Find words whose first and last three letters are equal

Find words whose first and last three letters are equal is a draft programming task. It is not yet considered ready to be promoted as a complete task, for reasons that should be found in its talk page.

Task

Using the dictionary  unixdict.txt

find the words whose first and last three letters are equal.

The length of any word shown should have a length   >  5.

Other tasks related to string operations:

Metrics

• Array length
• String length
• Copy a string
• Empty string  (assignment)

Counting

• Word frequency
• Letter frequency
• Jewels and stones
• I before E except after C
• Bioinformatics/base count
• Count occurrences of a substring
• Count how many vowels and consonants occur in a string

Remove/replace

• XXXX redacted
• Conjugate a Latin verb
• Remove vowels from a string
• String interpolation (included)
• Strip block comments
• Strip comments from a string
• Strip a set of characters from a string
• Strip whitespace from a string -- top and tail
• Strip control codes and extended characters from a string

Anagrams/Derangements/shuffling

• Word wheel
• ABC problem
• Sattolo cycle
• Knuth shuffle
• Ordered words
• Superpermutation minimisation
• Textonyms (using a phone text pad)
• Anagrams
• Anagrams/Deranged anagrams
• Permutations/Derangements

Find/Search/Determine

• ABC words
• Odd words
• Word ladder
• Semordnilap
• Word search
• Wordiff  (game)
• String matching
• Tea cup rim text
• Alternade words
• Changeable words
• State name puzzle
• String comparison
• Unique characters
• Unique characters in each string
• Extract file extension
• Levenshtein distance
• Palindrome detection
• Common list elements
• Longest common suffix
• Longest common prefix
• Compare a list of strings
• Longest common substring
• Find common directory path
• Words from neighbour ones
• Change e letters to i in words
• Non-continuous subsequences
• Longest common subsequence
• Longest palindromic substrings
• Longest increasing subsequence
• Words containing "the" substring
• Sum of the digits of n is substring of n
• Determine if a string is numeric
• Determine if a string is collapsible
• Determine if a string is squeezable
• Determine if a string has all unique characters
• Determine if a string has all the same characters
• Longest substrings without repeating characters
• Find words which contains all the vowels
• Find words which contains most consonants
• Find words which contains more than 3 vowels
• Find words which first and last three letters are equals
• Find words which odd letters are consonants and even letters are vowels or vice_versa

Formatting

• Substring
• Rep-string
• Word wrap
• String case
• Align columns
• Literals/String
• Repeat a string
• Brace expansion
• Brace expansion using ranges
• Reverse a string
• Phrase reversals
• Comma quibbling
• Special characters
• String concatenation
• Substring/Top and tail
• Commatizing numbers
• Reverse words in a string
• Suffixation of decimal numbers
• Long literals, with continuations
• Numerical and alphabetical suffixes
• Abbreviations, easy
• Abbreviations, simple
• Abbreviations, automatic

Song lyrics/poems/Mad Libs/phrases

• Mad Libs
• Magic 8-ball
• 99 Bottles of Beer
• The Name Game (a song)
• The Old lady swallowed a fly
• The Twelve Days of Christmas

Tokenize

• Text between
• Tokenize a string
• Word break problem
• Tokenize a string with escaping
• Split a character string based on change of character

Sequences

• Show ASCII table
• De Bruijn sequences
• Self-referential sequences
• Generate lower case ASCII alphabet

11l

L(word) File(‘unixdict.txt’).read().split("\n")
   I word.len > 5 & word[0.<3] == word[(len)-3..]
      print(word)

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Action!

In the following solution the input file unixdict.txt is loaded from H6 drive. Altirra emulator automatically converts CR/LF character from ASCII into 155 character in ATASCII charset used by Atari 8-bit computer when one from H6-H10 hard drive under DOS 2.5 is used.

