Help with "mv" wildcard command
Kurt M. Hockenbury
: Without writing a script, is there a unix command that I can use to
: change a number of "files.htm" to" files.html"? I have tried "mv *.htm
: *.html" and got an error message about the ending arguement needing to be
: a directory. Thanks!!!
Well, it's a command line, not a script, but... :-) [why are people afraid
of writing shell scripts?]
Under sh/ksh/bash:
for i in *.htm ; do mv $i "$i"l ; done
Under csh/tcsh:
foreach i (*.htm)
mv $i "$i"l
end
Don't forget to update the href's in your files to reflect the new names.
-Kurt
snail://USA/07030/NJ/Hoboken/PO Box 5136/Kurt M. Hockenbury
http://linux.stevens-tech.edu/kmh/kmh-fun.html
John Wingate
: Without writing a script, is there a unix command that I can use to
: change a number of "files.htm" to" files.html"? I have tried "mv *.htm
: *.html" and got an error message about the ending arguement needing to be
: a directory. Thanks!!!
You can use a loop entered by hand at the shell prompt. A ksh example
(which works for bash too) has been posted. A csh version would have
different syntax.
The reason for your error message is that the shell handles the
wild-card expansion; mv, cp, and other commands do not use wild cards
at all. If you have file1.htm, file2.htm, file3.htm and newfile.html
in your directory, the command executed for "mv *.htm *.html" will be
"mv file1.htm file2.htm file3.htm newfile.html", probably not what you
want.
Having the shell do the expansion automatically gives a consistent
semantic interpretation to asterisks wherever they occur. With
"mv *.htm *.html" as you tried it, the "*.htm" refers to existing files
and the "*.html" to (presumably) nonexisting files with names generated
from the others. This is possible with the corresponding DOS and VMS
commands, which don't take an arbitrary list of arguments. Also
handling more than one wild-card under a DOS-like interpretation is
problematic: supposing you had a file named "abracadabra", how would
you want "mv a*b*a a*x*a" to behave?
I prefer the consistency of the unix shell approach, even if it does
make some things a little harder to do.
________________________________________________________________________
John Wingate win...@clark.net jwi...@nswc.navy.mil
James Piwowarczyk
Scott Frohman
James-
'Fraid not. When the shell sees "*.html" in your command, it looks in the
current directory and tries to fill in that wildcard spec with actual files.
Since they don't exist (yet), it will not succeed. But the script isn't too
dangerous! See (ksh):
$ for FILE in *htm
> do
> mv FILE ${FILE%htm}html
> done(ksh):
the ${FILE%htm} stuff says, "take the rightmost part of the variable $FILE,
chopping off `htm'."
You could write a script to do this using something else in place for "htm" &
"html"...
Regards,
Scott Frohman
--
___ ____
/___ /__ sfro...@csci.csc.com GE d++ s++: C+ UH+++
\_____/cott / rohman Computer Sciences Corporation P+ W+ N++ w--- M++
(301) 897-1178 V-- PS+ t+ tv e+++
The opinions expressed here do not necessarily make sense.
Jim Davis
> Without writing a script, is there a unix command that I can use to
> change a number of "files.htm" to" files.html"? I have tried "mv *.htm
> *.html" and got an error message about the ending arguement needing to be
> a directory. Thanks!!!
ls *.htm |xargs -i -t mv {} {}l
Jim
Scott Frohman
'$' before the FILE on the 3rd line. It's corrected below. Sorry.
-S
$ for FILE in *htm
> do
> mv $FILE ${FILE%htm}html
> done(ksh):
^^^^^^
Gokhan Gezmisoglu
--
gokhan
Max Heffler
>Without writing a script, is there a unix command that I can use to
>change a number of "files.htm" to" files.html"? I have tried "mv *.htm
>*.html" and got an error message about the ending arguement needing to be
>a directory. Thanks!!!
Bourne shell:
for i in *.htm
do
mv $i $i{l}
done
C shell:
foreach i (*.htm)
mv $i ${i}l
done
--
Max Heffler, Senior Software Design Engineer home: m...@texsys.com
Home page: http://www.texmicro.com/~max - (an exercise in egomania)
___________________________________________________
/ // /// // /// // /// // /// // /// // /// // ///
---------------------------------------------------
Andrew Fitzgibbon
In article <412kan$k...@daily-planet.execpc.com> j...@earth.execpc.com (James Piwowarczyk) writes:
> Without writing a script, is there a unix command that I can use to
> change a number of "files.htm" to" files.html"? I have tried "mv *.htm
> *.html" and got an error message about the ending arguement needing to be
> a directory. Thanks!!!
