grep question re: counting DOS newlines

__John_L. · 11-10-2022, 06:22 PM

I've a file, temp.txt, that has DOS newline characters at the end of each line.

I thought the following would work:

Code:

grep -c -E "\r\n" temp.txt

however, the following warnings are returned:

Code:

grep: warning: stray \ before r
grep: warning: stray \ before n

Thanks in advance for any clarification you can provide.

michaelk · 11-10-2022, 06:55 PM

Try:

Code:

grep -c $'\015' temp.txt

015 is the octal code for carriage return.

If each line only has one carriage return you can just count the number of lines in the file with wc.

BudiKusasi · 11-10-2022, 06:56 PM

On Linux always read info by --help

...

-M, --multiline run in multiline mode
-m, --max-count=number stop after <number> matched lines
-N, --newline=type set newline type (CR, LF, CRLF, ANYCRLF, ANY, or NUL)
-n, --line-number print line number with output lines

...

...

-C, --context=number set number of context lines, before & after
-c, --count print only a count of matching lines per FILE
-D, --devices=action how to handle devices, FIFOs, and sockets
-d, --directories=action how to handle directories

...

NevemTeve · 11-10-2022, 11:38 PM

Instead of counting, you can remove them:

Code:

tr -d '\r' dos.txt >unix.txt

pan64 · 11-11-2022, 01:11 AM

if you want to count the lines use: wc -l
if you want to remove \r use: dos2unix
if you want to fix the grep warning issue: please provide more info, what os is it and which grep version is it, for me the command worked perfectly (did not give any warning).
you cannot use grep to find \n, because grep will split the file into lines and those lines do not contain this \n (and will search in these lines).

MadeInGermany · 11-12-2022, 09:18 AM

Yes, a RE always ends at the end of the line, to be matched as a $.
So it would be

Code:

grep -c '\r$'

And your grep version might not recognize \r

awk works:

Code:

awk '/\r$/{++c}END{print c+0}'