BYTE FUNC IsValidWord(CHAR ARRAY word)
  BYTE len

  len=word(0)
  IF len<=5 THEN RETURN (0) FI
  IF word(1)#word(len-2) THEN RETURN(0) FI
  IF word(2)#word(len-1) THEN RETURN(0) FI
  IF word(3)#word(len) THEN RETURN(0) FI
RETURN (1)

PROC FindWords(CHAR ARRAY fname)
  CHAR ARRAY line(256)
  CHAR ARRAY tmp(256)
  BYTE dev=[1]

  Close(dev)
  Open(dev,fname,4)
  WHILE Eof(dev)=0
  DO
    InputSD(dev,line)
    IF IsValidWord(line) THEN
      PrintE(line)
    FI
  OD
  Close(dev)
RETURN

PROC Main()
  CHAR ARRAY fname="H6:UNIXDICT.TXT"

  FindWords(fname)
RETURN

Output:

Screenshot from Atari 8-bit computer

antiperspirant calendrical einstein hotshot murmur oshkosh tartar testes

Ada

with Ada.Text_Io;
with Ada.Strings.Fixed;

procedure Find_Three_Equals is
   use Ada.Text_Io;
   use Ada.Strings.Fixed;

   Filename : constant String := "unixdict.txt";
   File     : File_Type;
begin
   Open (File, In_File, Filename);
   while not End_Of_File (File) loop
      declare
         Word  : constant String  := Get_Line (File);
         First : String renames Head (Word, 3);
         Last  : String renames Tail (Word, 3);
      begin
         if First = Last and Word'Length > 5 then
            Put_Line (Word);
         end if;
      end;
   end loop;
   Close (File);
end Find_Three_Equals;

ALGOL 68

# find 6 (or more) character words with the same first and last 3 letters #
IF  FILE input file;
    STRING file name = "unixdict.txt";
    open( input file, file name, stand in channel ) /= 0
THEN
    # failed to open the file #
    print( ( "Unable to open """ + file name + """", newline ) )
ELSE
    # file opened OK #
    BOOL at eof := FALSE;
    # set the EOF handler for the file #
    on logical file end( input file, ( REF FILE f )BOOL:
                                     BEGIN
                                         # note that we reached EOF on the #
                                         # latest read #
                                         at eof := TRUE;
                                         # return TRUE so processing can continue #
                                         TRUE
                                     END
                       );
    INT count := 0;
    WHILE STRING word;
          get( input file, ( word, newline ) );
          NOT at eof
    DO
        IF INT w len = ( UPB word + 1 ) - LWB word;
           w len > 5
        THEN
            IF word[ 1 : 3 ] = word[ w len - 2 : ]
            THEN
                count +:= 1;
                print( ( word, " " ) );
                IF count MOD 5 = 0
                THEN print( ( newline ) )
                ELSE FROM w len + 1 TO 14 DO print( ( " " ) ) OD
                FI
            FI
        FI
    OD;
    print( ( newline, "found ", whole( count, 0 ), " words with the same first and last 3 characters", newline ) );
    close( input file )
FI

Output:

antiperspirant calendrical    einstein       hotshot        murmur
oshkosh        tartar         testes
found 8 words with the same first and last 3 characters

Arturo

words: read.lines relative "unixdict.txt"
equalHeadTail?: function [w][
    equal? first.n: 3 w last.n: 3 w
]

loop words 'word [
    if 5 < size word [
        if equalHeadTail? word ->
            print word
    ]
]

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

AutoHotkey

FileRead, db, % A_Desktop "\unixdict.txt"
for i, word in StrSplit(db, "`n", "`r")
    if StrLen(word) < 6
        continue
    else if (SubStr(word, 1, 3) = SubStr(word, -2))
        result .= word "`n"
MsgBox, 262144, , % result
return

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

AWK

# syntax: GAWK -f FIND_WORDS_WHICH_FIRST_AND_LAST_THREE_LETTERS_ARE_EQUALS.AWK unixdict.txt
(length($0) >= 6 && substr($0,1,3) == substr($0,length($0)-2,3))
END {
    exit(0)
}

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

C++

#include <cstdlib>
#include <fstream>
#include <iostream>

int main(int argc, char** argv) {
    const char* filename(argc < 2 ? "unixdict.txt" : argv[1]);
    std::ifstream in(filename);
    if (!in) {
        std::cerr << "Cannot open file '" << filename << "'.\n";
        return EXIT_FAILURE;
    }
    std::string word;
    int n = 0;
    while (getline(in, word)) {
        const size_t len = word.size();
        if (len > 5 && word.compare(0, 3, word, len - 3) == 0)
            std::cout << ++n << ": " << word << '\n';
    }
    return EXIT_SUCCESS;
}

