UNIX for Dummies Questions & Answers

Hello,

I am relatively new to UNIX scripting and am learning a lot. I have already tried several searches on this website and have tried various syntax options suggested to no avail. I am obviously not writing the script correctly. I really do appreciate any and all the help.

Below is an outline of what I am looking to accomplish. The text in blue highlights the sections I am struggling with.

GOALS:
IF
The file, /test/me/test.csv is less than 295 bytes
THEN
Do commands here (already have this part written successfully)

ELIF
The file, /test/me/test.csv is >= 295 bytes AND if the FIRST row in that file contains the header string that begins with, ie. Tick_Num,Line_No,*
THEN
Do commands here (already have this part written successfully)

ELIF
The file /test/me/test.csv = 0 bytes
THEN
Do commands here (already have this part written successfully)
FI

Again, thank you very much.

D

In your elif, how can a file have no size and still have a header row? Perhaps there's a typo?

The elif-part doesn't make sense, because when the file has size 0 it can't have any content to be found by grep, as !s means the opposite, ie. <=0.

If you want to issue a command inside a test, you coud try like this:

Code:

if [[ -s /etc/passwd && $(head -1 /etc/passwd| grep -q root > /dev/null 2>&1; echo $?) == 0 ]]; then
    echo yo
else
    echo no
fi

... or as a separate sub if/then/fi below the check if the file exists and has a size >0.

To grep for more than one pattern as a logic "OR", you can use this:

Code:

grep -E "^Tick_Num|^Line_No" infile

The ^ tells it that it is the start of a line, ie. 1st character to follow and the pipe | separates the patterns.
A * would be writtens as .* in Regular Expressions and this would make no sense, as this could be everything and sowith always be true.

Another hint is to declare a variable, that contains the file name, since it is awful to type it over and over and when you want to change it, you change it at just 1 place.

Edit:
If "Tick_Num,Line_No,*" is a complete pattern, forget the lines above about grep -E

First of all, it would be helpful if you posted the error message too.
Second, you can get additional debugging output if you insert the line set -x near the start of your script.
Third, there's an unbalanced quote in your if:

Code:

if [ -s /test/me/test.csv && grep Tick_Num,Line_No,*' /test/me/test.csv ]

Here's one way, that while may not be the slickest, breaks down the issue into logical parts that you are testing for and may be useful for understanding:

Code:

#!/bin/ksh

FILE=test.csv  # Path to file to test.

if [[ -s $FILE ]]; then  # If it exists and has size...
   ##  See if the 1st line starts with the header string
   ##  Throw away the output and test for the return code
   head -1 $FILE | grep '^Tick_Num,Line_No' >/dev/null 2>&1
   if (( $? == 0 )); then
     print "$0: Found a header in the first row"
   else
     print "$0: $FILE does not have a header row"
     exit 1
   fi
else
  print "$0: $FILE does not exist or has no size"
  exit 2
fi

exit 0

If requirements changed, then post this as a reply please. Don't expect people to check your original post all the time to see if something changed, thanks.

Sorry folks - I did have a bunch of typos in my original posting. I'm really sorry about that. I will try the suggestions provided so far and let y'all know!

Thanks,
D

---------- Post updated at 10:19 AM ---------- Previous update was at 09:52 AM ----------

I think since my requirements drastically changed, a lot of the suggestions may not apply...If anyone can forgive my ignorance and update or provide other solutions, I would appreciate it.

Again, apologies.

D