hi guys. I have written the following shell script..
What the script does is it traverses a directory & compares files in that directory..
it traverses a file and if that has a include line it cuts out the name present in the include <> and tries to find a file of that name in the same directory..if the names are garbled then it rectifies it...

however when i execute it, it does work good for initial files and then hangs up..as in it doesn;t traverse the whole directory..which is wierd

here is the script.

#!/bin/sh
# Script for checking the CaSe of filenames in include statements

#rootdir is where all the child files reside
rootdir=/vobs/wsCore/mibs

child_file_count=0

# The for loop for going through the entire directory but checking only those files with mib in their names

for file in `ls $rootdir | grep -i "mib"`
do

# parse all include statements in the next child file by a case-insensitive search
# cut relevant fields remove the quotes redirect the output to a temporary file called test

grep -i "<include>:" $rootdir\/$file | cut -d' ' -f4- | sed 's/\"//g' > test

#Also store this in a variable for future filename comparison with parent files
var_child_i=`grep -i "<include>:" $rootdir\/$file | cut -d' ' -f4- | sed 's/\"//g'`

#now we deal with the temporary file i.e. test file

for i in `cat test`
do
ls $rootdir > list
# A case-sensitive search on list of files in root directory to check if filenames match
# capture parent filename
var_par_s=`grep "$i" list`
echo $var_par_s

#perform a case-insensitive search on the list of files in root directory to capture
#anyhow the parent filename for future comparisons
var_par_i=`grep -i "$i" list`
echo $var_par_i

#check if var_par_s is empty, if empty, no file with that name exists in mib folder
# i.e. case is wrong in original file

if [ -n "$var_par_s" ]
then
echo Parent file found.. Modification of child file $file NOT needed
else
echo Parent file NOT found.
echo Hence we need to modify the child file \= $file containing the erroroneous name \"$var_child_i\"

#check OUT the child file
cleartool checkout -unreserved -comment "For checking include statement" $rootdir\/$file
echo File $file Checked OUT of VOB

#Modify the include statement in the child file to reflect the parents name
echo Now every instance of \"$var_child_i\" will be replaced by \"$var_par_i\"
sed "s/$var_child_i/$var_par_i/g" $rootdir\/$file > aliasfile
mv aliasfile $rootdir\/$file
echo File $file\'s modification successful

#Check IN the child file
#cleartool checkin $rootdir\/$file
echo File $file checked back IN the VOB
fi

child_file_count=`expr $child_file_count + 1`

done

done

echo The total number of child files traversed \= $child_file_count
echo Program execution ends\!

Also...is there someway i can check if the traversing with the for loop reaches the end of the directory ?? or does the for loop take care of that itself??

I'd just keep a list of files that have already been processed and another list of files to be processed.

Proc A) You could have a little routine that will add a file to the "to do" list only if not in the "done" list. This would be easy to test separately.

Proc B) Next you'd want something that takes a file and extracts all the new files from it. Again easy to test.

Next you'd set up a loop that pulls a file from the "to do" list. You don't want to do foreach file in `cat to_do_list` because you'll be adding new files to the list as you go along.

Call proc B) to get all the files accessed by that file and then use proc A) to add them to the "to do" list if not already in there.

Once a file has been loaded then it goes into the "done" list.

Unsure if that made sense but at least it's easier to read than unformatted shell script written by someone else

Mike

Thanks man..
your idea looks great...I think that will work..
In case i run into problems, i ll get back..