A semordnilap is a word (or phrase) that spells a different word (or phrase) backward. "Semordnilap" is a word that itself is a semordnilap.

Task
Semordnilap
You are encouraged to solve this task according to the task description, using any language you may know.

Example: lager and regal

Task

This task does not consider semordnilap phrases, only single words. Using only words from this list, report the total number of unique semordnilap pairs, and print 5 examples. Two matching semordnilaps, such as lager and regal, should be counted as one unique pair. (Note that the word "semordnilap" is not in the above dictionary.)


Other tasks related to string operations:
Metrics
Counting
Remove/replace
Anagrams/Derangements/shuffling
Find/Search/Determine
Formatting
Song lyrics/poems/Mad Libs/phrases
Tokenize
Sequences



11l

Translation of: Python
V wordset = Set(File(‘unixdict.txt’).read().split("\n"))
V revlist = wordset.map(word -> reversed(word))
V pairs = Set(zip(wordset, revlist).filter((wrd, rev) -> wrd < rev & rev C :wordset))

print(pairs.len)
print(sorted(Array(pairs), key' p -> (p[0].len, p))[(len)-5..])
Output:
158
[(damon, nomad), (lager, regal), (leper, repel), (lever, revel), (kramer, remark)]

8th

We're using a map to keep track of what's been seen, and an array to store the results. We load the "unixdict.txt" as an "asset", meaning a file stored alongside the program code:

[] var, results

: processline \ m s --
  clone nip
  tuck s:rev
  m:exists? if 
    results @ rot a:push drop
  else
    swap true m:!
  then ;

{} "unixdict.txt" app:asset >s
' processline s:eachline 

results @ dup a:len . " pairs" . cr
a:shuffle
( a:shift dup .  " is the reverse of " . s:rev . cr ) 5 times bye
Output:
158 pairs
trap is the reverse of part
nab is the reverse of ban
la is the reverse of al
ta is the reverse of at
tin is the reverse of nit

Ada

Before tackling the real problem, we specify a package String_Vectors and a class String_Vectors.Vec, to store the list of words in the dictionary:

with Ada.Containers.Indefinite_Vectors, Ada.Text_IO;

package String_Vectors is

   package String_Vec is new Ada.Containers.Indefinite_Vectors
     (Index_Type => Positive, Element_Type => String);

   type Vec is new String_Vec.Vector with null record;

   function Read(Filename: String) return Vec;
     -- uses Ada.Text_IO to read words from the given file into a Vec
     -- requirement: each word is written in a single line

   function Is_In(List: Vec;
                  Word: String;
                  Start: Positive; Stop: Natural) return Boolean;
     -- checks if Word is in List(Start .. Stop);
     -- requirement: the words in List are sorted alphabetically
end String_Vectors;

The specified class String_Vectors.Vec has been derived from Ada.Containers.Indefinite_Vectors.Vector and provides two additional primitive operations Read and Is_In. Here is the implementation:

package body String_Vectors is

   function Is_In(List: Vec;
                  Word: String;
                  Start: Positive; Stop: Natural) return Boolean is
      Middle: Positive;
   begin
      if Start > Stop then
         return False;
      else
         Middle := (Start+Stop) / 2;
            if List.Element(Middle) = Word then
               return True;
            elsif List.Element(Middle) < Word then
               return List.Is_In(Word, Middle+1, Stop);
            else
               return List.Is_In(Word, Start, Middle-1);
            end if;
      end if;
   end Is_In;

   function Read(Filename: String) return Vec is
      package IO renames Ada.Text_IO;
      Persistent_List: IO.File_Type;
      List: Vec;
   begin
      IO.Open(File => Persistent_List, Name => Filename, Mode => IO.In_File);
      while not IO.End_Of_File(Persistent_List) loop
         List.Append(New_Item => IO.Get_Line(Persistent_List));
      end loop;
      IO.Close(Persistent_List);
      return List;
   end Read;

end String_Vectors;

This is the main program:

with String_Vectors, Ada.Text_IO, Ada.Command_Line;

procedure Semordnilap is

   function Backward(S: String) return String is
   begin
      if S'Length < 2 then
         return S;
      else
         return (S(S'Last) & Backward(S(S'First+1 .. S'Last-1)) & S(S'First));
      end if;
   end Backward;

   W: String_Vectors.Vec := String_Vectors.Read(Ada.Command_Line.Argument(1));

   Semi_Counter: Natural := 0;

begin
   for I in W.First_Index .. W.Last_Index loop
      if W.Element(I) /= Backward(W.Element(I)) and then
        W.Is_In(Backward(W.Element(I)), W.First_Index, I) then
         Semi_Counter := Semi_Counter + 1;
         if Semi_Counter <= 5 then
            Ada.Text_IO.Put_Line(W.Element(I) & " - " & Backward(W.Element(I)));
         end if;
      end if;
   end loop;
   Ada.Text_IO.New_Line;
   Ada.Text_IO.Put("pairs found:" & Integer'Image(Semi_Counter));
end Semordnilap;
Output:
>./semordnilap unixdict.txt 
ca - ac
dab - bad
diva - avid
dna - and
drab - bard

pairs found: 158

Aime

integer p, z;
record r;
file f;
text s, t;

f.affix("unixdict.txt");

p = 0;

while (f.line(s) != -1) {
    if (r_o_integer(z, r, t = b_reverse(s))) {
        p += 1;
        if (p <= 5) {
            o_(s, " ", t, "\n");
        }
    }

    r[s] = 0;
}

o_form("Semordnilap pairs: ~\n", p);
Output:
ca ac
dab bad
diva avid
dna and
drab bard
Semordnilap pairs: 158

ALGOL 68

Works with: ALGOL 68G version Any - tested with release 2.8.3.win32

The Algol 68 G "read" PRAGMA is used to include the associative array code from Associative_array/Iteration.

# find the semordnilaps in a list of words                           #
# use the associative array in the Associate array/iteration task    #
PR read "aArray.a68" PR

# returns text with the characters reversed                          #
OP REVERSE = ( STRING text )STRING:
     BEGIN
         STRING reversed  := text;
         INT    start pos := LWB text;
         FOR end pos FROM UPB reversed BY -1 TO LWB reversed
         DO
             reversed[ end pos ] := text[ start pos ];
             start pos +:= 1
         OD;
         reversed
    END # REVERSE # ;

# read the list of words and store the words in an associative array #
# check for semordnilaps                                             #
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
                       );
    REF AARRAY words := INIT LOC AARRAY;
    STRING word;
    INT    semordnilap count := 0;
    WHILE NOT at eof
    DO
        STRING word;
        get( input file, ( word, newline ) );
        STRING reversed word = REVERSE word;
        IF ( words // reversed word ) = ""
        THEN
            # the reversed word isn't in the array                   #
            words // word := reversed word
        ELSE
            # we already have this reversed - we have a semordnilap  #
            semordnilap count +:= 1;
            IF semordnilap count <= 5
            THEN
                print( ( reversed word, " & ", word, newline ) )
           FI
        FI
    OD;
    close( input file );
    print( ( whole( semordnilap count, 0 ), " semordnilaps found", newline ) ) 
FI
Output:
ac & ca
bad & dab
avid & diva
and & dna
bard & drab
158 semordnilaps found

Arturo

words: read.lines "http://wiki.puzzlers.org/pub/wordlists/unixdict.txt"
pairs: []
loop words 'wrd [ 
    if and? contains? words reverse wrd 
            wrd <> reverse wrd [
        'pairs ++ @[@[wrd reverse wrd]]
        print [wrd "-" reverse wrd]
    ]
]

unique 'pairs

print map 1..5 => [sample pairs]

AutoHotkey

Works with: AutoHotkey_L
S := [], M := []

FileRead, dict, unixdict.txt
Loop, Parse, dict, `n, `r`n
{
	r := Reverse(A_LoopField)
	if (S[r])
		M.Insert(r " / " A_LoopField)
	else
		S[A_LoopField] := 1
}

Loop, 5
	Out .= "`t" M[A_Index] "`n"

MsgBox, % "5 Examples:`n" Out "`nTotal Pairs:`n`t" M.MaxIndex()

Reverse(s) {
	Loop, Parse, s
		r := A_LoopField . r
	return r
}
Output:
5 Examples:
	ac / ca
	bad / dab
	avid / diva
	and / dna
	bard / drab

