Flat file to csv conversion

Shell Programming and Scripting
1

Hi Guy's can someone help me in converting the following

I have a flat text file which has several thousand lines which I need to convert to a csv it's got a consistent format but basically want every time it hit's txt to create a new line with the subsequent lines comma delimited for example

txt.1,line 1, line 2, line n
txt.2,line 1, line 2, line n
etc

The input file looks like this

txt.1
line 1
line 2
line n
txt.2
line 1
line 2
line n
txt.n
line n

Thanks All

2 
>cat p1.txt
txt.1
line 1
line 2
line n
txt.2
line 1
line 2
line n
txt.n
line n

>sed 's/txt/~txt/g' <p1.txt | tr "\n" "," | tr "~" "\n" | sed 's/,$//'

txt.1 ,line 1 ,line 2 ,line n
txt.2 ,line 1 ,line 2 ,line n
txt.n ,line n
3

Perfect - thank you very much!!!!

4

Or with awk:

awk '/txt/{if(s){print s}{s=$0;next}}{s=s "," $0}' file
5

One last question - my fault for not being precise - what if the file name varies the other way ie

1.txt
2.txt

this produces slightly different result for sed, and awk is only providing one result for the file

6

The given command should work, but to be more precise, if the lines are ending with txt you can use the $ anchor:

awk '/txt$/{if(s){print s}{s=$0;next}}{s=s "," $0}' file

More about anchors:

Regexp Operators - The GNU Awk User's Guide

7

One in perl,

perl -lne '$\=""; $v =(/^txt/)?"\n$_":" $_"; print $v;' file