Shell Programming and Scripting

hi all
this is the part i am facing a problem
eg data: filename : tr1

+ T 40

this is a sample record in that file ... the value of T can be anything, but will be a single character.

i need to cut from field two, and i am using this command
cut -d " " -f2 tr1 >tr3
and the o/p is
40

if i try to replace the spaces between + and T using sed

sed 's/\t/ /g' tr1>tr3

o/p :
+ T 40

it is not a tab.....

so i tried to replace it like this .....

tr -s " " "" tr1>tr3

o/p :

+ T 40 ...

it is not reading it as a space ....
more wierd is this ..
i tried to cut the file column wise ...


> cut -c 1 test1
+
> cut -c 2 test1

> cut -c 3 test1
T

i copied the record and pasted it in a notepad and i found that it is tab.

i cannot find out whether it is a space or a tab ....

can any one give me a solution ...

i am trying to do the following but not sure how to do ...
i want to replace everything that is there before tht "T" or "F" with a space... is it possible...?

First of all, i am sorry to say that i didnt understood ur problem , but whatever i understood

cat test
+ T 40

cat test | tr -d " " | sed -e 's/^./ /g;s/\( \)\(.\)/\1\2 /'
T 40

You can use something like the following

Code:

>cat test | od -An -t dC -w10

This will display the ASCII codes for your input file test.
That will tell you what is in your file. A space would be code 32, while a tab would be a 9. Carriage return and line feed are 13 and 10 respectively.

Take a look at this:

Code:

> echo "+ T 40" | od -An -t dC -w10
   43   32   84   32   52   48   10

This shows the two 32 characters and that they are spaces.

Aside: Both of your commands to remove spaces or tabs do not work. This is why you were misled.
Try:
To remove spaces:
tr -d " "
To remove tabs:
TAB=`echo "\0011"`
tr -d "${TAB}"

Notwithstanding "joeyg" good advice. To get a feel for whether white space is spaces or tabs you can substitute them both to make them visible.
TAB=`echo "\0011"`
cat filename|sed -e "s/${TAB}/TAB/g"|sed -e "s/ /SPACE/g"|pg
(Shell purists may cringe at this inefficient pipeline but it is only an example).

In your case the original problem is with using "cut" which requires a fixed delimiter.
If your use "awk" to get the second field it won't matter whether its a space or a tab.
cat filename|awk '{print $2}'

This code should be like this

cat test |sed -e 's/^.*+//g'|tr -d "\t"

It will take care any number of "tabs" or "spaces" present before the "+" sign. No matter its a "space or tabs". It will give the desired output. First give this a try at least

guys please don't use cat command when you are going to use commands like awk,grep and sed because it can read the file
regards

thanks to everyone it worked