Total Pairs:
	158

AWK

# syntax: GAWK -f SEMORDNILAP.AWK unixdict.txt
{ arr[$0]++ }
END {
    PROCINFO["sorted_in"] = "@ind_str_asc"
    for (word in arr) {
      rword = ""
      for (j=length(word); j>0; j--) {
        rword = rword substr(word,j,1)
      }
      if (word == rword) { continue } # palindrome
      if (rword in arr) {
        if (word in shown || rword in shown) { continue }
        shown[word]++
        shown[rword]++
        if (n++ < 5) { printf("%s %s\n",word,rword) }
      }
    }
    printf("%d words\n",n)
    exit(0)
}
Output:
able elba
abut tuba
ac ca
ah ha
al la
158 words

BBC BASIC

      INSTALL @lib$+"SORTLIB"
      Sort% = FN_sortinit(0,0)
      
      DIM dict$(26000*2)
      
      REM Load the dictionary, eliminating palindromes:
      dict% = OPENIN("C:\unixdict.txt")
      IF dict%=0 ERROR 100, "No dictionary file"
      index% = 0
      REPEAT
        A$ = GET$#dict%
        B$ = FNreverse(A$)
        IF A$<>B$ THEN
          dict$(index%) = A$
          dict$(index%+1) = B$
          index% += 2
        ENDIF
      UNTIL EOF#dict%
      CLOSE #dict%
      Total% = index%
      
      REM Sort the dictionary:
      C% = Total%
      CALL Sort%, dict$(0)
      
      REM Find semordnilaps:
      pairs% = 0
      examples% = 0
      FOR index% = 0 TO Total%-2
        IF dict$(index%)=dict$(index%+1) THEN
          IF examples%<5 IF LEN(dict$(index%))>4 THEN
            PRINT dict$(index%) " " FNreverse(dict$(index%))
            examples% += 1
          ENDIF
          pairs% += 1
        ENDIF
      NEXT
      
      PRINT "Total number of unique pairs = "; pairs%/2
      END
      
      DEF FNreverse(A$)
      LOCAL I%, L%, P%
      IF A$="" THEN =""
      L% = LENA$ - 1
      P% = !^A$
      FOR I% = 0 TO L% DIV 2
        SWAP P%?I%, L%?(P%-I%)
      NEXT
      = A$
Output:
damon nomad
kramer remark
lager regal
leper repel
lever revel
Total number of unique pairs = 158

Bracmat

( get'("unixdict.txt",STR):?dict
& new$hash:?H
& 0:?p
& ( @( !dict
     :   ?
         ( [!p ?w \n [?p ?
         & (H..insert)$(!w.rev$!w)
         & ~
         )
     )
  |   0:?N
    &   (H..forall)
      $ (
        =
          .     !arg:(?a.?b)
              & !a:<!b
              & (H..find)$!b
              & !N+1:?N:<6
              & out$(!a !b)
            |
        )
    & out$(semordnilap !N dnuoF)
  )
);
Output:
tv vt
ir ri
ac ca
eh he
ku uk
semordnilap 158 dnuoF

C

#include <stdio.h>
#include <stdlib.h>
#include <alloca.h> /* stdlib.h might not have obliged. */
#include <string.h>

static void reverse(char *s, int len)
{
    int i, j;
    char tmp;

    for (i = 0, j = len - 1; i < len / 2; ++i, --j)
        tmp = s[i], s[i] = s[j], s[j] = tmp;
}

/* Wrap strcmp() for qsort(). */
static int strsort(const void *s1, const void *s2)
{
    return strcmp(*(char *const *) s1, *(char *const *) s2);
}

int main(void)
{
    int i, c, ct = 0, len, sem = 0;
    char **words, **drows, tmp[24];
    FILE *dict = fopen("unixdict.txt", "r");

    /* Determine word count. */
    while ((c = fgetc(dict)) != EOF)
        ct += c == '\n';
    rewind(dict);

    /* Using alloca() is generally discouraged, but we're not doing
     * anything too fancy and the memory gains are significant. */
    words = alloca(ct * sizeof words);
    drows = alloca(ct * sizeof drows);

    for (i = 0; fscanf(dict, "%s%n", tmp, &len) != EOF; ++i) {
        /* Use just enough memory to store the next word. */
        strcpy(words[i] = alloca(len), tmp);

        /* Store it again, then reverse it. */
        strcpy(drows[i] = alloca(len), tmp);
        reverse(drows[i], len - 1);
    }

    fclose(dict);
    qsort(drows, ct, sizeof drows, strsort);

    /* Walk both sorted lists, checking only the words which could
     * possibly be a semordnilap pair for the current reversed word. */
    for (c = i = 0; i < ct; ++i) {
        while (strcmp(drows[i], words[c]) > 0 && c < ct - 1)
            c++;
        /* We found a semordnilap. */
        if (!strcmp(drows[i], words[c])) {
            strcpy(tmp, drows[i]);
            reverse(tmp, strlen(tmp));
            /* Unless it was a palindrome. */
            if (strcmp(drows[i], tmp) > 0 && sem++ < 5)
                printf("%s\t%s\n", drows[i], tmp);
        }
    }

    printf("Semordnilap pairs: %d\n", sem);
    return 0;
}
Output:
ca	ac
dab	bad
diva	avid
dna	and
drab	bard
Semordnilap pairs: 158

C#

using System;
using System.Net;
using System.Collections.Generic;
using System.Linq;
using System.IO;

public class Semordnilap
{
    public static void Main() {
        var results = FindSemordnilaps("http://www.puzzlers.org/pub/wordlists/unixdict.txt").ToList();
        Console.WriteLine(results.Count);
        var random = new Random();
        Console.WriteLine("5 random results:");
        foreach (string s in results.OrderBy(_ => random.Next()).Distinct().Take(5)) Console.WriteLine(s + " " + Reversed(s));
    }
    
    private static IEnumerable<string> FindSemordnilaps(string url) {
        var found = new HashSet<string>();
        foreach (string line in GetLines(url)) {
            string reversed = Reversed(line);
            //Not taking advantage of the fact the input file is sorted
            if (line.CompareTo(reversed) != 0) {
                if (found.Remove(reversed)) yield return reversed;
                else found.Add(line);
            }
        }
    }
    
    private static IEnumerable<string> GetLines(string url) {
        WebRequest request = WebRequest.Create(url);
        using (var reader = new StreamReader(request.GetResponse().GetResponseStream(), true)) {
            while (!reader.EndOfStream) {
                yield return reader.ReadLine();
            }
        }
    }
    
    private static string Reversed(string value) => new string(value.Reverse().ToArray());
}
Output:
158
5 random results:
keep peek
lever revel
ix xi
avid diva
gar rag

C++

#include <fstream>
#include <iostream>
#include <set>
#include <string>

int main() {
    std::ifstream input("unixdict.txt");
    if (input) {
        std::set<std::string> words; // previous words
        std::string word; // current word
        size_t count = 0; // pair count

        while (input >> word) {
            std::string drow(word.rbegin(), word.rend()); // reverse
            if (words.find(drow) == words.end()) {
                // pair not found
                words.insert(word);
            } else {
                // pair found
                if (count++ < 5)
                    std::cout << word << ' ' << drow << '\n';
            }
        }
        std::cout << "\nSemordnilap pairs: " << count << '\n';
        return 0;
    } else
        return 1; // couldn't open input file
}
Output:
ca ac
dab bad
diva avid
dna and
drab bard

Semordnilap pairs: 158

Clojure

(ns rosettacode.semordnilaps
  (:require [clojure.string  :as str])
            [clojure.java.io :as io ]))

(def dict-file
  (or (first *command-line-args*) "unixdict.txt"))

(def dict (-> dict-file io/reader line-seq set))

(defn semordnilap? [word] 
  (let [rev (str/reverse word)] 
    (and (not= word rev) (dict rev))))

(def semordnilaps
  (->> dict
       (filter semordnilap?)
       (map #([% (str/reverse %)]))
       (filter (fn [[x y]] (<= (compare x y) 0)))))

(printf "There are %d semordnilaps in %s.  Here are 5:\n" 
  (count semordnilaps) 
  dict-file)

(dorun (->> semordnilaps shuffle (take 5) sort (map println)))
Output:
There are 158 semordnilaps in unixdict.txt.  Here are 5:
[bog gob]
[gnaw wang]
[it ti]
[los sol]
[mot tom]

