Shell Parameter expansion

Question

I've got done the following parameter shell expansion:

trihead="$(cat ../FASTA_SEC/"$i".fa)"
echo "${trihead#"${trihead%%[!A]*}"}" > ../FASTA_SEC/"$i".fa

I'd like to trim the leading A's from the second line of a fasta file (a kind of plain text format .fa)

the input would be something like:

>B4-0K032_18670_015
AAAAAAANNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNCNNNGNNNTAGATACAAGCGAGCGGC

and i'd like the output to be like:

>B4-0K032_18670_015
NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNCNNNGNNNTAGATACAAGCGAGCGGC

By the way, does anybody know some side where shell parameter expansion is well explained?

Answer 1

To trim the leading consecutive As out of the second line of a file (and then replace the original file with the result):

$ sed '2s/^A*//' data.fa >data.out  &&  mv data.out data.fa

The sed command applies the substitution command (s) to line 2 specifically in this example. It will substitute any number of As at the start of the line with nothing.

The pattern is ^A* which means "match zero or more (*) of the single character A at the start of the line (^)".

The reason I don't use sed -i (for "in-place editing") is that the -i flag is horribly non-portable between sed implementations.

Answer 2

This will remove all leading As from a line following a line that start with >:

sed '/^>/n;s/^A*//' /path/to/file

If the output is what you want, and you want to edit the original file, use:

sed --in-place '/^>/n;s/^A*//' /path/to/file