Adding lines to files based on their names

Shell Programming and Scripting
1

I have several files on a folder something like this:

If the file starts with 1a, I would like to add the following at the very beggining of the file

If the file starts with 1b then I should add the following

For files starting with 6

So on and so forth. Thus, at the end each file will have a reference sequence at the very top 'matching' the beggining of the file name.
Thanks in advance.

2

wrong solutions. deleted.

3

Assuming that you have files which contain the data that needs to be added to each file named after the reference. For example, file "1a.header" contains the header for reference id "1a".


function add_reference
{
  typeset file=$1
  typeset ref_id=${file%%-*}
  typeset temp=${file}.tmp
  cat ${ref_id}.header $file > $temp
  mv $temp $file
}

for file in *.fas
do
  add_reference $file
done

If you want to keep all the reference strings within your script (instead of external files), you can use associative arrays. Since, ksh does not support associative arrays, you will have to use either ksh93 or awk.

4

another way, you can build init files by name as 1a, 1b, 1 with related contents.

such as:

$ cat 1a
>Reference1a
AGCATGACGATCAGTGTGCGGGTGA

then use below scripts.

for file in `ls *.fas`
do
  str=`echo $file|cut -d \- -f1`
  cat $str $file >> $file_new
done
5

I rather have all the References within my script (I really do not like the idea of having external files)

for file in `ls 1a*.fas`
do 
  echo ">Reference1a
AGCATGACGATCAGTGTGCGGGTGA " >> $file
done

I am not getting the expected result. ALl the information in the file is being replaced by the reference sequence.

I have been trying with awk, I just cannot get it to do what I want.

6

I am not big with awk, but the original code I posted can be modified to use ksh93 as follows:

#!/bin/ksh93

typeset -A reference_headers

##
## Add all the valid references in this associative array
##
reference_headers=(
[1a]=">Reference1a
AGCATGACGATCAGTGTGCGGGTGA"
[1b]=">Reference1b
AGCATGTGTGCGGGTGA"
);

function add_reference
{
  typeset file=$1
  typeset ref_id=${file%%-*}
  typeset temp=${file}.tmp
  (echo "${reference_headers[${ref_id}]}"; cat $file) > $temp
  mv $temp $file
}

for file in *.fas
do
  add_reference $file
done

Note that, on my Mac OS X, the default ksh shell itself is ksh93, but on lots of Unix systems there are separate executables for ksh and ksh93. If your system has only ksh, if the following command gives an output similar to what is shown, its ksh93, else not:

$ echo ${.sh.version}
Version M 1993-12-28 s+
7

I will try that. Hopefully someone will post a code using awk to accomplish the same task.

8

So long as the reference ids (1a, 1b, 6) are valid sh varname sequences (alphanumeric and underscore), and the replacement text does not contain a "." on a line by itself (this tells ed to exit insert mode):

#!/bin/sh

_r1a='>Reference1a
AGCATGACGATCAGTGTGCGGGTGA'

_r1b='>Reference1b
AGCATGTGTGCGGGTGA'

_r6='>Reference6
AGCATGTGTGCGGGTGATTTTTT'

for f in *.fas; do
    printf '1i\n%s\n.\nw\nq\n' "$(eval echo \"\$_r"${f%%-*}"\")" | ed -s "$f"
done
9

alister,

I have tried your code

#!/bin/sh

_r1a='>Reference1a
AGCATGACGATCAGTGTGCGGGTGA'

_r1b='>Reference1b
AGCATGTGTGCGGGTGA'

_r6='>Reference6
AGCATGTGTGCGGGTGATTTTTT'

for f in *.fas; do
    printf '1i\n%s\n.\nw\nq\n' "$(eval echo \"\$_r"${f%%-*}"\")" | ed -s "$f"
done

And this is what I am getting

I am missing something?

10

Did you create that script under Windows? If that is the case, first run:

dos2unix script.sh

Then rerun your script.

11

No, I used gedit. Is it working in your end?

12

I didn't try it on my test machine, but that error you are seeing means that illegal control character is appended to the end of the first line. Did you try using "dos2unix" on that file? As for the gedit.. Maybe you have character encoding misconfigured there? If it is set to something like DOS, then editor will insert those "^M" characters at the end of every line. Did you use gedit to write scripts earlier without problems?

13

I did use it and I this is what I am getting

14

I would suggest to use some other editor to write your script.. You can use

nano

it is very easy to use, with all CTRL+ commands printed on the bottom. I'm sure that if you use it to create that script, it will run.

15

I use nano to write the script and I am still getting the same result.

16

Hmm.. Did you maybe "copy&paste" that script from the browser's window to nano? Or did you wrote it "by hand"? I'm running out of ideas here;P

17

I typically copy & paste -always works.
I then type the whole thing and I am still getting the same thing.

18

Please do 'cat -vet script.sh' (or whatever you named you script file' and post it here using code tags.
Also, what OS are you under?

19 
#!/bin/sh$
$
_r1a='>Reference1a$
AGCATGACGATCAGTGTGCGGGTGA'$
$
_r1b='>Reference1b$
AGCATGTGTGCGGGTGA'$
$
_r6='>Reference6$
AGCATGTGTGCGGGTGATTTTTT'$
$
for f in *.fas$
do$
^Iprintf '1i\n%s\n.\nw\nq\n' "$(eval echo \"\$_r"${f%%-*}"\")" | ed -s "$f"$
done$

---------- Post updated at 05:16 PM ---------- Previous update was at 04:32 PM ----------

This does not look that good but it is getting the job done. Now the reference matches the file extension (Sample1.1a, Sample2.1b, Sample3, 1a, etc)

#!/bin/sh
for f in *.1a
do
printf ">Reference1a\nTTGATGTGCCAACTGCGCAAATAGCAGCAGCACTACCAGGACCTTCGCCCA\n" | cat - ${f} > Converted${f}
done
for i in *.1b
do
printf ">Reference1b\nTTGATGTGCCTACCAGGACCTTCGCCCA\n" | cat - ${i} > Converted${i}
done
for e in *.3
do
printf ">Reference3\nTTGATTCGCCCA\n" | cat - ${e} > Converted${e}
done

Any other ideas?