Common Lisp

(defun semordnilaps (word-list)
  (let ((word-map (make-hash-table :test 'equal)))
    (loop for word in word-list do
         (setf (gethash word word-map) t))
    (loop for word in word-list
       for rword = (reverse word)
       when (and (string< word rword) (gethash rword word-map))
       collect (cons word rword))))

(defun main ()
  (let ((words
         (semordnilaps
          (with-open-file (s "unixdict.txt")
            (loop for line = (read-line s nil nil)
               until (null line)
               collect (string-right-trim #(#\space #\return #\newline) line))))))
    (format t "Found pairs: ~D" (length words))
    (loop for x from 1 to 5
       for word in words
       do (print word)))
  (values))
Output:
* (main)
Found pairs: 158
("able" . "elba")
("abut" . "tuba")
("ac" . "ca")
("ah" . "ha")
("al" . "la")

Crystal

require "set"

UNIXDICT = File.read("unixdict.txt").lines

def word?(word : String)
  UNIXDICT.includes?(word)
end

# is it a word and is it a word backwards?
semordnilap = UNIXDICT.select { |word| word?(word) && word?(word.reverse) }

# consolidate pairs like [bad, dab] == [dab, bad]
final_results = semordnilap.map { |word| [word, word.reverse].to_set }.uniq

# sets of N=1 mean the word is identical backwards
# print out the size, and 5 random pairs
puts final_results.size, final_results.sample(5)
Output:
246
[Set{"s's"}, Set{"eire", "erie"}, Set{"bag", "gab"}, Set{"mat", "tam"}, Set{"gel", "leg"}]

D

Simple Imperative Version

void main() {
    import std.stdio, std.file, std.string, std.algorithm;

    bool[string] seenWords;
    size_t pairCount = 0;

    foreach (const word; "unixdict.txt".readText.toLower.splitter) {
        //const drow = word.dup.reverse();
        auto drow = word.dup;
        drow.reverse();
        if (drow in seenWords) {
            if (pairCount++ < 5)
                writeln(word, " ", drow);
        } else
            seenWords[word] = true;
    }

    writeln("\nSemordnilap pairs: ", pairCount);
}
Output:
ca ac
dab bad
diva avid
dna and
drab bard

Semordnilap pairs: 158

A More Functional Version

void main() {
    import std.stdio, std.file, std.algorithm, std.string, std.range;

    auto words = "unixdict.txt".readText.split.zip(0.repeat).assocArray;
    auto pairs = zip(words.byKey, words.byKey.map!(w => w.dup.reverse))
                 .filter!(wr => wr[0] < wr[1] && wr[1] in words)
                 .zip(0.repeat).assocArray;
    writeln(pairs.length, "\n", pairs.byKey.take(5));
}
Output:
158
[Tuple!(string, char[])("bag", "gab"), Tuple!(string, char[])("pat", "tap"), Tuple!(string, char[])("avis", "siva"), Tuple!(string, char[])("haw", "wah"), Tuple!(string, char[])("rot", "tor")]

Delphi

program Semordnilap;

{$APPTYPE CONSOLE}
{$R *.res}

uses
  System.SysUtils,
  System.Classes,
  System.StrUtils,
  System.Diagnostics;

function Sort(s: string): string;
var
  c: Char;
  i, j, aLength: Integer;
begin
  aLength := s.Length;

  if aLength = 0 then
    exit('');

  Result := s;

  for i := 1 to aLength - 1 do
    for j := i + 1 to aLength do
      if result[i] > result[j] then
      begin
        c := result[i];
        result[i] := result[j];
        result[j] := c;
      end;
end;

function IsAnagram(s1, s2: string): Boolean;
begin
  if s1.Length <> s2.Length then
    exit(False);

  Result := Sort(s1) = Sort(s2);

end;

function CompareLength(List: TStringList; Index1, Index2: Integer): Integer;
begin
  result := List[Index1].Length - List[Index2].Length;
  if Result = 0 then
    Result := CompareText(Sort(List[Index2]), Sort(List[Index1]));
end;

function IsSemordnilap(word1, word2: string): Boolean;
begin
  Result := SameText(word1, ReverseString(word2));
end;

var
  SemordnilapDict, Dict: TStringList;
  Count, Index, i, j: Integer;
  words: string;
  StopWatch: TStopwatch;

begin
  Randomize;
  StopWatch := TStopwatch.Create;
  StopWatch.Start;

  Dict := TStringList.Create();
  Dict.LoadFromFile('unixdict.txt');

  SemordnilapDict := TStringList.Create;

  Dict.CustomSort(CompareLength);

  Index := Dict.Count - 1;
  words := '';
  Count := 1;

  while Index - Count >= 0 do
  begin
    if IsAnagram(Dict[Index], Dict[Index - Count]) then
    begin
      if IsSemordnilap(Dict[Index], Dict[Index - Count]) then
      begin
        words := Dict[Index] + ' - ' + Dict[Index - Count];
        SemordnilapDict.Add(words);
      end;
      Inc(Count);
    end
    else
    begin
      if Count > 2 then
        for i := 1 to Count - 2 do
          for j := i + 1 to Count - 1 do
          begin
            if IsSemordnilap(Dict[Index - i], Dict[Index - j]) then
            begin
              words := Dict[Index - i] + ' - ' + Dict[Index - j];
              SemordnilapDict.Add(words);
            end;
          end;

      Dec(Index, Count);
      Count := 1;
    end;
  end;

  StopWatch.Stop;

  Writeln(Format('Time pass: %d ms [i7-4500U Windows 7]', [StopWatch.ElapsedMilliseconds]));

  writeln(#10'Semordnilap found: ', SemordnilapDict.Count);
  writeln(#10'Five random samples:'#10);

  for Index := 0 to 4 do
    writeln('  ', SemordnilapDict[Random(SemordnilapDict.Count)]);

  SemordnilapDict.SaveToFile('Semordnilap.txt');
  SemordnilapDict.Free;
  Dict.Free;
  Readln;
end.
Output:
Time pass: 558 ms [i7-4500U Windows 7]

Semordnilap found: 158

Five random samples:

  on - no
  me - em
  peek - keep
  ton - not
  viva - aviv

EchoLisp

We use the words library, and the french dictionary delivered with EchoLisp.

(lib 'struct)
(lib 'sql)
(lib 'words)

(lib 'dico.fr.no-accent) ;; load dictionary
(string-delimiter "")

;; check reverse r of w is a word
;; take only one pair : r < w
(define (semordnilap? w)
	    (define r (list->string (reverse (string->list w))))
	    (and (word? r) (string<? r w)))
	    
;; to get longest first
(define (string-sort a b) (> (string-length a) (string-length b)))
	    
(define (task)
;; select unique words into the list 'mots'
    (define mots (make-set (words-select #:any null 999999)))
    (define semordnilap 
	    (list-sort string-sort (for/list ((w mots))
	    #:when (semordnilap? w)
	    w )))
   (writeln 'pairs ' (length semordnilap))
   (writeln 'longest ' (take semordnilap 5)))
Output:
(task)
    pairs       →     345    
    longest     →     (rengager tresser strasse reveler retrace)    

Eiffel

First the programm reads the wordlist into an array. Then it mirrors each word and searchs for it across the array using binary search.

class
	SEMORDNILAP

create
	make

feature

	make
			--Semordnilaps in 'solution'.
		local
			count, i, middle, upper, lower: INTEGER
			reverse: STRING
		do
			read_wordlist
			create solution.make_empty
			from
				i := 1
			until
				i > word_array.count
			loop
				word_array [i].mirror
				reverse := word_array [i]
				from
					lower := i + 1
					upper := word_array.count
				until
					lower >= upper
				loop
					middle := (upper - lower) // 2 + lower
					if reverse.same_string (word_array [middle]) then
						count := count + 1
						upper := 0
						lower := 1
						solution.force (word_array [i], count)
					elseif reverse.is_less (word_array [middle]) then
						upper := middle - 1
					else
						lower := middle + 1
					end
				end
				if lower < word_array.count and then reverse.same_string (word_array [lower]) then
					count := count + 1
					upper := 0
					lower := 1
					solution.force (word_array [i], count)
				end
				i := i + 1
			end
		end

	solution: ARRAY [STRING]

	original_list: STRING = "unixdict.txt"


feature {NONE}

	read_wordlist
			-- Preprocessed word_array for finding Semordnilaps.
		local
			l_file: PLAIN_TEXT_FILE
			wordlist: LIST [STRING]
		do
			create l_file.make_open_read_write (original_list)
			l_file.read_stream (l_file.count)
			wordlist := l_file.last_string.split ('%N')
			l_file.close
			create word_array.make_empty
			across
				1 |..| wordlist.count as i
			loop
				word_array.force (wordlist.at (i.item), i.item)
			end
		end

	word_array: ARRAY [STRING]

end

Test:

class
	APPLICATION

create
	make

feature

	make
		local
			test: ARRAY [STRING]
			s: STRING
		do
			create se.make
			test := se.solution
			create sort.sort (test)
			across
				test.subarray (1, 5) as t
			loop
				s := t.item
				io.put_string (t.item + "%T")
				s.mirror
				io.put_string (s)
				io.new_line
			end
			io.put_string ("Total number of semordnilaps: ")
			io.put_integer (test.count)
		end

	se: SEMORDNILAP

	sort: MERGE_SORT [STRING]

end
Output:
ca      ac
dab     bad
diva    avid
dna     and
drab    bard
Total number of semordnilaps: 158