Output:

1. antiperspirant
2. calendrical
3. einstein
4. hotshot
5. murmur
6. oshkosh
7. tartar
8. testes

F#

// First and last three letters are equal. Nigel Galloway: February 18th., 2021
let fN g=if String.length g<6 then false else g.[..2]=g.[g.Length-3..]
seq{use n=System.IO.File.OpenText("unixdict.txt") in while not n.EndOfStream do yield n.ReadLine()}|>Seq.filter fN|>Seq.iter(printfn "%s")

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Factor

Read entire file

This version reads the entire dictionary file into memory and filters it. This is the fastest version by far. Factor is optimized for making multiple passes over data; it actually takes longer if we combine the two filters into one, either with short-circuiting or non-short-circuiting and.

USING: io io.encodings.ascii io.files kernel math sequences ;

"unixdict.txt" ascii file-lines
[ length 5 > ] filter
[ [ 3 head-slice ] [ 3 tail-slice* ] bi = ] filter
[ print ] each

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Read file line by line

This version reads the dictionary file line by line and prints out words that fit the criteria. This ends up being a bit more imperative and deeply nested, but unlike the version above, we only load one word at a time, saving quite a bit of memory.

USING: combinators.short-circuit io io.encodings.ascii io.files
kernel math sequences ;

"unixdict.txt" ascii [
    [
        readln dup
        [
            dup
            {
                [ length 5 > ]
                [ [ 3 head-slice ] [ 3 tail-slice* ] bi = ]
            } 1&&
            [ print ] [ drop ] if
        ] when*
    ] loop
] with-file-reader

Output:

As above.

Lazy file I/O

This version lazily reads the input file by treating a stream like a lazy list with the llines word. This allows us the nice style of the first example with the memory benefits of the second example. Unlike in the first example, combining the filters would buy us some time here, as lazy lists aren't as efficient as sequences.

USING: io io.encodings.ascii io.files kernel lists lists.lazy
math sequences ;

"unixdict.txt" ascii <file-reader> llines
[ length 5 > ] lfilter
[ [ 3 head-slice ] [ 3 tail-slice* ] bi = ] lfilter
[ print ] leach

Output:

As above.

Forth

Works with: Gforth

: first-last-three-equal { addr len -- ? }
  len 5 <= if false exit then
  addr 3 addr len 3 - + 3 compare 0= ;

256 constant max-line

: main
  0 0 { count fd-in }
  s" unixdict.txt" r/o open-file throw to fd-in
  begin
    here max-line fd-in read-line throw
  while
    here swap 2dup first-last-three-equal if
      count 1+ to count
      count 1 .r ." . " type cr
    else
      2drop
    then
  repeat
  drop
  fd-in close-file throw ;

main
bye

Output:

1. antiperspirant
2. calendrical
3. einstein
4. hotshot
5. murmur
6. oshkosh
7. tartar
8. testes

FreeBASIC

#define NULL 0

type node
    word as string*32   'enough space to store any word in the dictionary
    nxt as node ptr
end type

function addword( tail as node ptr, word as string ) as node ptr
    'allocates memory for a new node, links the previous tail to it,
    'and returns the address of the new node
    dim as node ptr newnode = allocate(sizeof(node))
    tail->nxt = newnode
    newnode->nxt = NULL
    newnode->word = word
    return newnode
end function

function length( word as string ) as uinteger
    'necessary replacement for the built-in len function, which in this
    'case would always return 32
    for i as uinteger = 1 to 32
        if asc(mid(word,i,1)) = 0 then return i-1
    next i
    return 999
end function

dim as string word
dim as node ptr tail = allocate( sizeof(node) )
dim as node ptr head = tail, curr = head, currj
dim as uinteger ln
tail->nxt = NULL
tail->word = "XXXXHEADER"

open "unixdict.txt" for input as #1
while true
    line input #1, word
    if word = "" then exit while
    if length(word)>5 then tail = addword( tail, word )
wend
close #1

while curr->nxt <> NULL
    word = curr->word
    ln = length(word)
    for i as uinteger = 1 to 3
        if mid(word,i,1) <> mid(word,ln-3+i,1) then goto nextword
    next i
    print word
    nextword:
    curr = curr->nxt
wend

