Grep case insensitive error

UNIX for Beginners Q & A
1

Hello,
My OS: ubuntu 20
I do not understand why I am unable print expected output in this simple case:
Update: After couple attempts, I have found a way to get expected output from command line but it does not work in script.
Scratching my hair.

db

tv cine edition
tv cine action
nat geo wild

pt.xml

<channel id="TVCineAction.pt"><display-name>TVCine Action
<channel id="TVCineEdition.pt"><display-name>TVCine Edition
<channel id="NationalGeographicWildPortugal.pt"><display-name>National Geographic Wild

Expected:

tv*cine*edition TVCineEdition.pt        TVCine Edition
tv*cine*action  TVCineAction.pt TVCine Action
nat*geo*wild NationalGeographicWildPortugal.pt National Geographic Wild

What I am doing:

grep -i -m1 "wild" pt.xml

it returns empty.
I also tried dos2unix and chmod 644 it but no luck.

ls -la pt.xml:

-rw-r--r-- 1 root root 10787 Nov  5 00:31 pt.xml

when I run like below it is okay but not okay in command line:
grep -r -i -m1 "wild" pt.xml

<channel id="NationalGeographicWildPortugal.pt"><display-name>National Geographic Wild

Don't understand why it does not work just for the keyword "wild" but others.
Also tried pgrep, fgrep, egrep but no way.

get.sh

while read -r line
do
keyword=`echo "$line" | sed "s| |*|g"`
output=`grep -i -m1 -A0 "$keyword" pt.xml | sed "s|<channel id=\"||g" | sed "s|\"><display-name>|\t|g"`
echo -e "$keyword"'\t'"$output"
sed -e "/$keyword/Id" pt.xml > pt2.xml ; mv pt2.xml pt.xml #case insensitive delete and overwrite pt.xml 
done<db > output

output:

tv*cine*edition TVCineEdition.pt        TVCine Edition
tv*cine*action  TVCineAction.pt TVCine Action
nat*geo*wild

As you can see C2 and C3 in third line are empty. Is this issue related to regexp or is there a special exception for the word wild in linux?

Thank you
Boris

2

Your regular expressions are incorrect.

The * is a shell glob match character. For a grep RE, you need .* to match any sequence of characters.

tv*cine*edition and tv*cine*action both work because the eE and eA are consecutive. It finds zero or more of the e in cine, and then matches the next characters.

It fails for NationalGeographicWild because the ional and graphic are not matched by the * -- they need .*.

3

... and to add ....

I recommend you use a regex checker when writing regex (when having issues). For example (one of many online regex checkers):