[cgjoker]

could I do something like
if -mtime +10 <= current time ?

[cgjoker]

i did this but i get the following error...

for file in *
do
if [ -f $file ]
then
flag=$(echo $file | cut -c1-2)
#echo "${flag}"
for f in `find . -type f -mtime ${flag} -name "${flag}${file}"`
do
echo "${flag}"
mv "$f" /directory/"b$f""_$date".dat
done
fi
done

error:
find: 0652-086 Specify a decimal integer for -mtime
Usage: find Path-list [Expression-list]
find: 0652-086 Specify a decimal integer for -mtime
Usage: find Path-list [Expression-list]
find: 0652-086 Specify a decimal integer for -mtime
Usage: find Path-list [Expression-list]

[cgjoker]

updated the flag value to make it an expression, now i dont get any errors but the files that are older than the flag and where the name is the same as the flag, are still not moving.

for file in *
do
if [ -f $file ]
then
flag=`expr $(echo $file | cut -c1-2)`
#echo "${flag}"
for f in `find . -type f -mtime ${flag} -name "${flag}${file}"`
do
echo "${flag}"
mv "$f" /directory/"b$f""_$date".dat
done
fi
done

[cgjoker]

im at a total loss.... it looks like the `find... ` command doesn't recognize predetermined variables. if it did i think this code would work fine...


for file in *
do
if [ -f $file ]
then
flag=`expr $(echo $file | cut -c1-2)`
echo "${flag}"
for f in `find . -type f -mtime +${flag} -name "${flag}${file}"`
do
echo "${flag}"
mv "$f" /directory/"b$f""_$date".dat
done
fi
done

[AGCurry]

Originally Posted by cgjoker

im at a total loss.... it looks like the `find... ` command doesn't recognize predetermined variables. if it did i think this code would work fine...


for file in *
do
if [ -f $file ]
then
flag=`expr $(echo $file | cut -c1-2)`
echo "${flag}"
for f in `find . -type f -mtime +${flag} -name "${flag}${file}"`
do
echo "${flag}"
mv "$f" /directory/"b$f""_$date".dat
done
fi
done

Well, your find can't find the file because it's looking for the wrong name.
${flag}${file} - for the file "10farkel.blah" - would be "1010farkel.blah".

find . -name "$file" -mtime +${flag} -exec mv {} /directory \;

[cgjoker]

I didn't even catch that... thanks.

Its working great now.. after all that these few lines does just what I need....

for file in `ls *`
do
flag=`expr $(echo $file | cut -c1-2)`
find . -name "$file" -mtime +${flag} -exec mv {} /directory/"b""$file".dat \;
done

less the removing of the 10 flag but I can do that.

Thanks so much for your help, you guys are great !
c.

[AGCurry]

Congratulations.

A couple of suggestions:

1. All you need in the first line is "for file in *". No need to load and run ls.

2. The line: flag=`expr $(echo $file | cut -c1-2)` bothers me. One reason is that the expr command should not be needed. The other is that IF there is a file which does not begin with numeric digits, find will fail. A test that $flag is numeric should probably be incorporated, e.g.,

if let $flag=$flag+0
then
find ...

Good luck!
Andy

[cgjoker]

I see your point about the test, however, if there is a file with non numeric values it would just be ignored and the others would still get processed would it not?

by the way, im trying to add something where if the file has a numeric value it does one thing, if it has a two numerics it does another.. im can't seem to get it.. can you help please?

if ( $file =~ /[^0-9][a-zA-Z]*/ )
then
newfile=`echo $file | cut -b2- | sed -e 's/.dat//g'`
else
newfile=`echo $file | cut -b3- | sed -e 's/.dat//g'`
fi

[AGCurry]

Originally Posted by cgjoker

I see your point about the test, however, if there is a file with non numeric values it would just be ignored and the others would still get processed would it not?

by the way, im trying to add something where if the file has a numeric value it does one thing, if it has a two numerics it does another.. im can't seem to get it.. can you help please?

if ( $file =~ /[^0-9][a-zA-Z]*/ )
then
newfile=`echo $file | cut -b2- | sed -e 's/.dat//g'`
else
newfile=`echo $file | cut -b3- | sed -e 's/.dat//g'`
fi

I often find it easiest to use "case" for this kind of thing:

case "$file" in
[0-9][0-9]*) do one thing ;;
[0-9]*) do another thing ;;
*) do something else ;;
esac

[cgjoker]

thanks again... c.

