merlyn@iwarp.intel.com (Randal Schwartz) (03/21/90)
Okay, here's my entry for the shortest self-reproducing Perl program.
Any takers for shorter? Any one liners?
print<<'E'x 2,"E\n";
print<<'E'x 2,"E\n";
E
The program must be fed to an unadorned Perl command 'perl' and
produce the output that would make 'perl' look like it has been
replaced with 'cat'. The null program is not a valid submission. (I
claim it first... :-)
$_='5O1v3v5y9)1b7u2q4x1i0e3u2"3S9n5w7s6&7o7h8k1l6k3u3/';s/(.)(.)/pack('C',ord($2)-$1)/eg;print;
--
/=Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095 ==========\
| on contract to Intel's iWarp project, Beaverton, Oregon, USA, Sol III |
| merlyn@iwarp.intel.com ...!any-MX-mailer-like-uunet!iwarp.intel.com!merlyn |
\=Cute Quote: "Welcome to Portland, Oregon, home of the California Raisins!"=/raymond@russet.berkeley.edu (Raymond Chen) (03/21/90)
>Any one liners?
[[ don't put a newline at the end ]]
$s='$s=%c%s%c;printf$s,39,$s,39;';printf$s,39,$s,39;
Unfortunately, it doesn't work even though it should; the output is
$s='$s=%c%sc;printf$s,39,$s,39;';printf$s,39,$s,39;
^ huh?
This happens with
$Header: perly.c,v 3.0.1.4 90/02/28 18:06:41 lwall Locked $
Patch level: 12
on a VAX8700 running ULTRIX.
Was this fixed in one of the more recent patches?
Adjust another curl attacker, [motto for a curling iron repair shop?]
--
raymond@math.berkeley.edu Maintainer of the csip Frequently Asked Questionstneff@bfmny0.UU.NET (Tom Neff) (03/21/90)
Well this works slightly differently: #! /usr/bin/perl system "cat $0"; -- To have a horror of the bourgeois (\( Tom Neff is bourgeois. -- Jules Renard )\) tneff@bfmny0.UU.NET
emv@math.lsa.umich.edu (Edward Vielmetti) (03/21/90)
In article <1990Mar20.215103.23101@agate.berkeley.edu> raymond@russet.berkeley.edu (Raymond Chen) writes: >Any one liners? [[ don't put a newline at the end ]] $s='$s=%c%s%c;printf$s,39,$s,39;';printf$s,39,$s,39; Unfortunately, it doesn't work even though it should; the output is $s='$s=%c%sc;printf$s,39,$s,39;';printf$s,39,$s,39; ^ huh? the spot marked "huh?" has an ascii NUL here, run it through cat -v or od -a to see it. --Ed
merlyn@iwarp.intel.com (Randal Schwartz) (03/21/90)
In article <1990Mar20.215103.23101@agate.berkeley.edu>, raymond@russet (Raymond Chen) writes: | >Any one liners? | | [[ don't put a newline at the end ]] | | $s='$s=%c%s%c;printf$s,39,$s,39;';printf$s,39,$s,39; | | Unfortunately, it doesn't work even though it should; the output is | | $s='$s=%c%sc;printf$s,39,$s,39;';printf$s,39,$s,39; | ^ huh? | | This happens with | | $Header: perly.c,v 3.0.1.4 90/02/28 18:06:41 lwall Locked $ | Patch level: 12 | | on a VAX8700 running ULTRIX. | | Was this fixed in one of the more recent patches? With pl15, it puts out a null in place of that last percent, yup. And, it looks pretty good in other terms, so I presume that's a bug. Larry? eval <<EOF; print "Just another Perl hacker," EOF -- /=Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095 ==========\ | on contract to Intel's iWarp project, Beaverton, Oregon, USA, Sol III | | merlyn@iwarp.intel.com ...!any-MX-mailer-like-uunet!iwarp.intel.com!merlyn | \=Cute Quote: "Welcome to Portland, Oregon, home of the California Raisins!"=/
lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (03/21/90)
In article <1990Mar20.203516.8429@iwarp.intel.com> merlyn@iwarp.intel.com (Randal Schwartz) writes:
: Okay, here's my entry for the shortest self-reproducing Perl program.
: Any takers for shorter? Any one liners?
