Using 'if' within a 'while' loop in Bash

Question

I have these diff results saved to a file:

bash-3.00$ cat /tmp/voo
18633a18634
> sashabrokerSTP
18634a18636
> sashatraderSTP
21545a21548
> yheemustr

I just really need the logins:

bash-3.00$ cat /tmp/voo | egrep ">|<"
> sashaSTP
> sasha
> yhee
bash-3.00$

But when I try to iterate through them and just print the names I get errors. I just do not understand the fundamentals of using "if" with "while loops". Ultimately, I want to use the while loop because I want to do something to the lines - and apparently while only loads one line into memory at a time, as opposed to the whole file at once.

bash-3.00$ while read line; do  if [[ $line =~ "<" ]] ; then  echo $line ; fi ;  done <  /tmp/voo
bash-3.00$
bash-3.00$
bash-3.00$ while read line; do  if [[ egrep "<" $line ]] ; then  echo $line ; fi ;  done    <  /tmp/voo
bash: conditional binary operator expected
bash: syntax error near `"<"'
bash-3.00$
bash-3.00$ while read line; do  if [[ egrep ">|<" $line ]] ; then  echo $line ; fi ;  done <  /tmp/voo
bash: conditional binary operator expected
bash: syntax error near `|<"'
bash-3.00$

There has to be a way to loop through the file and then do something to each line. Like this:

bash-3.00$ while read line; do  if [[ $line =~ ">" ]];
 then echo $line |  tr ">" "+" ;
 if [[ $line =~ "<" ]];
 then echo $line | tr "<" "-" ;
 fi ;
 fi ;
 done  < /tmp/voo


+ sashab
+ sashat
+ yhee
bash-3.00$

Solution 1

You should be checking for >, not <, no?

while read line; do
    if [[ $line =~ ">" ]]; then
        echo $line
    fi
done < /tmp/voo

Solution 2

Do you really need regex here? The following shell glob can also work:

while read line; do [[ "$line" == ">"* ]] && echo "$line"; done < /tmp/voo

OR use AWK:

awk '/^>/ { print "processing: " $0 }' /tmp/voo

Solution 3

grep will do:

$ grep -oP '> \K\w+' <<END
18633a18634
> sashabrokerSTP
18634a18636
> sashatraderSTP
21545a21548
> yheemustr
END

sashabrokerSTP
sashatraderSTP
yheemustr