creating a customized stat file out of files a current directory

Question

I have files like these:

$ cat A9E27.txt
person,42%
person,60%
$ cat ffn2eG6.txt
train,85%
$ cat sBHFgkv.txt  
person,85%
person,73%
chair,44%
person,57%
$ cat A8eVAmK.txt

etc etc Some of my txt files are actually even empty. how can I create a file that is like this

1,A9E27.txt,person,42%,person,60%
2,ffn2eG6.txt,train,85%
3,sBHFgkv.txt,person,85%,person,73%,chair,44%,person,57%
4,A8eVAmK.txt
5, etc etc 

where 1 shows the file number when I sort my file names alphabetically something like ls -1 | sort and A9E27.txt is the file name, and each line of file contents comes in front of it separated by comma

UPDATE: SchrodingersScat's answer from Ubuntu IRC is almost correct on stdout:

$ for i in * ; do echo -n "$i"; cat "$i" | sed ':a;$!{N;s/\n/, /;ba;}' ; done
5Co16.txtA9E27.txtperson,42%, person,60%
ffn2eG6.txttrain,85%
Kkc4e.txtsBHFgkv.txtperson,85%, person,73%, chair,44%, person,57%

gives correct answer but how can I save this result shown on stdout into a file correctly?

Answer 1

kerframil from bash IRC:

$  i=1; for f in *; do [[ -f $f ]] || continue; data=$(tr '\n' ',' < "$f"); printf '%s,%s,%s\n' "$((i++))" "$f" "${data%,}"; done > ../../in-your-file-correctly.txt

notes:

/join #bash 

Answer 2

How about using perl in slurp mode?

$ perl -F'\n' -0777ne 'print join(",", ++$n, $ARGV, @F), "\n"' *.txt
1,A8eVAmK.txt
2,A9E27.txt,person,42%,person,60%
3,ffn2eG6.txt,train,85%
4,sBHFgkv.txt,person,85%,person,73%,chair,44%,person,57%

Note that the shell *.txt glob causes the files to be processed in the default collation order of your locale - that might not be what you consider to be "alphabetic" (although it should be the same as that produced by ls -1, I think).

To save the result to a file, redirect it:

perl -F'\n' -0777ne 'print join(",", ++$n, $ARGV, @F), "\n"' *.txt > somefile