Prev: FAQ 4.74 How do I print out or copy a recursive data structure?
Next: Where to get the document for a library function in command line (linux)?
From: Jorge on 20 May 2010 17:44
Yeah Uri and J gave the program a good going-over so I have plenty to
look at there and now I can add your pointers to the list.

You mention one very good point that should have been obvious yet it
completely got past me ... that beinging I am searching every file for
each single pattern where I should be looking at each file only once
for all patterns.

Ihanks again, Jorge


On May 20, 2:30 pm, Jim Gibson <jimsgib...(a)gmail.com> wrote:
> In article
> <6bf31a76-1cd6-48be-8bea-a91b1c6b8...(a)y21g2000vba.googlegroups.com>,
>
>
>
>
>
> Jorge <awks...(a)yahoo.com> wrote:
> > I have a Linux directory full of shell scripts (~100 scripts with
> > average
> > size of ~60 lines) with many of them calling other scripts as
> > depencencies.
>
> > Not being the author of the scripts and with the immediate need to
> > become familiar with
> > the overall scope of this script directory, I needed to create a cross-
> > reference chart
> > that shows what script depends on what other script.
>
> > My Perl script (below) works in terms of doing what I need, however,
> > IMO, it appears to run a bit slow. I timed it at ~3.5 sec. Possibly
> > that is it's max performance but something tells me (from other Perl
> > script experiences) this one just isn't up to speed.
>
> > I would appreciate any ideas or comments.
>
> > Thanks, Jorge
>
> Uri and J. have given you thorough critiques of your posted program. I
> can only add a few additional suggestions.
>
> Concerning speedup, you have implemented an O(N**2) algorithm. You are
> reading each of N files N-1 times, looking for a different script name
> each time you read each file. You should read each file only once, and
> look for all of your script names in each line. You shouldn't exclude
> the script your are reading from this search, since scripts can call
> themselves. I would also not look in comments, since it is very common
> to put script names in comments, and you don't want those.
>
> For how to search for many things at once, see the advice in 'perldoc
> -q many' "How do I efficiently match many regular expressions at once?"
>
> You are concatenating the directory name and file name into a path,
> then taking many steps to strip the directory name from the path to
> retrieve the file name. Don't bother, since you are only working in one
> directory and the $dir variable only contains one value. Add $dir to
> the open statement or a temp variable:
>
>   my $filename = "$dir/$scripts";
>   open( my $f, '<', $filename ) ...
>
> Then you don't need the @paths array at all.
>
> [program snipped]
>
> --
> Jim Gibson- Hide quoted text -
>
> - Show quoted text -

From: Tad McClellan on 20 May 2010 21:11
Jorge <awkster(a)yahoo.com> wrote:
> Yeah Uri and J gave the program a good going-over so I have plenty to
> look at there and now I can add your pointers to the list.

[ snip TOFU. Please stop doing that! ]

TOFU is "Top Over Fullquote Under" also known as "top posting".

You were asked nicely (twice!) to stop doing that years ago,
yet you continue.


I hereby banish you to perpetual invisibility!

*plonk*


--
Tad McClellan
email: perl -le "print scalar reverse qq/moc.liamg\100cm.j.dat/"
The above message is a Usenet post.
I don't recall having given anyone permission to use it on a Web site.
From: bugbear on 21 May 2010 04:01
Jorge wrote:
> I have a Linux directory full of shell scripts (~100 scripts with
> average
> size of ~60 lines) with many of them calling other scripts as
> depencencies.
>
> Not being the author of the scripts and with the immediate need to
> become familiar with
> the overall scope of this script directory, I needed to create a cross-
> reference chart
> that shows what script depends on what other script.
>
> My Perl script (below) works in terms of doing what I need, however,
> IMO, it appears to run a bit slow. I timed it at ~3.5 sec. Possibly
> that is it's max performance but something tells me (from other Perl
> script experiences) this one just isn't up to speed.
>
> I would appreciate any ideas or comments.

This is a run-once script. Any time spent optimising
it (e.g. 10 minutes) is a net loss to you of ten minutes,
bar what ever time your optimisation saves from 3.5 seconds.

10 minutes then.

BugBear
From: Jorge on 21 May 2010 12:05
On May 20, 6:11 pm, Tad McClellan <ta...(a)seesig.invalid> wrote:
> Jorge <awks...(a)yahoo.com> wrote:
> > Yeah Uri and J gave the program a good going-over so I have plenty to
> > look at there and now I can add your pointers to the list.
>
> [ snip TOFU. Please stop doing that! ]
>
> TOFU is "Top Over Fullquote Under" also known as "top posting".
>
> You were asked nicely (twice!) to stop doing that years ago,
> yet you continue.
>
> I hereby banish you to perpetual invisibility!
>
> *plonk*
>
> --
> Tad McClellan
> email: perl -le "print scalar reverse qq/moc.liamg\100cm.j.dat/"
> The above message is a Usenet post.
> I don't recall having given anyone permission to use it on a Web site.

Yes I was warned (and IMO not in a nice way) but it has been 3 years
since I posted and simply had forgotten about that incident. Sometimes
sh*t happens.

I will make every effort not to do it again.
From: Jorge on 21 May 2010 12:10
On May 21, 1:01 am, bugbear <bugbear(a)trim_papermule.co.uk_trim> wrote:
> Jorge wrote:
> > I have a Linux directory full of shell scripts (~100 scripts with
> > average
> > size of ~60 lines) with many of them calling other scripts as
> > depencencies.
>
> > Not being the author of the scripts and with the immediate need to
> > become familiar with
> > the overall scope of this script directory, I needed to create a cross-
> > reference chart
> > that shows what script depends on what other script.
>
> > My Perl script (below) works in terms of doing what I need, however,
> > IMO, it appears to run a bit slow. I timed it at ~3.5 sec. Possibly
> > that is it's max performance but something tells me (from other Perl
> > script experiences) this one just isn't up to speed.
>
> > I would appreciate any ideas or comments.
>
> This is a run-once script. Any time spent optimising
> it (e.g. 10 minutes) is a net loss to you of ten minutes,
> bar what ever time your optimisation saves from 3.5 seconds.
>
> 10 minutes then.
>
>     BugBear- Hide quoted text -
>
> - Show quoted text -

Point taken. I probably should have pointed out that the focus of my
interest in the program execution speed is not for just this one-time
script. Certainly you (and others) are spot on that 3 seconds is of
little consequence. However, for future use, I.E., when the day comes
that I have to deal with data 100x the size of this data I want to be
using the most efficient method(s).

Thanks, Jorge
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: FAQ 4.74 How do I print out or copy a recursive data structure?
Next: Where to get the document for a library function in command line (linux)?