Bash refers to keywords in a file and extracts parts common to the string.

If you use hoge_list like a dictionary and try to find a match and delete it…, it's quite troublesome, so make it simpler,

• Delete the string if it starts with Chuo Ward, Tokyo
• Delete the string if it starts with Meguro Ward, Tokyo
• Delete the string if it starts with Suginami Ward, Tokyo
• If the string starts with Yokohama City, Kanagawa Prefecture, delete it

It is useful to use sed for 一致remove matches 削除

s|^Chuo Ward, Tokyo||
s|^Meguro Ward, Tokyo||
s|^Suginami Ward, Tokyo||
s|^Yokohama City, Kanagawa Prefecture||

Create a file named hoge_list.sed and

 echo 'Asagaya 1-2-3' in Suginami Ward, Tokyo | sed-foge_list.sed

and

Asagaya 1-2-3

is obtained.

Since hoge_list actually has quite a few lines, you want to automate the creation of hoge_list.sed based on it.
Including that part, the shell script looks like this:

#!/bin/bash

cathoge_list | sed-e's / \(.*\)/s |^\1||/'>hoge_list.sed

sed-foge_list.sed | sed-e's /^//'

Save this script as test.sh and run it.

$echo'Tokyo Suginami Ward Asagaya 1-2-3'|./test.sh
Asagaya 1-2-3
$ 

You can see the output:

I tried Awk.

$INPUT='Asagaya 1-2-3 in Suginami Ward, Tokyo'
$ awk-vinput="$INPUT"'
    input~"^"$0{
      sub("^"$0"[\t]*", "", input)
      exit
    }END {print input}
  ' hoge_list

Asagaya 1-2-3    

If only the bash function can be realized, I think it would be like the following.

#!/bin/bash
INPUT="Asagaya 1-2-3 in Suginami Ward, Tokyo"
shopt-extglob
list = $(<hoge_list)
pattern=${list//$'\n'/|}
echo${INPUT#@($pattern)}