Elixir

words = File.stream!("unixdict.txt")
        |> Enum.map(&String.strip/1)
        |> Enum.group_by(&min(&1, String.reverse &1))
        |> Map.values
        |> Enum.filter(&(length &1) == 2)
IO.puts "Semordnilap pair: #{length(words)}"
IO.inspect Enum.take(words,5)
Output:
Semordnilap pair: 158
[["dab", "bad"], ["drib", "bird"], ["marc", "cram"], ["soma", "amos"],
 ["tab", "bat"]]

Erlang

Translation of: Clojure
#!/usr/bin/env escript
main([]) -> main(["unixdict.txt"]);

main([DictFile]) -> 
  Dict = sets:from_list(read_lines(DictFile)),
  Semordnilaps = 
    lists:filter(fun([W,R]) -> W < R end,
      lists:map(fun(W) -> [W, lists:reverse(W)] end, 
        semordnilaps(Dict))),
  io:fwrite("There are ~b semordnilaps in ~s~n",
            [length(Semordnilaps), DictFile]),
  lists:map(fun([W,R]) -> io:fwrite("~s/~s~n", [W, R]) end,
            lists:sort(lists:sublist(shuffle(Semordnilaps),1,5))).

read_lines(Filename) when is_list(Filename) ->
  { ok, File } = file:open(Filename, [read]),
  read_lines(File);

read_lines(File) when is_pid(File) ->
   case file:read_line(File) of
    {ok, Data} -> [chop(Data) | read_lines(File)];
    eof        -> []
   end.

is_semordnilap(Word, Dict) -> 
  Rev = lists:reverse(Word),
  sets:is_element(Word, Dict) and sets:is_element(Rev, Dict).

semordnilaps(Dict) -> 
  lists:filter(fun(W) -> is_semordnilap(W, Dict) end, sets:to_list(Dict)).

shuffle(List) -> 
  [X||{_,X} <- lists:sort([ {random:uniform(), N} || N <- List])].

chop(L) -> [_|T] = lists:reverse(L), lists:reverse(T).
Output:
There are 158 semordnilaps in unixdict.txt
aryl/lyra
caw/wac
cram/marc
dine/enid
dual/laud

F#

Using a mutable dictionary.

open System

let seen = new System.Collections.Generic.Dictionary<string,bool>()

let lines = System.IO.File.ReadLines("unixdict.txt")

let sems = seq {
    for word in lines do
        let drow = new String(Array.rev(word.ToCharArray()))
        if fst(seen.TryGetValue(drow)) then yield (drow, word)
        seen.[drow] <- true
        seen.[word] <- true
}

let s = Seq.toList sems
printfn "%d" s.Length
for i in 0 .. 4 do printfn "%A" s.[i]
Output:
158
("ac", "ca")
("bad", "dab")
("avid", "diva")
("and", "dna")
("bard", "drab")

Factor

USING: assocs combinators.short-circuit formatting
io.encodings.utf8 io.files kernel literals locals make
prettyprint random sequences ;
IN: rosetta-code.semordnilap

CONSTANT: words $[ "unixdict.txt" utf8 file-lines ]

: semordnilap? ( str1 str2 -- ? )
    { [ = not ] [ nip words member? ] } 2&& ;

[
    [let
        V{ } clone :> seen words
        [
            dup reverse 2dup
            { [ semordnilap? ] [ drop seen member? not ] } 2&&
            [ 2dup [ seen push ] bi@ ,, ] [ 2drop ] if 
        ] each
    ]
] H{ } make >alist
[ length "%d semordnilap pairs.\n" printf ] [ 5 sample . ] bi
Output:
158 semordnilap pairs.
{
    { "pan" "nap" }
    { "lac" "cal" }
    { "tang" "gnat" }
    { "wolf" "flow" }
    { "mac" "cam" }
}

Forth

This code uses a Forth wordlist to contain the dictionary, and uses the Forth-2012 TRAVERSE-WORDLIST to walk through it (a simpler way would be to check for the presence of the reversed word when putting the word into the wordlist).

One interesting issue is how I get each pair only once and exclude palindromes: I accept only pairs where nt<nt2. A type checking bigot will likely argue that nts should not be compared with <, because they are opaque data types. But their implementation does not matter for this check: Whatever bit patterns these two nts get, either it's the same nt, then nt<nt2 will return false, as desired; and if they are different, exactly one of nt<nt2 and nt2<nt will return true.

The code uses two Gforth-specific words: EXECUTE-PARSING (implementable in standard Forth, but not easy) for allowing to provide the name of the defined word on the stack; and FIND-NAME-IN to look up the reversed word (could be replaced with a use of the standard SEARCH-WORDLIST, but the code would become a little more complicated).

wordlist constant dict

: load-dict ( c-addr u -- )
    r/o open-file throw >r
    begin
	pad 1024 r@ read-line throw while
	    pad swap ['] create execute-parsing
    repeat
    drop r> close-file throw ;

: xreverse {: c-addr u -- c-addr2 u :}
    u allocate throw u + c-addr swap over u + >r begin ( from to r:end)
	over r@ u< while
	    over r@ over - x-size dup >r - 2dup r@ cmove
	    swap r> + swap repeat
    r> drop nip u ;

: .example ( c-addr u u1 -- )
    5 < if
	cr 2dup type space 2dup xreverse 2dup type drop free throw then
    2drop ;

: nt-semicheck ( u1 nt -- u2 f )
    dup >r name>string xreverse 2dup dict find-name-in dup if ( u1 c-addr u nt2) 
	r@ < if ( u1 c-addr u ) \ count pairs only once and not palindromes
	    2dup 4 pick .example
	    rot 1+ -rot then
    else
	drop then
    drop free throw r> drop true ;

get-current dict set-current s" unixdict.txt" load-dict set-current

0 ' nt-semicheck dict traverse-wordlist cr .
cr bye
Output:
suez zeus
paz zap
way yaw
pay yap
may yam
158 

Fortran

Please read the comments at the beginning of the f90 source to see the compilation instructions and output of 5 random words from a run. Note that program Semordnilap opens the file unixdict.txt . It does not read from stdin, hence the command line redirection from unixdict.txt is irrelevant. I haven't bothered to change it.


