[SOLVED] bash script- to read file name with two digits

smart1seo · 06-20-2011, 10:18 PM

I have many files with name in order of number.
e.g)
u0101.asc
u0102.asc
u0103.asc
.
.
u0110.asc
u0111.asc
.
.

I am trying to read file using for loop.

for ((date=01; date<=31; date++))
do
echo ${date}
done

but '01' is read(print) as '1'
How can I make it read from '1' to '01'??

grail · 06-20-2011, 11:33 PM

Well you would need to use printf to get the leading 0 to appear in output.
However, if the question is how to cycle through your files with the for loop:

Code:

for x in u*.asc
do
    do_your_stuff_here
done

smart1seo · 06-21-2011, 12:02 PM

Quote:

Originally Posted by grail

Well you would need to use printf to get the leading 0 to appear in output.
However, if the question is how to cycle through your files with the for loop:

Code:

for x in u*.asc
do
    do_your_stuff_here
done

I could't quite done what I really need to do. If you can help me....

I need to remove text in the file(First 5 lines) for each existing file and save as same name.
The name of files are in order of date and hour. e.g) u0101.asc, u0102.asc... u0110.asc ..u0212.asc...u3123.asc...
So, I used 'tail' command ( I need last 94 lines)

for ((date=01; date<=31; date++))
for ((hour=00; hour<=23; hour++))
do
tail -94 directory/u${date}${hour}.asc > directory/u${date}${hour}.asc
done

But the variables ${date}and${hour} are recognized as one digit '1' instead of '01'.
Please help me to solve this case.

ssrameez · 06-21-2011, 12:16 PM

for i in `ls u01*`
do
j=`cat $i|wc -l`
k=$j-5
cat $i|tail $k>temp
mv temp $i
done

Not Tested.. Just wrote it on the fly.. Please test it and modify accordingly.
variable j will have the number of lines in the file.
variable k will have the number deducted by 5(or whatever you need).
temp file will have temp content.
then move it to the actual name.

colucix · 06-21-2011, 12:32 PM

As previously mentioned use the printf command to retrieve leading zeros:

Code:

for ((date=1; date<=31; date++))
do
  for ((hour=0; hour<=23; hour++))
  do
    echo u$(printf %02d $date)$(printf %02d $hour).asc
  done
done

Newer versions of bash (version 4) can do this by means of the enhanced extended brace expansion, that now accepts zero-padding:

Code:

for date in {01..31}
do
  for hour in {00..23}
  do
    echo u$date$hour.asc
  done
done

chrism01 · 06-21-2011, 06:15 PM

As with Grail on post #2, if you're just cycling through the files, you don't need to know what the nums mean, just process each file.
Re post #4; bash arithmetic http://tldp.org/LDP/abs/html/arithexp.html

In fact, you should bookmark these
http://rute.2038bug.com/index.html.gz
http://tldp.org/LDP/Bash-Beginners-G...tml/index.html
http://www.tldp.org/LDP/abs/html/

grail · 06-22-2011, 12:40 AM

I am not even sure you need the loop structure at all:

Code:

sed -i '1,5 d' u01*