Print lines where any word begins and ends with the same letter in linux

Question

I have input like

sie%Qu7s Kuux"oh9 ohc9ahG% hoe8Toh: Eix*ohd1 doh:bo2U Cu0doo|t zo`L9xaW
fie5Du[h Phe8aid# Opu&fai5 ieZ


where each word in the column constitutes a password ،  i am trying to print lines where where any word begins and ends with the same letter , with this we do not distinguish between uppercase and lowercase letters
i know with command  grep i can do this
cat passwords.txt | grep -e ' $[A-Z]$......\1 ' -e ' $[a-z]$......\1 '

but here the word will start and end only with same latter (uppercase or lowercase letters)  ,
like
Foh:x2zi aLoo'qu2 Ia6aig-e La{vie1E IeFoh{cI Au_h7Hee Se)f4ebi Cah$yu7m

expected output
    eif\e8AE Ieb,ing4 reph1oW* eeSh'ee8 Ah+ei4ai Oi0Ca,vu Esh1xe?e Wei&k4ic
    sie%Qu7s Kuux"oh9 ohc9ahG% hoe8Toh: Eix*ohd1 doh:bo2U Cu0doo|t zo`L9xaW
    ue5OhQu. aaf-i8uP eedae%T5 sei?M9Pu ieH[oh2l ieh~ah8A aev"oo9A Ohf"i8de
    Foh:x2zi aLoo'qu2 Ia6aig-e La{vie1E IeFoh{cI Au_h7Hee Se)f4ebi Cah$yu7m
    ahqu5To/ ahtie[H3 ioK&u5Ai nei1Za#d poo_Th9r gu|aGh7h uZ%io2ah IeNah&v7

user_11329890 · Answer

Use GNU grep:

grep -i -P '(?


The -i option turns on case insensitivity, -P turns on PCRE flavor (supports lookbehinds/lookaheads).
See regex proof.
EXPLANATION
--------------------------------------------------------------------------------
  (?

user_5866580 · Answer

With your shown samples, please try following awk program. Written and tested with GNU awk, should work in any awk.

awk '
{
  for(i=1;i<=NF;i++){
    if(tolower(substr($i,1,1))==tolower(substr($i,length($i)))){
      print
      next
    }
  }
}
'  Input_file

Explanation: Simple explanation would be, going through all the fields(columns) in all the lines and comparing their 1st and last values(each field's) if they are equal then print that line and move on to next line and so on.

Bonus solution: Just a little tweak of above solution, in case you have GNU awk then use IGNORECASE and you could shorten above to following code then.

awk -v IGNORECASE="1" '
{
  for(i=1;i<=NF;i++){
    if(substr($i,1,1)==substr($i,length($i))){
      print
      next
    }
  }
}
' Input_file

Print lines where any word begins and ends with the same letter in linux

2 Answers