Here's a perl script which does this... not well tested but it works in your case.
For safety, it only prints out the moves, you should pipe them through sh
to execute them. Obviously this can be changed.
For ultimate convenience, you need to persuade your shell not to glob using
an alias. In tcsh, that's
alias mva '(set noglob; exec $HOME/bin/share/mva \!*)'
For bash, my best guess is
alias mva='set -o noglob ; exec_and_reset_glob $HOME/bin/share/mva '
exec_and_reset_glob () { $* ; set +o noglob }
A.
-----------------------------------------------------------------------------
#!/usr/local/bin/perl4 -w
# Copyright July 1995
# Andrew Fitzgibbon (andr...@ed.ac.uk)
sub usage {
print '
Usage: mva "<template>" "<template>"
<template>s may contain the following globules:
* converted to regexp (.*)
{a,b,c} converted to regexp (a|b|c)
Beware of ~expansions.
';
exit;
};
open (STDERR, "/dev/null");
print STDERR "MVA: ";
# Check that two args have been specified
&usage if ($#ARGV != 1);
# Divide source and dest into an array of basic blocks.
# The block types are Star, Alternative and Literal.
$src = $ARGV[0];
$dest = $ARGV[1];
$glob1="\\*";
$glob2="{[^}]+}";
$glob="($glob1)|($glob2)";
@src_array = split(/$glob/, $src);
@dest_array = split(/$glob/, $dest);
# Convert the shell glob into a perl regexp, sources as expected,
# destinations to "\$matchindex++"
$re_src = "^";
$re_dest = "";
$matchindex = 0;
for ($aryindex = 0; $aryindex <= $#src_array; $aryindex++ ) {
$cmp = $src_array[$aryindex];
if ($cmp) {
print STDERR "Parsing [$cmp]: ";
if ($cmp eq "*") {
print STDERR "Star\n";
$re_src = $re_src . "(.*)";
$matchindex++;
$re_dest = $re_dest . "\\$matchindex";
} elsif ($cmp =~ $glob2) {
print STDERR "Alternative\n";
$cmp =~ s/[{}]//g; #
$cmp =~ s/,/|/g;
$re_src = $re_src . "($cmp)";
$matchindex++;
$re_dest = $re_dest . "\\$matchindex";
} else {
print STDERR "Literal\n";
$cmp =~ s/(\W)/\\$1/g;
$re_src = $re_src . $cmp;
$re_dest = $re_dest . $dest_array[$aryindex];
}
}
}
$re_src = $re_src . "\$";
print STDERR "$src -> $re_src\n";
print STDERR "Dest = $re_dest\n";
# use csh to glob original pattern as sh doesn't do {} construct
open(FILELIST, "/bin/csh -fc \"echo $src\" | tr -s ' \t\r\f' '\\012\\012\\012\\012'|");
while ($_ = <FILELIST>) {
chop;
print "mv $_\t";
eval "s!$re_src!$re_dest!";
print STDERR "1: $1 2: $2\n";
print "$_\n";
}
-----------------------------------------------------------------------------
--
Andrew Fitzgibbon (Research Associate), andr...@ed.ac.uk
Artificial Intelligence, Edinburgh University. +44 031 650 4504
<a href=http://www.dai.ed.ac.uk/staff/personal_pages/andrewfg> Home Page </a>
"Never say there is no way" -- me.
Roman Fietze
> Without writing a script, is there a unix command that I can use to
> change a number of "files.htm" to" files.html"? I have tried "mv *.htm
> *.html" and got an error message about the ending arguement needing to be
> a directory. Thanks!!!
for file in *.htm; do mv $file ${file}l; done
... and e.g to change *.abc to *.xyz you could do
for file in *.abc; do mv $file ${file%.abc}.xyz; done
This is for sh (???) and ksh.
Roman
--
Roman Fietze (Mail Code 5023) Kodak AG Stuttgart/Germany
fie...@kagcpd01.ag01.kodak.COM fie...@kodak.COM