regex - Find and replace strings in file with $ selectors? -

i have file , i'm looking use () select piece of text, , $1 , $2 respectivley replace string.

http://127.0.0.1:5080/tarballs/(.+)/(.+).tgz  https://registry.npmjs.org/$1/-/$1-$2.tgz 

how can in unix? want run command this:

$ regex-replace ./file.txt "http://127.0.0.1:5080/tarballs/(.+)/(.+).tgz" "https://registry.npmjs.org/$1/-/$1-$2.tgz" 

you can use sed:

s='http://127.0.0.1:5080/tarballs/user/file.tgz'  sed -e 's~http://127\.0\.0\.1:5080/tarballs/(.+)/(.+.tgz)$~https://registry.npmjs.org/\1/-/\1-\2~' <<< "$s"  https://registry.npmjs.org/user/-/user-file.tgz 

• dot special metacharacter in regex needs escaped, otherwise match character.
• ~ used regex delimiter in s (substitute) command of sed.
• -e used extended regex support.
• \1 , \2 back-references of 2 captured groups.