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Go

package main

import (
    "bytes"
    "fmt"
    "io/ioutil"
    "log"
    "unicode/utf8"
)

func main() {
    wordList := "unixdict.txt"
    b, err := ioutil.ReadFile(wordList)
    if err != nil {
        log.Fatal("Error reading file")
    }
    bwords := bytes.Fields(b)
    count := 0
    for _, bword := range bwords {
        s := string(bword)
        if utf8.RuneCountInString(s) > 5 && (s[0:3] == s[len(s)-3:]) {
            count++
            fmt.Printf("%d: %s\n", count, s)
        }
    }
}

Output:

1: antiperspirant
2: calendrical
3: einstein
4: hotshot
5: murmur
6: oshkosh
7: tartar
8: testes

J

   >(#~ ((3&{. -: _3&{.)*5<#)@>) cutLF fread 'unixdict.txt'
antiperspirant
calendrical   
einstein      
hotshot       
murmur        
oshkosh       
tartar        
testes

jq

Works with: jq

Works with gojq, the Go implementation of jq

select(length > 5 and .[:3] == .[-3:])

Output:

Invocation example: jq -rRM -f program.jq unixdict.txt

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Julia

See Alternade_words#Julia for the foreachword function.

matchfirstlast3(word, _) = length(word) > 5 && word[1:3] == word[end-2:end] ? word : ""
foreachword("unixdict.txt", matchfirstlast3, numcols=4)

Output:

Word source: unixdict.txt

antiperspirant calendrical    einstein       hotshot
murmur         oshkosh        tartar         testes

Ksh

#!/bin/ksh

# Find list of words (> 5 chars) where 1st 3 and last 3 letters are the same

#	# Variables:
#
dict='../unixdict.txt'
integer MIN_LEN=5
integer MATCH_NO=3

 ######
# main #
 ######

 while read word; do
	(( ${#word} <= MIN_LEN )) && continue

	first=${word:0:${MATCH_NO}}
	last=${word:$((${#word}-MATCH_NO)):${#word}}

	[[ ${first} == ${last} ]] && print ${word}

 done < ${dict}

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Mathematica/Wolfram Language

dict = Once[Import["https://web.archive.org/web/20180611003215/http://www.puzzlers.org/pub/wordlists/unixdict.txt"]];
dict //= StringSplit[#, "\n"] &;
dict //= Select[StringLength /* GreaterThan[5]];
Select[dict, StringTake[#, 3] === StringTake[#, -3] &]

Output:

{"antiperspirant", "calendrical", "einstein", "hotshot", "murmur", "oshkosh", "tartar", "testes"}

Nim

for word in "unixdict.txt".lines:
  if word.len > 5:
    if word[0..2] == word[^3..^1]:
      echo word

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Perl

as one-liner ..

// 20210212 Perl programming solution

perl -ne '/(?=^(.{3}).*\1$)^.{6,}$/&&print' unixdict.txt

# minor variation

perl -ne 's/(?=^(.{3}).*\1$)^.{6,}$/print/e' unixdict.txt

Phix

with javascript_semantics
function flaste(string word) return length(word)>5 and word[1..3]=word[-3..-1] end function
sequence flastes = filter(unix_dict(),flaste)
printf(1,"%d words: %s\n",{length(flastes),join(shorten(flastes,"",3))})

Output:

8 words: antiperspirant calendrical einstein hotshot murmur oshkosh tartar testes

PL/I

firstAndLast3Equal: procedure options(main);
    declare dict file;
    open file(dict) title('unixdict.txt');
    on endfile(dict) stop;
    
    declare word char(32) varying, (first3, last3) char(3);
    do while('1'b);
        get file(dict) list(word);
        first3 = substr(word, 1, 3);
        last3 = substr(word, length(word)-2, 3);
        if length(word) > 5 & first3 = last3 then
            put skip list(word); 
    end;    
end firstAndLast3Equal;

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Python

Tested on Python 3+, the file download will work only if the link is still active. It is possible that you may be able to fetch the file in your browser but download via code may still fail. Check whether you are connected to a VPN, it works on open networks

import urllib.request
urllib.request.urlretrieve("http://wiki.puzzlers.org/pub/wordlists/unixdict.txt", "unixdict.txt")

dictionary = open("unixdict.txt","r")

wordList = dictionary.read().split('\n')

dictionary.close()

for word in wordList:
    if len(word)>5 and word[:3].lower()==word[-3:].lower():
        print(word)