[cgjoker]

So im trying to figure out how I can use the case loop for this? So the idea again was that I want to be able to get either files that have 2 digits or 1 digit in the file name.
In the code below, it is only going after 2 digits.

cd /olddirectory

for file in `find . -name "00*.*"`
do
if [ $file ]
then
fnow=`echo $file | cut -b4- | sed -e 's/.dat//g'`
mv "$file" /directory/"$fnow""_$date".dat
done

for ffile in *
do
if [ $ffile ]
then
fflag=`echo $ffile | cut -c2`
newffile=`echo $ffile | cut -b3- | sed -e 's/.dat//g'`
find . -name "$ffile" -mtime +${fflag} -exec mv {} /directory/"$newffile""_$date".dat \;

fi

[cgjoker]

i have an idea ill try and let you know.

[cgjoker]

my idea didn't work... ill have to rethink this approach...

for fnow in `find . -name "0*.*"`
do
if [ $fnow ]
then
case "$fnow" in
[0-9][aA-zZ]*) now=`echo $fnow | cut -b3- | sed -e 's/.dat//g'`;;
[0-9][0-9]*) now=`echo $fnow | cut -b4- | sed -e 's/.dat//g'`;;
*) echo "$dt: no files found, continuing processing" ;;
esac
mv "$fnow" /directory/"b$fnow""_$date".dat

done

for ffile in *
do
if [ $ffile ]
then
fflag=`echo $ffile | cut -c1-2`
case "$ffile" in
[0-9][aA-zZ]*) newffile=`echo $ffile | cut -b3- | sed -e 's/.dat//g'`;;
[0-9][0-9]*) newffile=`echo $ffile | cut -b4- | sed -e 's/.dat//g'`;;
*) echo "$dt: no files found, continuing processing"
;;
esac
find . -name "$fblfile" -mtime +${fflag} -exec mv {} /directory/"b$newfile""_$date".dat \;
echo "$dt: ${ffile} file found equal to 1
done

[AGCurry]

Originally Posted by cgjoker

my idea didn't work... ill have to rethink this approach...

for fnow in `find . -name "0*.*"`
do
if [ $fnow ]
then
case "$fnow" in
[0-9][aA-zZ]*) now=`echo $fnow | cut -b3- | sed -e 's/.dat//g'`;;
[0-9][0-9]*) now=`echo $fnow | cut -b4- | sed -e 's/.dat//g'`;;
*) echo "$dt: no files found, continuing processing" ;;
esac
mv "$fnow" /directory/"b$fnow""_$date".dat

done

problems:
find command will find files beginning with 0 only.
if [ $fnow ] should be if [ -f $fnow ] ( you want only normal files ).
case is built to recognize file names that find will never find.
cut -b3 - why do you want the 3rd byte? aren't you looking for the first and/or second bytes?

for ffile in *
do
if [ $ffile ]
then
fflag=`echo $ffile | cut -c1-2`
case "$ffile" in
[0-9][aA-zZ]*) newffile=`echo $ffile | cut -b3- | sed -e 's/.dat//g'`;;
[0-9][0-9]*) newffile=`echo $ffile | cut -b4- | sed -e 's/.dat//g'`;;
*) echo "$dt: no files found, continuing processing"
;;
esac
find . -name "$fblfile" -mtime +${fflag} -exec mv {} /directory/"b$newfile""_$date".dat \;
echo "$dt: ${ffile} file found equal to 1
done

Better. But...
if [ $ffile ] doesn't make sense (same reason as above).
fflag=`echo $ffile | cur -c1-2` gets the first two characters. Your case, again, isn't really set up to match the possibilities correctly.

Here's a shot:

for file in *
do
if [ -f $file ]
then
case "$file" in
[0-9][A-z]*) agedays=$(echo $file | cut -c1) ;; # one digit
[0-9][0-9]*) agedays=$(echo $file | cut -c1-2) ;; # two digits
*) agedays=999 ;;
esac

if [ $agedays -ne 999 -a $agedays -gt 0 ]
then
do whatever you need to do
fi
fi
done


case "$file"