!-*- mode: compilation; default-directory: "/tmp/" -*-
!Compilation started at Sun May 19 21:50:08
!
!a=./F && make $a && $a < unixdict.txt
!f95 -Wall -ffree-form F.F -o F
! 5 of          158 semordnilaps            
!yaw                     
!room                    
!xi                      
!tim                     
!nova                    
!
!
!Compilation finished at Sun May 19 21:50:08
!
!
!
!
!                       unixdict.txt information
! wc -l unixdict.txt                                         #--> 25104                    25 thousand entries
! gawk 'length(a)<length($0){a=$0}END{print a}' unixdict.txt #--> electroencephalography   longest word has 22 characters
! gawk '/[A-Z]/{++a}END{print a}' unixdict.txt               #--> <empty>                  the dictionary is lower case
! sort unixdict.txt | cmp - unixdict.txt                     #--> - unixdict.txt differ: byte 45, line 12
!                                                                                          the dictionary is unsorted
!     mmmmm the dictionary is sorted, according to subroutine bs.  There's something about the ampersands within unixdict.txt I misunderstand.

program Semordnilap
  implicit none
  integer :: i, ios, words, swords
  character(len=24), dimension(32768) :: dictionary, backword
  real, dimension(5) :: harvest
  ! read the dictionary
  open(7,file='unixdict.txt')
  do words = 1, 32768
    read(7, '(a)', iostat = ios) dictionary(words)
    if (ios .ne. 0) exit
  enddo
  close(7)
  if (iachar(dictionary(words)(1:1)) .eq. 0) words = words-1
  ! sort the dictionary
  call bs(dictionary, words)
  !do i = 1, words
  !  write(6,*) dictionary(i)(1:len_trim(dictionary(i))) ! with which we determine the dictionary was ordered
  !enddo
  swords = 0
  do i = 1, words
    call reverse(dictionary(i), backword(swords+1))
    if ((binary_search(dictionary, words, backword(swords+1))) &      !     the reversed word is in the dictionary
      .and. (.not. binary_search(backword, swords, dictionary(i))) &  ! and it's new
      .and. (dictionary(i) .ne. backword(swords+1))) then             ! and it's not a palindrome
      swords = swords + 1
      call bs(backword, swords)
    endif
  enddo
  call random_number(harvest)
  call reverse('spalindromes', backword(swords+1))
  write(6, *) '5 of ', swords, backword(swords+1)
  write(6,'(5(a/))') (backword(1+int(harvest(i)*(swords-2))), i=1,5)

contains

  subroutine reverse(inp, outp)
    character(len=*), intent(in) :: inp
    character(len=*), intent(inout) :: outp
    integer :: k, L
    L = len_trim(inp)
    do k = 1, L
      outp(L+1-k:L+1-k) = inp(k:k)
    enddo
    do k = L+1, len(outp)
      outp(k:k) = ' '
    enddo
  end subroutine reverse

  subroutine bs(a, n) ! ok, despite having claimed that bubble sort should be unceremoniously buried, I'll use it anyway because I expect the dictionary is nearly ordered.  It's also not a terrible sort for less than 5 items.
    ! Please note, I tested bs using unixdict.txt randomized with sort --random .
    character(len=*),dimension(*),intent(inout) :: a
    integer, intent(in) :: n
    integer :: i, j, k
    logical :: done
    character(len=1) :: t
    do i=n-1, 1, -1
      done = .true.
      do j=1, i
        if (a(j+1) .lt. a(j)) then
          done = .false.
          do k = 1, max(len_trim(a(j+1)), len_trim(a(j)))
            t = a(j+1)(k:k)
            a(j+1)(k:k) = a(j)(k:k)
            a(j)(k:k) = t(1:1)
          enddo
        endif
      enddo
      if (done) return
    enddo
  end subroutine bs

  logical function binary_search(source, n, target)
    character(len=*),dimension(*),intent(in) :: source
    character(len=*),intent(in) :: target
    integer, intent(in) :: n
    integer :: a,m,z
    a = 1
    z = n
    do while (a .lt. z)
      m = a + (z - a) / 2
      if (target .lt. source(m)) then
        z = m-1
      else
        if (m .eq. a) exit
        a = m
      endif
    enddo
    binary_search = (target .eq. source(a)) .or. (target .eq. source(z))
  end function binary_search

end program Semordnilap

FreeBASIC

' version 20-06-2015
' compile with: fbc -s console

Function reverse(norm As String) As String

    Dim As String rev
    Dim As Integer i, l = Len(norm) -1

    rev = norm
    For i = 0 To l
        rev[l-i] = norm[i]
    Next

    Return rev

End Function

' ------=< MAIN >=------

Dim As Integer i, j, count, amount, ff = FreeFile
Dim As String in_str, rev, big = " "  ' big needs to start with a space
Dim As String norm(27000), result(270, 2)

Print
Print "Start reading unixdict.txt";

Open "unixdict.txt" For Input As #ff

While Not Eof(ff)                  ' read to end of file

    Line Input #ff, in_str         ' get line = word
    in_str = Trim(in_str)          ' we don't want spaces
    If Len(in_str) > 1 Then        ' if length > 1 then reverse
        rev = reverse(in_str)
        If in_str <> rev Then      ' if in_str is not a palingdrome
            count = count + 1      ' increase counter
            norm(count) = in_str   ' store in the array
            big = big + rev + " "  ' create big string with reversed words
        End If
    End If

Wend

Close #ff

Print " ... Done"
Print : Print "Start looking for semordnilap"

For i = 1 To count
    For j = 1 To amount ' check to avoid the double
        If result(j, 2) = norm(i) Then Continue For, For
    Next
    j = InStr(big, " " + norm(i) + " ")
    If j <> 0 Then                          ' found one
        amount = amount + 1                 ' increase counter
        result(amount,1) = norm(i)          ' store normal word
        result(amount,2) = reverse(norm(i)) ' store reverse word
    End If
Next

Print : Print "Found"; amount; " unique semordnilap pairs"
Print : Print "Display 5 semordnilap pairs"
Print

count = 0
For i = 1 To amount
    If Len(result(i,1)) >= 5 Then
        count = count + 1
        Print result(i, 1), result(i, 2)
        If count >= 5 Then Exit For
    EndIf
Next

Print

' empty keyboard buffer 
While InKey <> "" : Wend
Print : Print "Hit any key to end program"
Sleep
End
Output:
Start reading unixdict.txt ... Done

Start looking for semordnilap

Found 158 unique semordnilap pairs

Display 5 semordnilap pairs

damon         nomad
kramer        remark
lager         regal
leper         repel
lever         revel

Go

package main

import (
    "fmt"
    "io/ioutil"
    "log"
    "strings"
)

func main() {
    // read file into memory as one big block
    data, err := ioutil.ReadFile("unixdict.txt")
    if err != nil {
        log.Fatal(err)
    }
    // copy the block, split it up into words
    words := strings.Split(string(data), "\n")
    // optional, free the first block for garbage collection
    data = nil
    // put words in a map, also determine length of longest word
    m := make(map[string]bool)
    longest := 0
    for _, w := range words {
        m[string(w)] = true
        if len(w) > longest {
            longest = len(w)
        }
    }
    // allocate a buffer for reversing words
    r := make([]byte, longest)
    // iterate over word list
    sem := 0
    var five []string
    for _, w := range words {
        // first, delete from map.  this prevents a palindrome from matching
        // itself, and also prevents it's reversal from matching later.
        delete(m, w)
        // use buffer to reverse word
        last := len(w) - 1
        for i := 0; i < len(w); i++ {
            r[i] = w[last-i]
        }
        rs := string(r[:len(w)])
        // see if reversed word is in map, accumulate results
        if m[rs] {
            sem++
            if len(five) < 5 {
                five = append(five, w+"/"+rs)
            }
        }
    }
    // print results
    fmt.Println(sem, "pairs")
    fmt.Println("examples:")
    for _, e := range five {
        fmt.Println("  ", e)
    }
}
Output:
158 pairs
examples:
   able/elba
   abut/tuba
   ac/ca
   ah/ha
   al/la

Groovy

def semordnilapWords(source) {
    def words = [] as Set
    def semordnilaps = []
    source.eachLine { word ->
        if (words.contains(word.reverse())) semordnilaps << word
        words << word
    }
    semordnilaps
}

Test Code

def semordnilaps = semordnilapWords(new URL('http://www.puzzlers.org/pub/wordlists/unixdict.txt'))
println "Found ${semordnilaps.size()} semordnilap words"
semordnilaps[0..<5].each { println "$it -> ${it.reverse()}" }
Output:
Found 158 semordnilap words
ca -> ac
dab -> bad
diva -> avid
dna -> and
drab -> bard

Haskell

import qualified Data.Set as S

semordnilaps
  :: (Ord a, Foldable t)
  => t [a] -> [[a]]
semordnilaps =
  let f x (s, w)
        | S.member (reverse x) s = (s, x : w)
        | otherwise = (S.insert x s, w)
  in snd . foldr f (S.empty, [])

main :: IO ()
main = do
  s <- readFile "unixdict.txt"
  let l = semordnilaps (lines s)
  print $ length l
  mapM_ (print . ((,) <*> reverse)) $ take 5 (filter ((4 <) . length) l)
Output:
158
("damon","nomad")
("kramer","remark")
("lager","regal")
("leper","repel")
("lever","revel")

Icon and Unicon

The following solution works in both Icon and Unicon:

procedure main(a)
    words := set()
    found := 0
    every word := map(!&input) do {
        if member(words, reverse(word)) then {
            if (found +:= 1) <= 5 then write("\t",reverse(word),"/",word)
            }
        else insert(words, word)
        }
    write("\nFound ",found," semordnilap words")
end

Sample run with unixdict:

->smp <unixdict.txt
        ac/ca
        bad/dab
        avid/diva
        and/dna
        bard/drab

Found 158 semordnilap words
->

J

We find all semordnilaps by filtering only words which, when reversed, are a member of the set of dictionary words and are not palindromes. We then find only unique semordnilaps by pairing them with their reversed instance, sorting each pair, and eliminating duplicates pairs:

   isSemordnilap=: |.&.> (~: *. e.) ]
   unixdict=: <;._2 freads 'unixdict.txt'
   #semordnilaps=: ~. /:~"1 (,. |.&.>) (#~ isSemordnilap) unixdict
158

We see that there are 158 semordnilaps.