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Quackery

  [ [] swap ]'[ swap
    witheach [
      dup nested
      unrot over do
      iff [ dip join ]
      else nip
    ] drop ]                   is filter ( [ --> [ )

  $ "unixdict.txt" sharefile drop nest$
  filter [ size 5 > ]
  filter [ 3 split -3 split nip = ]
  witheach [ echo$ cr ]

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

R

dict <- scan("https://web.archive.org/web/20180611003215/http://www.puzzlers.org/pub/wordlists/unixdict.txt", what = character())
dict[nchar(dict) > 5 & substr(dict, 1, 3) == substr(dict, nchar(dict) - 2, nchar(dict))]

Racket

#lang racket

(define ((prefix-and-suffix-match? len) str)
  (let ((l (string-length str)))
    (and (>= l (* 2 len))
         (string=? (substring str 0 len)
                   (substring str (- l len))))))

(module+ main
  (filter (prefix-and-suffix-match? 3) (file->lines "../../data/unixdict.txt")))

Output:

'("antiperspirant" "calendrical" "einstein" "hotshot" "murmur" "oshkosh" "tartar" "testes")

Raku

# 20210210 Raku programming solution

my ( \L, \N, \IN ) = 5, 3, 'unixdict.txt';

for IN.IO.lines { .say if .chars > L and .substr(0,N) eq .substr(*-N,*) }

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Red

Red[]

foreach word read/lines %unixdict.txt [
    if all [
        greater? length? word 5
        equal? take/part copy word 3 take/part/last copy word 3
    ][
        print word
    ]
]

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

REXX

This REXX version doesn't care what order the words in the dictionary are in,   nor does it care what
case  (lower/upper/mixed)  the words are in,   the search for the words and vowels is   caseless.

The program verifies that the first and last three characters are, indeed, letters.

It also allows the length (3) of the first and last number of letters to be specified,   and also the minimum length of the
words to be searched on the command line (CL) as well as specifying the dictionary file identifier.

/*REXX pgm finds words in an specified dict. which have the same 1st and last 3 letters.*/
parse arg minL many iFID .                       /*obtain optional arguments from the CL*/
if minL=='' | minL=="," then minL=  6            /* "      "         "   "   "     "    */
if many=='' | many=="," then many=  3            /* "      "         "   "   "     "    */
if iFID=='' | iFID=="," then iFID='unixdict.txt' /* "      "         "   "   "     "    */

              do #=1  while lines(iFID)\==0      /*read each word in the file  (word=X).*/
              x= strip( linein( iFID) )          /*pick off a word from the input line. */
              @.#= x                             /*save:  the original case of the word.*/
              end   /*#*/
#= # - 1                                         /*adjust word count because of DO loop.*/
say copies('─', 30)     #     "words in the dictionary file: "       iFID
finds= 0                                         /*word count which have matching end.  */
                                                 /*process all the words that were found*/
     do j=1  for #;          $= @.j;    upper $  /*obtain dictionary word; uppercase it.*/
     if length($)<minL  then iterate             /*Word not long enough?   Then skip it.*/
     lhs= left($, many);     rhs= right($, many) /*obtain the left & right side of word.*/
     if \datatype(lhs || rhs, 'U')  then iterate /*are the left and right side letters? */
     if lhs \== rhs                 then iterate /*Left side match right side?  No, skip*/
     finds= finds + 1                            /*bump count of only "e" vowels found. */
     say right( left(@.j, 30),  40)              /*indent original word for readability.*/
     end        /*j*/
                                                 /*stick a fork in it,  we're all done. */
say copies('─', 30)  finds  " words found that the left "   many   ' letters match the' ,
                            "right letters which a word has a minimal length of "     minL

output   when using the default inputs:

────────────────────────────── 25104 words in the dictionary file:  unixdict.txt
          antiperspirant
          calendrical
          einstein
          hotshot
          murmur
          oshkosh
          tartar
          testes
────────────────────────────── 8  words found that the left  3  letters match the right letters which a word has a minimal length of  6

Ring

load "stdlib.ring"

cStr = read("unixdict.txt")
wordList = str2list(cStr)
num = 0

see "working..." + nl
see "Words are:" + nl

ln = len(wordList)
for n = ln to 1 step -1
    if len(wordList[n]) < 6
       del(wordList,n)
    ok
next

for n = 1 to len(wordList)
    if left(wordList[n],3) = right(wordList[n],3) 
       num = num + 1
       see "" + num + ". " + wordList[n] + nl
    ok
next

see "done..." + nl

Output:

working...
Words are:
1. antiperspirant
2. calendrical
3. einstein
4. hotshot
5. murmur
6. oshkosh
7. tartar
8. testes
done...

