Hi Everyone,
file.txt
+++
a
b
c
+++
d
+++
asdf fefe
fff
Would like to have the output:
file1.txt
+++
a
b
c
file2.txt
+++
d
file3.txt
+++
asdf fefe
fff
Any simple way to do that, like awk, perl.
Thanks
Hi,
Do you want to separate the records into file base on '+++' ?
Could you explain more about your issue ?
Thanks, yes, want to split is when there is '+++'. (just now made a mistake for file3.txt, already corrected the output)
Try this:
awk /^+++/{c++}{print > "file" c ".txt"}' file.txt
$ cat buf
+++
a
b
c
+++
d
+++
asdf fefe
fff
$ awk '/^\+/ { f=++i".txt" } { print >> f }' buf
$ ls [1-3].txt
1.txt 2.txt 3.txt
$ cat -n [1-3].txt
1 +++
2 a
3 b
4 c
5
1 +++
2 d
3
4
1 +++
2 asdf fefe
3 fff
work perfect, if i want to remove those empty lines (only beginning and ending of each file, not the middle empty line) also in each file[1-3].txt, please advice. Thanks
To remove empty lines:
$ awk '/^\+/ { f=++i".txt" } !/^$/{ print >> f }' buf
Thanks, but seems cannot remove the empty line in the beginging and ending of each file (not middle empty lines).
assume 1.txt, yours output is
+++
a
b
c
but if in between b and c got few empty lines, those middle empty lines cannot be removed.
Thanks
Try this:
awk '/^\+/{c++; s=""}
!NF && c {s=s?s ORS s:ORS;next}
{print s $0 > "file" c ".txt"; s=""}' file.txt
only bash
while read L
do [ -z "$L" ] && continue
[ "$L" = "+++" ] && ((i++))
echo "$L" >> file$i.txt
done < file.txt
Thanks Franklin52, also Frans
i learnt a lot about awk from you.
Thanks
---------- Post updated at 06:52 AM ---------- Previous update was at 06:45 AM ----------
i got another question, to see whehter can do together.
right now, the file name is "file1.txt", "file2.txt", and "file3.txt".
instead of fix the file name into [1-3].txt, wheter i can have the file name like:
file1.txt to a.txt
file2.txt to d
file3.txt to asdf fefe
means, the file name is each file's 2nd row value. Please advice.
Thanks
Try:
awk '/^\+/{p=$0;getline;f=$0;$0=p RS $0;s=""}
!NF && f {s=s?s ORS s:ORS;next}
{print s $0 > f; s=""}' file.txt
modified to match your needs
while read L
do
[ -z "$L" ] && continue
[ "$L" = "+++" ] && { read L; F="$L.txt"; echo "+++" > "$F"; }
echo "$L" >> "$F"
done < file.txt
Thanks Franklin52 and Frans, your guys are
---------- Post updated at 08:01 AM ---------- Previous update was at 07:50 AM ----------
Hi Frank, in your awk, which part is the one "2nd row value to be the file name", if it is the 3rd row, or other row. (assume the row has the value)
Thanks
i tried many ways to modify your awk, but fail to get other rows value as the file name. 
please advice.
That happens on this line:
/^\+/{p=$0;getline;f=$0;$0=p RS $0;s=""}
Explanation:
/^\+/ -> When the line starts with a '+' then
p=$0 -> Store the line in p
getline -> get the next line (with the new file name) and proceed with the next command
f=$0 -> Store the filename in variable f
$0=p RS $0 -> Combine previous line (++++) with the current line
Hope this helps.
Thanks Frank,
i modified below, failed, actually i tried lots of combinations.
awk '/^\+/{p=$0;getline;{if($NR==2) {f=$0}};$0=p RS $0;s=""} !NF && f {s=s?s ORS s:ORS;next} {print s $0 > f; s=""}' tmp
Please guide me where wrong. Thanks
Can you explain what you're trying to achieve?
Hi Frank,
let's say
file.txt
+++
a1
a2
a4
+++
b1
b2
b3
The output:
a2.txt
+++
a1
a2
a4
b2.txt
+++
b1
b2
b3
by suing below awk i modified from your awk, but fail to do that:
awk '/^\+/{p=$0;getline;{if($NR==3) {f=$0}};$0=p RS $0;s=""} !NF && f {s=s?s ORS s:ORS;next} {print s $0 > f; s=""}' file.txt
Please advice.
Thanks
I still don't understand what you're trying to achieve..
Did you get errors or a wrong output with my last command?
Hi Frank,
Your awk is working perfect, but the output new files are a1.txt and b1.txt.
if i want to a2.txt, b2.txt, means it is not the line after the line "+++".
Sorry again if i did not explain well. Thanks