Shell Programming and Scripting

i need a simple text parser which can parse a data file created by a softwre so that i can export it to my mysqldb,, datafile created as one record per line with different number of fields.
e.g datafile contains following.
a=1, b=3, c=4
a=1, c=55, d=abcd
a=5, b=hello, c=99, d=help

now i want to parse them as
1,3,4
1,55,abcd
5,hello,99,help
i want to remove field name and make data as comma seperated so that i can export them to my db. or anyother way to export these fields to mysql where field names are same as a,b,c,d.
a,b,c and d are fields.

Code:

cat datafile | sed 's/[^,]*=//g'

thnx BMDan it realy works. i used it (sed) with awk in bash script instead of cat and got the data i wanted.

one more question in the folder thr are many data files created daily named e.g data20080729_0000. one or two files created daily. i want to use the bash script so that i chk the folder for all files named data***** parse them and at the end move them to other folder. makeing the current folder empty.

If your script is named "my_parser_script.sh":

Code:

for file in $(ls -1d data*); do ./my_parser_script.sh "$file"; mv $file archive/; done

For future questions, please specify the meanings of "parse" and "other folder", though I've made assumptions here.

Lastly, if I may offer a suggestion--train yourself to type carefully and well. If you type things like "thnx", "realy", "thr", "chk", and "makeing", imagine what mistakes like that would translate to on a *NIX command line! Double the amount of time you budget for a reply, type more slowly, and read what you wrote. By taking things a little more slowly on these forums, you'll quickly discover you learn a lot more, too, and you'll sometimes discover that you know the answer to your question already when you've taken the time to carefully describe it. For example, there is another way to accomplish this that I'd bet you could have come up with:

Code:

ls -1d data* | xargs -n 1 ./my_parser_script.sh; mv data* archive/

At any rate, enough lecture. It just saddens me to see people who work in a discipline in which the difference between rm -rf and rm -f is so large being so careless with what's coming out of their keyboard.

Thank you. I appriciate. basically i am not a programmer. I am trying to solve this because of my own ease for things I have to do manualy everyday. But I appriciate your help.
One problem I didnt notices. When using AWK and SED i get the desired results on screen. But when I put ' >> filename ' at the end so that the results goes to a specific file, in that file i get trailing '\' after comma in every field.
e.g
1\,3\,4
1\,55\,abcd
5\,hello\,99\,help

Not seeing what actually you you're doing is making it hard to help...
Help us to help you......

pealse read my first post. I want to change the data file created by a sofware to the simple one. by removing the left side of '=' and keeping the right side. one line per record to a new datafile. All are text files. I am using awk in bash with sed to remove the '=' and left side. but output in screen is ok, but in bash script if i try to use '>> newfilename' it adds '\' (back slash) at the end of each comma seperated field. I dont know may be its blank or empty space which is shown as '\' or something else.