Hi all,
i have a text file of several thousand lines, and each lines needs to be outputed in a separate text file.

Split command doesn´t work, because of suffix failure in the case of so many lines. This is what I have tried so far:

while read LINE; do echo $LINE>$LINE.txt; done <all_texts_new.txt

This works perfectly until the ">$LINE.txt" part, here it seems to fail.

Any ideas? Thanks a lot

Try while read LINE; do echo ${LINE}>"${LINE}".txt; done <all_texts_new.txt (By default bash will "tokenize" LINE, and any blanks, etc., in LINE will make bash try the parts of LINE after the blank as commands to be executed if the quotes are missing.)

There are other, more sophisticated, ways to do this sort of thing. And the .txt, while it may be useful as a "tag" for you, is not normally required. (Linux files a categorized by an internal "magic number," and, although extensions can often help users,they are't needed by most programs.)

By the way, you're going to end up with a directory containing files whose contents is redundant with the file's name. You could, instead, create empty files without loosing any information: while read LINE; touch "${LINE}"; done <all_texts_new.txt

1 members found this post helpful.

Thank you for your suggestion, PTrenholme! Unfortunately, this did not work in my case, probably because lines are too long to function as file names???

Anyways, found the solution just a couple of minutes ago:

awk '{print >> "s" sprintf("%03d",++c) ".txt"}' all_texts_new.txt

Phew, glad this is done, it was a loong evening!

What if the line is blank? Do you want an empty file for that line?

IIRC, the file name length limit is about 256 characters, but it may be longer if you're using a 64-bit OS.

Anyhow, if you just wanted to name the file sequentially, you should have said so in your first post. The AWK code to do what you first said you wanted would be something like

gawk '{print > "\"" gensub(/\"/,"\\\"","g",$0) "\".txt"}' all_texts_new.txt

(Note that the gensub function is a gawk extension to ANSI AWK.)

And the bash code to do what you ended up doing would be:

c=0;while read LINE; do c=$((++c));echo $LINE>$(printf "s%03d" ${c});done <all_texts_new.txt

(But the AWK would be much faster . . .)

Well I will assume you do not wish for blank files, but also that the line an entry is on is important, so this could work:

Please use ***[code][/code] tags*** around your code and data, to preserve formatting and to improve readability. Please do not use quote tags, colors, or other fancy formatting.


The "++" increment operator resets the variable contents directly, so there's no need to use the "=".

In bash or ksh, you can use the ((..)) arithmetic operator. For portability, prefix $((..)) with the true command or use let.

In this case however, we want to pass the value directly to printf, so we use $((..)).

Also, don't forget to quote your expansions, to avoid word splitting.

So, to flesh out the loop (assuming bash):

The [[ -n $line ]] test catches cases where there's no final newline in the input text.

Finally, since environment variables are generally all upper-case, it's good practice to keep your own user variables in lower-case or mixed-case to help differentiate them.