Here's 5 of them, picked arbitrarily:

   (5?.158) { semordnilaps
┌────┬────┐
│kay │yak │
├────┼────┤
│nat │tan │
├────┼────┤
│avis│siva│
├────┼────┤
│flow│wolf│
├────┼────┤
│caw │wac │
└────┴────┘

Java

Translation of: D
Works with: Java version 7+
import java.nio.file.*;
import java.util.*;

public class Semordnilap {

    public static void main(String[] args) throws Exception {
        List<String> lst = Files.readAllLines(Paths.get("unixdict.txt"));
        Set<String> seen = new HashSet<>();
        int count = 0;
        for (String w : lst) {
            w = w.toLowerCase();
            String r = new StringBuilder(w).reverse().toString();
            if (seen.contains(r)) {
                if (count++ < 5)
                    System.out.printf("%-10s %-10s\n", w, r);
            } else seen.add(w);
        }
        System.out.println("\nSemordnilap pairs found: " + count);
    }
}
ca         ac        
dab        bad       
diva       avid      
dna        and       
drab       bard      

Semordnilap pairs found: 158

JavaScript

Node.js

Translation of: Clojure
#!/usr/bin/env node
var  fs = require('fs');
var sys = require('sys');

var dictFile = process.argv[2] || "unixdict.txt";

var dict = {};
fs.readFileSync(dictFile)
  .toString()
  .split('\n')
  .forEach(function(word) {
    dict[word] = word.split("").reverse().join("");
  });

function isSemordnilap(word) { return dict[dict[word]]; };

var semordnilaps = []
for (var key in dict) {
  if (isSemordnilap(key)) {
    var rev = dict[key];
    if (key < rev) {
      semordnilaps.push([key,rev]) ;
    }
  }
}

var count = semordnilaps.length;
sys.puts("There are " + count + " semordnilaps in " +
         dictFile + ".  Here are 5:" );

var indices=[]
for (var i=0; i<count; ++i) {
  if (Math.random() < 1/Math.ceil(i/5.0)) {
    indices[i%5] = i 
  }
}
indices.sort()
for (var i=0; i<5; ++i) {
  sys.puts(semordnilaps[indices[i]]);
}

Rhino

Works with: Rhino version 1.7
#!/usr/bin/env rhino

importPackage (java.io)

var dictFile = arguments[0] || "unixdict.txt";

var reader = new BufferedReader(new FileReader(dictFile));
var dict = {};
var word;
while (word = reader.readLine()) {
  dict[word] = word.split("").reverse().join("");
}

function isSemordnilap(word) { return dict[dict[word]]; };

var semordnilaps = []
for (var key in dict) {
  if (isSemordnilap(key)) {
    var rev = dict[key];
    if (key < rev) {
      semordnilaps.push([key,rev]) ;
    }
  }
}

var count = semordnilaps.length;
print("There are " + count + " semordnilaps in " +
      dictFile + ".  Here are 5:" );
var indices=[]
for (var i=0; i<count; ++i) {
  if (Math.random() < 1/Math.ceil(i/5.0)) {
     indices[i%5] = i 
  }
}
indices.sort()
for (var i=0; i<5; ++i) {
  print(semordnilaps[indices[i]]);
}
Output:
There are 158 semordnilaps in unixdict.txt.  Here are 5:
loot,tool
ah,ha
dial,laid
dine,enid
haw,wah

macOS JavaScript for Automation

Translation of: Python
(() => {
    'use strict';

    // semordnilap :: [String] -> String
    const semordnilap = xs => {
        const go = ([s, ws], w) =>
            s.has(w.split('').reverse().join('')) ? (
                [s, [w].concat(ws)]
            ) : [s.add(w), ws];
        return xs.reduce(go, [new Set(), []])[1];
    };

    const main = () => {
        
        // xs :: [String]
        const xs = semordnilap(
            lines(readFile('unixdict.txt'))
        );

        console.log(xs.length);
        xs.filter(x => 4 < x.length).forEach(
            x => showLog(...[x, x.split('').reverse().join('')])
        )
    };


    // GENERIC FUNCTIONS ----------------------------

    // lines :: String -> [String]
    const lines = s => s.split(/[\r\n]/);

    // readFile :: FilePath -> IO String
    const readFile = fp => {
        const
            e = $(),
            uw = ObjC.unwrap,
            s = uw(
                $.NSString.stringWithContentsOfFileEncodingError(
                    $(fp)
                    .stringByStandardizingPath,
                    $.NSUTF8StringEncoding,
                    e
                )
            );
        return undefined !== s ? (
            s
        ) : uw(e.localizedDescription);
    };

    // showLog :: a -> IO ()
    const showLog = (...args) =>
        console.log(
            args
            .map(JSON.stringify)
            .join(' -> ')
        );

    // MAIN ---
    return main();
})();
Output:
158
"revel" -> "lever"
"repel" -> "leper"
"remark" -> "kramer"
"regal" -> "lager"
"nomad" -> "damon"

jq

Works with: jq version 1.4

The following program illustrates several points about jq:

  • jq can be used to process text, as well as JSON;
  • for text-processing tasks for which awk is well-suited, jq may be slightly slower;
  • jq objects (i.e. JSON objects) can be used to define time-efficient mappings from strings.

Here are some running times on the same machine:

  • awk program:

user 0m0.134s; sys 0m0.012s

  • time /usr/local/bin/jq -M -s -R -r -f semordnilap.jq unixdict.txt

user 0m0.440s; sys 0m0.010s

# Produce a stream
def report:
  split("\n") as $list
  # construct the dictionary:
  | (reduce $list[] as $entry ({}; . + {($entry): 1})) as $dict
  # construct the list of semordnilaps:
  | $list[]
  | select( (explode|reverse|implode) as $rev
            | (. < $rev and $dict[$rev]) );

[report] | (.[0:5][],  "length = \(length)")
Output:
able
abut
ac
ah
al
length = 158

Julia

raw = readdlm("unixdict.txt",String)[:]
inter = intersect(raw,map(reverse,raw)) #find the matching strings/revstrings
res = String[b == 1 && a != reverse(a) && a < reverse(a) ? a : reverse(a) for a in inter, b in 1:2] #create pairs
res = res[res[:,1] .!= res[:,2],:] #get rid of duplicates, palindromes
julia> length(res[:,1])
158

julia> res[1:5,:]
5x2 String Array:
 "able"  "elba"
 "abut"  "tuba"
 "ac"    "ca"  
 "ah"    "ha"  
 "al"    "la"  

Kotlin

Translation of: D
// version 1.2.0

import java.io.File
 
fun main(args: Array<String>) {
    val words = File("unixdict.txt").readLines().toSet()
    val pairs = words.map { Pair(it, it.reversed()) }
            .filter { it.first < it.second && it.second in words } // avoid dupes+palindromes, find matches
    println("Found ${pairs.size} semordnilap pairs")
    println(pairs.take(5))
}
Output:
Found 158 semordnilap pairs
[(able, elba), (abut, tuba), (ac, ca), (ah, ha), (al, la)]

Ksh

#!/bin/ksh

# Semordnilap

#	# Variables:
#
integer MIN_WORD_LEN=1 TRUE=1 FALSE=0
dict='/home/ostrande/prj/roscode/unixdict.txt'

integer i j=0 k=0
typeset -A word

#	# Functions:
#

#	# Function _flipit(string) - return flipped string
#
function _flipit {
	typeset _buf ; _buf="$1"
	typeset _tmp ; unset _tmp

	for (( _i=$(( ${#_buf}-1 )); _i>=0; _i-- )); do
		_tmp="${_tmp}${_buf:${_i}:1}"
	done

	echo "${_tmp}"
}

#	# Function _isword(word, wordlist) - return 1 if word in wordlist
#
function _isword {
	typeset _word ; _word="$1"
	typeset _wordlist ; nameref _wordlist="$2"

