Shell Programming and Scripting

Hello everyone,

I am manipulating a large CSV file and am trying to read it into a program and started running into trouble. The have manually edited the file trying to make it correctly run through the program and have made progress. However, I am know stuck with an issue involving too many commas in lines. The correct number of commas in each line should be two but some lines have more than that. Here is an example of a clean line:

Code:

4343434,MyName,I went to a doctor

A messy line might look like this:

Code:

3232323,SomeName,I bought a car for 9007,and,9222

Notice the commas at the front are correct. So is there any way to remove any commas that appear in a line after the second occurrence? I was thinking about using awk or sed for this task but am not entirely sure how to go about this? Any help to resolve this problem would be greatly appreciated!

Code:

echo "3232323,SomeName,I bought a car for 9007,and,9222" | sed ':begin;s/,/ /3;t begin'
3232323,SomeName,I bought a car for 9007 and 9222

Try:

Code:

perl -a -F"," -ple '$_="$F[0],$F[1]," . join " ",@F[2..$#F]' file

Actually this thread can now be disregarded. I just found a solution to my problem using this regular expression and notepad++. I went and reverted my previous edits and started fresh with the file.

Code:

^([0-9]{4,9}+\s)([#_0-9A-Za-z]+(\s))

I then replace with:

$1,$2,

What I was trying to do is solved.

---------- Post updated at 02:44 PM ---------- Previous update was at 02:43 PM ----------

I also thank you guys for your replies. They are educational for me and now I know a little bit more about scripting. I certainly appreciate it.

Moderator's Comments:

edit by bakunin: you are welcome. Changed your threads title accordingly.