how to process last x lines of a file - shell

Question

I want to Analyse a logfile for specific Errors.
Therefore i want to be able to loop through the last x lines of the files and check every line with a specific REGEX Pattern and then define a specific return value.
The logfile Looks in case of success as follows at the Moment when i want to check it.
….
sftp> get blahblah/blahblah
sftp> bye
In case of an Error there is something between the two sftp lines.
What i allready tried is to solve the Problem with a specific regex which worked fine on some online Regex testers but couldn´t get it to work in ksh.
My current Approach is the following
LOG_FIL="test_log"
MODE="${1}"
check_log_file() {
ERRNBR=${1}
REGEX=${2}
TAIL=${3}
RETURN="0"
echo "ERRNBR = ${ERRNBR}"
echo "REGEX = ${REGEX}"
echo "TAIL = ${TAIL}"
while read line; do
echo "${line}"
if [[ "${line}" =~ ${REGEX} ]]; then
RETURN="0"
echo "bin hier"
else
RETURN=${ERRNBR}
echo "bin wo anders"
break
fi
done <<<$(tail -${TAIL} ${LOG_FIL})
echo "${RETURN}"
return ${RETURN}
}
echo "sftp> get cwi/cdk_final*" >> ${LOG_FIL}
if [ "${MODE}" == "1" ]; then
echo "Werner ist der beste" >>${LOG_FIL}
fi
check_log_file "22" "^(sftp> ).*$" "1"
echo "$?"
echo "sftp> bye" >> ${LOG_FIL}
check_log_file "21" "((sftp> ).*|(sftp> bye))" "2"
echo "$?"
The results i get are the following
edv> sh cdk_test4sftp.sh 1
ERRNBR = 22
REGEX = ^(sftp> ).*$
TAIL = 1
Werner ist der beste
bin wo anders
22
22
ERRNBR = 21
REGEX = ((sftp> ).*|(sftp> bye))
TAIL = 2
Werner ist der beste sftp> bye
bin hier
0
0
What i hoped to achieve was that the Output coming from the tail command would be seperated. So that i ccould test each line individually.

Answer 1

Your second regex:
((sftp> ).*|(sftp> bye))
Matches the following line, which is why your function returns 0:
Werner ist der beste sftp> bye
Since you want to match the following pattern on each line:
sftp> get blahblah/blahblah
sftp> bye
Your regex should look more like the first one you used to match:
^(sftp> ).*$