	[[ ${_word} == @(${_wordlist}) ]] && return $TRUE
	return $FALSE
}

 ######
# main #
 ######
#	# Due to the large number of words in unixdist.txt subgroup by 1st letter and length
#	# only accept words containing alpha chars and > 1 chars
#
while read; do
	[[ $REPLY != *+(\W)* ]] && [[ $REPLY != *+(\d)* ]] &&	\
	  (( ${#REPLY} > MIN_WORD_LEN )) && word[${REPLY:0:1}][${#REPLY}]+=( $REPLY )
done < ${dict}

print Examples:
for fl in ${!word[*]}; do				# Over $fl first letter
	for len in ${!word[${fl}][*]}; do	# Over $len word length
		for ((i=0; i<${#word[${fl}][${len}][*]}; i++)); do 
			Word=${word[${fl}][${len}][i]}		# dummy
			Try=$(_flipit ${Word})
			if [[ ${Try} != ${Word} ]]; then	# no palindromes
				unset words
				oldIFS="$IFS" ; IFS='|' ; words=${word[${Try:0:1}][${#Try}][*]} ; IFS="${oldIFS}"
				_isword "${Try}" words
				if (( $? )); then
					if [[ ${Try} != @(${uniq%\|*}) ]]; then
						((++j))
						(( ${#Word} >= 5 )) && (( k<=5 )) && print $((++k)). ${Word} ${Try}
						uniq+="${Try}|${Word}|"
					fi
				fi
			fi
		done
	done
done
echo ; print ${j} pairs found.
Output:

Examples: 1. damon nomad 2. kramer remark 3. lager regal 4. leper repel 5. lever revel

158 pairs found.

Lasso

local(
	words		= string(include_url('http://www.puzzlers.org/pub/wordlists/unixdict.txt')) -> split('\n'),
	semordnilaps	= array,
	found_size,
	example,
	haveexamples	= false,
	examples	= array
)

#words -> removeall('')

with word in #words do {
	local(reversed = string(#word) -> reverse&)
	if(not(#word == #reversed) and not(#semordnilaps >> #word) and not(#semordnilaps >> #reversed) and #words >> #reversed) => {
		#semordnilaps -> insert(#word = #reversed)
	}
}

#found_size = #semordnilaps -> size

while(not(#haveexamples)) => {
	#example = #semordnilaps -> get(integer_random(#found_size, 1))
	not(#examples >> #example -> name) ? #examples -> insert(#example)
	#examples -> size >= 5 ? #haveexamples = true
}
'Total found: '
#found_size
'<br />'
#examples
Output:
Total found: 158
array((dew = wed), (are = era), (den = ned), (oat = tao), (eel = lee))

Liberty BASIC

print "Loading dictionary."
open "unixdict.txt" for input as #1
while not(eof(#1))
   line input #1, a$
   dict$=dict$+" "+a$
wend
close #1

print "Dictionary loaded."
print "Seaching for semordnilaps."

semo$=" "   'string to hold words with semordnilaps

do
    i=i+1
    w$=word$(dict$,i)
    p$=reverseString$(w$)
    if w$<>p$ then
    p$=" "+p$+" "
    if instr(semo$,p$) = 0 then
        if instr(dict$,p$) then
          pairs=pairs+1
          print w$+" /"+p$
          semo$=semo$+w$+p$
        end if
      end if
    end if
    scan
loop until w$=""

print "Total number of unique semordnilap pairs is ";pairs
wait

Function isPalindrome(string$)
    string$ = Lower$(string$)
    reverseString$ = reverseString$(string$)
    If string$ = reverseString$ Then isPalindrome = 1
End Function

Function reverseString$(string$)
    For i = Len(string$) To 1 Step -1
        reverseString$ = reverseString$ + Mid$(string$, i, 1)
    Next i
End Function
Output:
able / elba
leper / repel
lever / revel
moor / room
suez / zeus
tort / trot
Total number of unique semordnilap pairs is 158

Lua

#!/usr/bin/env lua
-- allow dictionary file and sample size to be specified on command line
local dictfile = arg[1] or "unixdict.txt"
local sample_size = arg[2] or 5;

-- read dictionary
local f = assert(io.open(dictfile, "r"))
local dict = {}
for line in f:lines() do
  dict[line] = line:reverse()
end
f:close()

-- find the semordnilaps
local semordnilaps = {}
for fwd, rev in pairs(dict) do
  if dict[rev] and fwd < rev then
    table.insert(semordnilaps, {fwd,rev})
  end
end

-- print the report
print("There are " .. #semordnilaps .. " semordnilaps in " .. dictfile .. ".  Here are " .. sample_size .. ":")

math.randomseed( os.time() )
for i = 1, sample_size do
  local j
  repeat 
    j = math.random(1,#semordnilaps)
  until semordnilaps[j]
  local f, r = unpack(semordnilaps[j])
  semordnilaps[j] = nil
  print(f .. " -> " .. r)
end
Output:
There are 158 semordnilaps in unixdict.txt.  Here are 5:
deer -> reed
rat -> tar
pus -> sup
meet -> teem
bat -> tab

M2000 Interpreter

Module semordnilaps {
      Document d$
      Load.Doc d$, "unixdict.txt"
      Inventory MyDict, Result
      Function s$(a$) {
            m$=a$:k=Len(a$):for i=1 to k {insert i, 1 m$=mid$(a$, k, 1):k--} : =m$
      }
      L=Doc.Par(d$)
      m=Paragraph(d$, 0) 
      If not Forward(d$,m) then exit
      i=1
      While m {
            word$=Paragraph$(d$,(m))
            Print Over $(0, 10), str$(i/L,"##0.00%"), Len(Result) : i++
            If Exist(MyDict, word$) then { if Exist(Result, word$) Then exit
                  Append Result, word$
            } Else.if  len(word$)>1 Then p$=s$(word$):if p$<>word$ Then Append MyDict, p$
      }
      Print
      Print "Semordnilap pairs: ";Len(Result)
      For i=0 to len(Result)-1 step  len(Result) div 5 {
            p$=Eval$(Result, i)
            Print s$(p$);"/";p$
      }
}
semordnilaps
Output:
Semordnilap pairs: 158
ac/ca
nap/pan
cos/soc
loot/tool
way/yaw

Mathematica/Wolfram Language

data = Import["http://www.puzzlers.org/pub/wordlists/unixdict.txt", "List"];
result = DeleteDuplicates[ Select[data, MemberQ[data, StringReverse[#]]
  && # =!= StringReverse[#] &], (# ===StringReverse[#2]) &];
Print[Length[result], Take[result, 5]]
Output:
158 {able,abut,ac,ah,al}

Nanoquery

Translation of: Java
import Nanoquery.IO

def reverse_str(string)
        ret = ""

        for char in list(string).reverse()
                ret += char
        end

        return ret
end

lst = split(new(File).open("rosetta-code/unixdict.txt").readAll(), "\n")
seen = list()
count = 0
for w in lst
        w = lower(w)
        r = reverse_str(w)
        if r in seen
                count += 1
                if count <= 5
                        print format("%-10s %-10s\n", w, r)
                end
        else
                seen.append(w)
        end
end

println "\nSemordnilap pairs found: " + count
Output:
ca         ac
dab        bad
diva       avid
dna        and
drab       bard

Semordnilap pairs found: 158

NetRexx

Translation of: REXX
/* NetRexx */
options replace format comments java crossref symbols nobinary

/* REXX ***************************************************************
* 07.09.2012 Walter Pachl
**********************************************************************/
fid = 'unixdict.txt'                   /* the test dictionary        */
ifi = File(fid)
ifr = BufferedReader(FileReader(ifi))
have = ''                              /* words encountered          */
pi = 0                                 /* number of palindromes      */
loop label j_ forever                  /* as long there is input     */
  line = ifr.readLine                  /* read a line (String)       */
  if line = null then leave j_         /* NULL indicates EOF         */
  w = Rexx(line)                       /* each line contains 1 word  */
  If w > '' Then Do                    /* not a blank line           */
    r = w.reverse                      /* reverse it                 */
    If have[r] > '' Then Do            /* was already encountered    */
      pi = pi + 1                      /* increment number of pal's  */
      If pi <= 5 Then                  /* the first 5 are listed     */
        Say have[r] w
      End
    have[w] = w                        /* remember the word          */
    End
  end j_
ifr.close

