Paste each single row from n separate text files into a new file containing n rows

Mike_V · 04-27-2009, 09:51 AM

Hi there,

Say I have 3 text files each with one row of numbers:

file1.txt contains:

Code:

9830987716238

file2.txt contains:

Code:

908095823098

file3.txt contains:

Code:

43098104982

and I want to paste them in one file called

file_all.txt containing:

Code:

9830987716238
908095823098
43098104982

how could I do that?

I use "paste" a lot to add files containing columns of text, but these are rows...

Thanks a lot!
Mike

druuna · 04-27-2009, 09:53 AM

Hi,

cat file1 file2 file3 fileN > newfile

Hope this helps.

Mike_V · 04-27-2009, 10:08 AM

so simple... and so helpful, thanks!
Mike

druuna · 04-27-2009, 10:11 AM

Hi,

Simple if you know how

BTW: if you have a lot of files that need to be processed, you can use wildcards: cat file* > newfile

Mike_V · 04-27-2009, 10:46 AM

Hi Druuna,

Indeed I was using cat file* > newfile.

One problem now. It worked on a test I did myself but not on my actual files. My actual files each contain one row of almost 400 numbers. For some reason they are pasted into a single row in newfile. The strange thing to me is that I see no difference between my test files and the actual files in terms of how the line ends or so...

Would you happen to know how I can be sure that each file will start on a new row in newfile?

~Mike

druuna · 04-27-2009, 10:57 AM

Hi,

What is the output of the following commands:

file name.actual.file
od -c name.actual.file

Mike_V · 04-27-2009, 11:09 AM

$ file list_-22_-10_-12_per_track_number_m.txt
list_-22_-10_-12_per_track_number_m.txt: character Computer Graphics Metafile

$ od -c list_-22_-10_-12_per_track_number_m.txt
0000000 0 7 0 0 0 0 0 0
0000020 0 0 0 0 0 0 0 0
0000040 0 1 4 0 0 0 0 0 0
0000060 0 0 0 0 0 1 0 0
0000100 0 0 0 0 0 2 0 0
0000120 0 0 1 0 0 0 0 0
0000140 0 0 0 0 0 0 0 0
*
0000260 0 0 2 0 0 0 0 0
0000300 0 0 0 0 0 0 0 0
*
0000420 0 0 0 0 0 0 1 6 0
0000440 0 0 0 0 0 0 0 0
*
0001220 0 0 0 0 1 4 0 0
0001240 0 0 0 0 0 0 0 0
*
0001440 0
0001442
$

and also here for the "test" file:

$ od -c test1.txt
0000000 0 7 0 0 0 0 0 0
0000020 0 0 0 0 0 0 0 0
0000040 0 1 4 \n
0000045

$ od -c test1.txt
0000000 0 7 0 0 0 0 0 0
0000020 0 0 0 0 0 0 0 0
0000040 0 1 4 \n
0000045

So I can already see a difference myself... the "\n" at the end that my actual files don't have...

I also noticed that "\n" is introduced after I close gedit (it wasn't there when I made the test file)

(One more piece of information: my actual files are 399 in total and their number of columns ranges between 1 and 399)

druuna · 04-27-2009, 11:31 AM

Hi,

Computer Graphics Metafile files are a binary format, the tools discussed are meant for text files.

I wouldn't know how to manipulate these.........

Mike_V · 04-27-2009, 01:11 PM

In the meantime I did find a solution:

I typed in the command line:

Code:

foreach file (list*)
    echo `cat $file` >> newfile.txt
end

NOTE 1: My file names start with "list..."
NOTE 2: The quotation marks are not ' but ` (same key as ~)

That did work.

Thanks

druuna · 04-27-2009, 01:20 PM

Hi,

That, indeed, works. Very nice workaround!

For those that wonder why this does work:

Because cat $file is placed between backticks, it is executed in a new shell. There's still no newline after the cat command is executed, but when the 'new shell' itself is closed, a newline is added before it is written to newfile.txt.

ghostdog74 · 04-27-2009, 11:27 PM

i do not have your sample files to try, but you can try this. if it doesn't work, so be it.

Code:

awk '{print $0 > 'newfile' }' list*

Mike_V · 04-27-2009, 11:51 PM

ghostdog74:

At first I got an error message but after changing some quotation marks this worked:

Code:

awk '{print $0 > "newfile.txt" }' file*

Very short, very nice.

Cheers,