Ruby

words = File.readlines("unixdict.txt").map(&:chomp)
puts words.select{|w| w.end_with?(w[0,3]) && w.size > 5}

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Sidef

File("unixdict.txt").open_r.each {|word|
    word.len > 5 || next
    if (word.ends_with(word.first(3))) {
        say word
    }
}

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes

Swift

import Foundation

do {
    try String(contentsOfFile: "unixdict.txt", encoding: String.Encoding.ascii)
        .components(separatedBy: "\n")
        .filter{$0.count > 5 && $0.prefix(3) == $0.suffix(3)}
        .enumerated()
        .forEach{print("\($0.0 + 1). \($0.1)")}
} catch {
    print(error.localizedDescription)
}

Output:

1. antiperspirant
2. calendrical
3. einstein
4. hotshot
5. murmur
6. oshkosh
7. tartar
8. testes

VBScript

After building a program checking for the 3 letters in any order, I found people just checked the same trigraph at start and end. I modified my program so it puts an asterisk after the words in the "standard" answer. Run the ssript with CScript.

with createobject("ADODB.Stream")
  .charset ="UTF-8"
  .open
  .loadfromfile("unixdict.txt")
  s=.readtext
end with  
a=split (s,vblf)

set d= createobject("Scripting.Dictionary")
for each aa in a
  x=trim(aa)
  l=len(x)
  if l>5 then
   d.removeall
   for i=1 to 3
     m=mid(x,i,1)
     if not d.exists(m) then d.add m,null
   next
   res=true
   for i=l-2 to l
     m=mid(x,i,1)
     if not d.exists(m) then 
       res=false:exit for 
      else
        d.remove(m)
      end if        
   next 
   if res then 
     wscript.stdout.write left(x & space(15),15)
     if left(x,3)=right(x,3) then  wscript.stdout.write "*"
     wscript.stdout.writeline
    end if 
  end if
next

Output:

alfalfa
antiperspirant *
calendrical    *
cataract
deadhead
earthenware
einstein       *
encumbrance
greenberg
hannah
hotshot        *
marjoram
murmur         *
oshkosh        *
tartar         *
teammate
tenement
testes         *

Wren

Library: Wren-fmt

import "io" for File
import "/fmt" for Fmt

var wordList = "unixdict.txt" // local copy
var count = 0
File.read(wordList).trimEnd().split("\n").
    where { |w|
        return w.count > 5 && (w[0..2] == w[-3..-1])
    }.
    each { |w|
        count = count + 1
        Fmt.print("$d: $s", count, w)
    }

Output:

1: antiperspirant
2: calendrical
3: einstein
4: hotshot
5: murmur
6: oshkosh
7: tartar
8: testes

XPL0

string 0;               \Use zero-terminated strings
int  I, Ch, Len;
char Word(100); \(longest word in unixdict.txt is 22 chars)
def  LF=$0A, CR=$0D, EOF=$1A;
[FSet(FOpen("unixdict.txt", 0), ^I);    \open dictionary and set it to device 3
OpenI(3);
repeat  I:= 0;
        loop    [repeat Ch:= ChIn(3) until Ch # CR;     \remove possible CR
                if Ch=LF or Ch=EOF then quit;
                Word(I):= Ch;
                I:= I+1;
                ];
        Word(I):= 0;                    \terminate string
        Len:= I;
        if Len >= 6 then
            if Word(0) = Word(Len-3) &
               Word(1) = Word(Len-2) &
               Word(2) = Word(Len-1) then
                  [Text(0, Word);  CrLf(0)];
until   Ch = EOF;
]

Output:

antiperspirant
calendrical
einstein
hotshot
murmur
oshkosh
tartar
testes