Shell Programming and Scripting

I'm working on a personal project, a multiplication quiz script for my kids. In it, the user's performance will be recorded and written to a file. After they've played it a little while, it will start to focus more on the ones that give them the most trouble-- that take a long time to answer or that they more frequently get wrong.

The information for each user will be written out to a file for future reference. The format will be something like:

Code:

factor1,factor2|trial1|trial2|...|trialn

where each trialx will look like:

Code:

YES/NO,answer_given,time_to_answer

YES/NO is simply the answer to the question, "Did the user give the right answer?" So, long explanation over... Here's what I'm wondering: What is the best way to append the most recent results to particular lines? Suppose that the user chooses to do a 3-question test and gets this output:

Code:

3,6|YES,18,1.635
9,10|YES,90,1.838
8,6|NO,47,4.510

Then I want to take "YES,18,1.635" and append it to the "3,6" line of the file. Similarly, "YES,90,1.838" will be appended to the "9,10" line, and so on. What's your suggestion for the best way to do this from within a script? I can load the external file into an array, append, and write out, for example, but is that the best way? How would you do this? Please let me know if the question is unclear. Many thanks in advance.

Assumed your filename is "file".
With sed:

Code:

factor="3,6"
trial="YES,18,1.635"
sed '/'"$factor"'|/ s/$/|'"$trial"'/' < file > file.new

With the shell built-ins it becomes a bit long,
so a function makes sense:

Code:

append () {
IFS="|"   
while read factor trial
do
  if [ "$factor" = "$1" ] 
  then
    echo "$factor|$trial|$2"
  else
    echo "$factor|$trial"
  fi
done < file > file.new
}

and call it

Code:

append "3,6" "YES,18,1.635"

---------- Post updated at 02:47 PM ---------- Previous update was at 02:38 PM ----------

The sed version has the problem that it does not have a true field separator.
Here is an awk version:

Code:

awk -F"|" '{if ($1==factor) {print $0 FS trial} else {print}}' factor="3,6" trial="YES,18,1.635" < file > file.new

Unfortunately, both of these examples only work if all possible values of factor are in file before you try to add a new trial. If you want to start with an empty file and add new entries (as well as update existing entries), you could try these minor updates:

Code:

append () {
IFS="|"
found=0
while read factor trial
do
  if [ "$factor" = "$1" ] 
  then
    echo "$factor|$trial|$2"
    found=1
  else
    echo "$factor|$trial"
  fi
done < file > file.new
if [ $found = 0 ]
then
    echo "$1|$2" >> file.new
fi
}

Code:

awk -F"|" '{if ($1==factor) {print $0 FS trial;found=1} else {print}END{if(! found) print factor FS trial}' factor="3,6" trial="YES,18,1.635" < file > file.new

Thank you for both replies. I apologize for not responding sooner and for bumping a thread so long after it was replied to. A family emergency prevented (or, at least distracted from) replying sooner. The suggestions are now being turned into a working solution.

One question, though... Since the separator of choice, "|", follows both the factors at the beginning of the line and each of the subsequent trial entries, wouldn't only one existent trial be preserved from the file? Suppose there's a line:

8,6|NO,47,6.110|NO,47,3.910

Wouldn't $factor be "8,6" and $trial be "NO,47,6.110", the first trial? It seems like the second would be discarded and no more than 2 trials would ever be preserved. But I could easily be reading wrong. If not, though, perhaps I would be better off with a file structure like:

Code:

factor1,factor2|trial1|trial2|...|trialn

and IFS=":". What think ye? Again, many thanks.

I apologize. In my last posting, the awk program I supplied was missing a closing brace. I should have pasted it in the form I used when I tested it instead of converting it back to a 1-liner. Try the following awk script (but use /usr/xpg4/bin/awk, /usr/xpg6/bin/awk, or nawk instead of awk if you're using a Solaris/SunOS system):

Code:

awk -F"|" '
{       if($1==factor) {
                print $0 FS trial
                found=1
        } else print
}
END{    if(! found) print factor FS trial
}' factor="3,6" trial="YES,18,1.635" < file > file.new

And, no. Changing the field separator from "|" to ":" will not make any real difference. This script will add new trial results to the line in file that has the given factors (or add a new line if the given factors are not already in file); not replace existing trial results. It will never replace existing trial information when adding a new trial. You just need to move or copy file.new to file after each trial is successfully added.