:
: print<<'E'x 2,"E\n";
: print<<'E'x 2,"E\n";
: E
:
: The program must be fed to an unadorned Perl command 'perl' and
: produce the output that would make 'perl' look like it has been
: replaced with 'cat'. The null program is not a valid submission. (I
: claim it first... :-)
An obvious abridgement is to delete all the E's:
print<<''x 2,"\n";
print<<''x 2,"\n";
Note the blank line at the end. A not-so-obvious improvement is that
the last semicolon is now optional. (For shame! You of all people
should have thought of this one, Randal!)
print<<''x 2,"\n"
print<<''x 2,"\n"
It's too bad you said the thing about cat. Otherwise
die<<''x 2."\n"
die<<''x 2."\n"
would qualifiy.
But, we can shorten the "\n" by two characters:
print<<''x 2,$/
print<<''x 2,$/
I haven't thought about one liners yet.
Larrylwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (03/21/90)
In article <1990Mar20.215103.23101@agate.berkeley.edu> raymond@math.berkeley.edu (Raymond Chen) writes:
: >Any one liners?
:
: [[ don't put a newline at the end ]]
:
: $s='$s=%c%s%c;printf$s,39,$s,39;';printf$s,39,$s,39;
:
: Unfortunately, it doesn't work even though it should; the output is
:
: $s='$s=%c%sc;printf$s,39,$s,39;';printf$s,39,$s,39;
: ^ huh?
:
: This happens with
:
: $Header: perly.c,v 3.0.1.4 90/02/28 18:06:41 lwall Locked $
: Patch level: 12
:
: on a VAX8700 running ULTRIX.
:
: Was this fixed in one of the more recent patches?
Depends on how recent you think patch 16 is. :-)
do_sprintf() was assuming that you wouldn't sprintf a string using itself
as the format, so it temporarily installed nulls in the string to
handle each % interpolation. Obviously a bad assumption...
Life's a beach. And then you dry.
Larrylwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (03/21/90)
In article <7490@jpl-devvax.JPL.NASA.GOV> lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) writes:
:
: print<<''x 2,$/
: print<<''x 2,$/
:
:
: I haven't thought about one liners yet.
Here is a one-liner. There may be shorter one-liners.
$_=q $_=qx;s/x/$"$_$"/;print ;s/x/$"$_$"/;print
Larrymerlyn@iwarp.intel.com (Randal Schwartz) (03/22/90)
In article <7490@jpl-devvax.JPL.NASA.GOV>, lwall@jpl-devvax (Larry Wall) writes: | Note the blank line at the end. A not-so-obvious improvement is that | the last semicolon is now optional. (For shame! You of all people | should have thought of this one, Randal!) I thought that was just for 'eval' (because of my original request). Does that mean the book is off? :-) $_="7072696e7420274a75737420616e6f74686572205065726c206861636b65722c27"; s/../pack('C',hex($&))/eg; eval; -- /=Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095 ==========\ | on contract to Intel's iWarp project, Beaverton, Oregon, USA, Sol III | | merlyn@iwarp.intel.com ...!any-MX-mailer-like-uunet!iwarp.intel.com!merlyn | \=Cute Quote: "Welcome to Portland, Oregon, home of the California Raisins!"=/
tneff@bfmny0.UU.NET (Tom Neff) (03/22/90)
The following ILLEGAL two-liner reproduces itself via descriptor 2: syntax error in file - at line 1, next 2 tokens "syntax error" Execution aborted due to compilation errors. -- "We must never forget that if the war in Vietnam \ $ Tom Neff is lost... the right of free speech will be X tneff@bfmny0.UU.NET extinguished throughout the world." -- RN 10/27/65 $ \ uunet!bfmny0!tneff
lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) (03/22/90)
In article <1990Mar21.163855.13099@iwarp.intel.com> merlyn@iwarp.intel.com (Randal Schwartz) writes: : In article <7490@jpl-devvax.JPL.NASA.GOV>, lwall@jpl-devvax (Larry Wall) writes: : | Note the blank line at the end. A not-so-obvious improvement is that : | the last semicolon is now optional. (For shame! You of all people : | should have thought of this one, Randal!) : : I thought that was just for 'eval' (because of my original request). That was for -e. eval has assumed a semicolon for a long time. But it was easier to assume ; on all perl scripts than just on -e. : Does that mean the book is off? :-) It just means it doesn't have to end with a semicolon. Now if we could just get it to reproduce itself... Actually, I'd settle for it PROducing itself. I'm still slogging through the tutorial... If you folks out there have any examples you think are instructive for novices, we'd sure like to see them. We can't promise to work them in, but we're particularly looking for small scripts that illustrate particular constructs constructively. You could get your name in print... Say, Randal, do you think we should have an appendix of signature scripts? Larry
tkacik@rphroy.UUCP (Tom Tkacik) (03/23/90)
In article <15274@bfmny0.UU.NET> tneff@bfmny0.UU.NET (Tom Neff) writes: >Well this works slightly differently: > #! /usr/bin/perl > system "cat $0"; Let's take this one to the extreme. And we don't even need perl ;-(. Just put the following line in the file junk, and make it executable. #!/bin/cat junk -- Tom Tkacik GM Research Labs, Warren MI 48090 uunet!edsews!rphroy!megatron!tkacik Work Ph: (313)986-1442 "Csh must go. This is non-negotiable!"
