being? Or was this just a proposed new switch?
> This, by the way (even as a test) was agravating to me because in
time when i want to edit the previous line only a little
: bit. i tried using ssfe (split screen front end, comes with sirc char
: client) and i saw no output due to pipe buffering. so i had to put a
: $|++ in the one liner to make it work. perl6 should have a command line
: option to enable $| for
4733 Aug 5 1999 Digest.pm
>From: Uri Guttman <[EMAIL PROTECTED]>
>Reply-To: Uri Guttman <[EMAIL PROTECTED]>
>To: [EMAIL PROTECTED]
>Subject: command line option: $|++
>Date: Tue, 15 Aug 2000 00:08:24 -0400
>
>i just found an interseting little oddity i want
using ssfe (split screen front end, comes with sirc char
client) and i saw no output due to pipe buffering. so i had to put a
$|++ in the one liner to make it work. perl6 should have a command line
option to enable $| for STDOUT.
i haven't see any threads about perl's command line options
36 matches
Mail list logo