jmm@eci386.uucp (John Macdonald) (03/23/90)
Rather than strain my brain, I let the computer find a short
self-reproducing perl "program" (well, it is a program in the
sense that its purpose in life is to be submitted to a Perl
interpreter to cause it to generate output, but it has no
other use). The shell script to generate the program, and
the program it found, are:
---- cut here for genrepro sh script ----
: genrepro - generate a self-reproducing perl program
echo junk >cur.prog
echo trash >cur.out
until cmp -s cur.prog cur.out
do
echo === not yet ===
cat cur.out
mv cur.out cur.prog
perl cur.prog 2>cur.out
done
mv cur.out cur.prog
echo === final version ===
cat cur.prog
---- end of genrepro sh script ----
---- cut here for cur.prog ----
syntax error in file cur.prog at line 1, next 2 tokens "syntax error"
Execution aborted due to compilation errors.
---- end of cur.prog ----
--
Algol 60 was an improvment on most | John Macdonald
of its successors - C.A.R. Hoare | jmm@eci386phillips@cs.ubc.ca (George Phillips) (03/27/90)
In article <7494@jpl-devvax.JPL.NASA.GOV> lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall) writes: > >Here is a one-liner. There may be shorter one-liners. > >$_=q $_=qx;s/x/$"$_$"/;print ;s/x/$"$_$"/;print How about: $_=q print"\$_=q\t$_\t;eval" ;eval The spaces are tabs, no newline at the end. There's also its more printable cousin: $_='print"\$_=\47$_\47;eval"';eval But the 2 digit octal sequences seem kinda bogus. George Phillips phillips@cs.ubc.ca {alberta,uw-beaver,uunet}!ubc-cs!phillips
jhpb@granjon.ATT.COM (Joseph H. Buehler) (03/28/90)
Here's the world's smallest. Quoted so you can see it: "" [granjon] $ perl junk > junk.out [granjon] $ ls -l junk* -rw-r--r-- 1 jhpb arch 0 Mar 27 19:01 junk -rw-r--r-- 1 jhpb arch 0 Mar 27 19:05 junk.out Joe Buehler
nwinton@dalriada.axion.bt.co.uk (Neil Winton) (03/28/90)
Um ... err ... I may have missed this one going by but how about exec 'cat', $0; Nahhh ... too much of a cheat ... but it does fulfill Randal's requirement that `"perl file" should behave exactly as "cat file"' :-) Neil E-Mail (UUCP) NWinton@axion.bt.co.uk (...!uunet!mcsun!axion.bt.co.uk!nwinton) Organisation British Telecom Research Laboratories (RT3134) Snail Mail 306 SSTF, BTRL, Martlesham Heath, IPSWICH IP5 7RE, UK Telephone +44 473 646079 (or +44 473 643210) *** This line intentionally left justified ***
merlyn@iwarp.intel.com (Randal Schwartz) (03/29/90)
In article <JHPB.90Mar27190649@granjon.ATT.COM>, jhpb@granjon (Joseph H. Buehler) writes: | | Here's the world's smallest. Quoted so you can see it: | | "" Go back to my original article. I already DQ'ed that. Nice try. :-) open(STDERR,">&STDOUT");warn "Just another Perl hacker,\n" -- /=Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095 ==========\ | on contract to Intel's iWarp project, Beaverton, Oregon, USA, Sol III | | merlyn@iwarp.intel.com ...!any-MX-mailer-like-uunet!iwarp.intel.com!merlyn | \=Cute Quote: "Welcome to Portland, Oregon, home of the California Raisins!"=/