Say pi 'words in' fid 'have a palindrome' /* total number found      */
return
Output:
ac ca 
bad dab 
avid diva 
and dna 
bard drab 
158 words in unixdict.txt have a palindrome

NewLisp

;;;	Get the words as a list, splitting at newline
(setq data
	(parse (get-url "http://wiki.puzzlers.org/pub/wordlists/unixdict.txt")
	"\n"))
;
;;;	destructive reverse wrapped into a function
(define (get-reverse x) (reverse x))
;
;;;	stack of the results
(setq res '())
;
;;;	Find the semordlinap and put them on the stack
(dolist (x data)
	(let (y (get-reverse x))
		(if (and
			(member y data)			; reverse is a dictionary word
			(!= x y)			; but not a palindrome
			(not (member y res)))		; not already stacked
				(push x res -1))))
;
;;;	Count results
(println "Found " (length res) " pairs.")
(println)
;;;	Show the longest ones
(dolist (x res)
	(if (> (length x) 4) (println x " -- " (get-reverse x))))
Output:
Found 158 pairs.

damon -- nomad
kramer -- remark
lager -- regal
leper -- repel
lever -- revel

Nim

import strutils, sequtils, sets, algorithm

proc reversed(s: string): string =
  result = newString(s.len)
  for i, c in s:
    result[s.high - i] = c

let
  words = readFile("unixdict.txt").strip().splitLines()
  wordset = words.toHashSet
  revs = words.map(reversed)
var pairs = zip(words, revs).filterIt(it[0] < it[1] and it[1] in wordset)

echo "Total number of semordnilaps: ", pairs.len
pairs = pairs.sortedByIt(it[0].len)
echo pairs[^5..^1]
Output:
Total number of semordnilaps: 158
@[(a: damon, b: nomad), (a: lager, b: regal), (a: leper, b: repel), (a: lever, b: revel), (a: kramer, b: remark)]

OCaml

module StrSet = Set.Make(String)

let str_rev s =
  let len = String.length s in
  let r = Bytes.create len in
  for i = 0 to len - 1 do
    Bytes.set r i s.[len - 1 - i]
  done;
  Bytes.to_string r

let input_line_opt ic =
  try Some (input_line ic)
  with End_of_file -> close_in ic; None

let () =
  let ic = open_in "unixdict.txt" in
  let rec aux set acc =
    match input_line_opt ic with
    | Some word ->
        let rev = str_rev word in
        if StrSet.mem rev set
        then aux set ((word, rev) :: acc)
        else aux (StrSet.add word set) acc
    | None ->
        (acc)
  in
  let pairs = aux StrSet.empty [] in
  let len = List.length pairs in
  Printf.printf "Semordnilap pairs: %d\n" len;
  Random.self_init ();
  for i = 1 to 5 do
    let (word, rev) = List.nth pairs (Random.int len) in
    Printf.printf " %s %s\n" word rev
  done
Output:
Semordnilap pairs: 158
 tar rat
 sera ares
 sub bus
 tic cit
 mid dim

Octave

a = strsplit(fileread("unixdict.txt"), "\n");
a = intersect(a, cellfun(@fliplr, a, "UniformOutput", false));
a = a(arrayfun(@(i) ismember(fliplr(a{i}), a(i+1:length(a))), 1:length(a)));
length(a)
arrayfun(@(i) printf("%s %s\n", a{i}, fliplr(a{i})), 1:5)

Output:

ans =  158

able elba
abut tuba
ac ca
ah ha
al la

Oforth

: semordnilap
| w wr wrds |
   ListBuffer new ->wrds
   ListBuffer new
   File new("unixdict.txt") forEach: w [ 
      wrds include(w reverse dup ->wr) ifTrue: [ [wr, w] over add ]
      w wr < ifTrue: [ wrds add(w) ]
      ] ;
Output:
>semordnilap dup size println left(5) println
158
[[ac, ca], [bad, dab], [avid, diva], [and, dna], [bard, drab]]

Perl

while (<>) {
	chomp;
	my $r = reverse;
	$seen{$r}++ and $c++ < 5 and print "$_ $r\n" or $seen{$_}++;
}

print "$c\n"

Phix

with javascript_semantics
sequence words = unix_dict(), semordnilap={}
for i=1 to length(words) do
    string word = words[i]
    if rfind(reverse(word),words,i-1) then
        semordnilap = append(semordnilap,word)
    end if
end for
printf(1,"%d semordnilap found, the first five are:\n",length(semordnilap))
for i=1 to 5 do
    printf(1,"%s - %s\n",{semordnilap[i],reverse(semordnilap[i])})
end for
Output:
158 semordnilap found, the first five are:
ca - ac
dab - bad
diva - avid
dna - and
drab - bard

Phixmonti

include ..\Utilitys.pmt

( ) ( )

"unixdict.txt" "r" fopen var f

true while
	f fgets
	dup -1 == if
		drop
		f fclose
		false
	else
		-1 del 
		0 put
		true
	endif	
endwhile

len while
	len 1 > if
		pop swap reverse find dup
		if
			extract rot swap 0 put swap
		else
			drop
		endif
		true
	else
		drop false
	endif
endwhile

( 50 54 ) for get dup reverse print " -> " print ? endfor nl
len print " pairs" ?
Output:
dew -> wed
dial -> laid
dim -> mid
dine -> enid
dog -> god

158 pairs

=== Press any key to exit ===

PHP

Translation of: Raku
<?php
// Read dictionary into array
$dictionary = array_fill_keys(file(
    'http://www.puzzlers.org/pub/wordlists/unixdict.txt',
    FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES
), true);
foreach (array_keys($dictionary) as $word) {
    $reversed_word = strrev($word);
    if (isset($dictionary[$reversed_word]) && $word > $reversed_word)
        $words[$word] = $reversed_word;
}
echo count($words), "\n";
// array_rand() returns keys, not values
foreach (array_rand($words, 5) as $word)
    echo "$word $words[$word]\n";
Output:
158
ti it
tide edit
top pot
tram mart
un nu

PicoLisp

(let Semordnilap
   (mapcon
      '((Lst)
         (when (member (reverse (car Lst)) (cdr Lst))
            (cons (pack (car Lst))) ) )
      (make (in "unixdict.txt" (while (line) (link @)))) )
   (println (length Semordnilap) (head 5 Semordnilap)) )
Output:
158 ("able" "abut" "ac" "ah" "al")

PL/I

find: procedure options (main); /* 20/1/2013 */
   declare word character (20) varying controlled;
   declare dict(*) character (20) varying controlled;
   declare 1 pair controlled,
              2 a character (20) varying, 2 b character (20) varying;
   declare (i, j) fixed binary;
   declare in file;

   open file(in) title ('/UNIXDICT.TXT,type(LF),recsize(100)');
   on endfile (in) go to completed_read;
   do forever;
      allocate word;
      get file (in) edit (word) (L);
   end;

completed_read:
   free word; /* because at the final allocation, no word was stored. */
   allocate dict(allocation(word));
   do i = 1 to hbound(dict,1);
      dict(i) = word; free word;
   end;

   /* Search dictionary for pairs: */
   do i = 1 to hbound(dict,1)-1;
      do j = i+1 to hbound(dict,1);
         if length(dict(i)) = length(dict(j)) then
            do;
               if dict(i) = reverse(dict(j)) then
                  do;
                     allocate pair; pair.a = dict(i); pair.b = dict(j);
                  end;
            end;
      end;
   end;

   put skip list ('There are ' || trim(allocation(pair)) || ' pairs.');

   do while (allocation(pair) > 0);
      put skip edit (pair) (a, col(20), a); free pair;
   end;
end find;
There are 158 pairs. 

5 values at random:

ward               draw
was                saw
wed                dew
wolf               flow
won                now

PowerShell

function Reverse-String ([string]$String)
{
    [char[]]$output = $String.ToCharArray()
    [Array]::Reverse($output)
    $output -join ""
}

[string]$url = "http://www.puzzlers.org/pub/wordlists/unixdict.txt"
[string]$out = ".\unixdict.txt"

(New-Object System.Net.WebClient).DownloadFile($url, $out)

[string[]]$file = Get-Content -Path $out

[hashtable]$unixDict    = @{}
[hashtable]$semordnilap = @{}

foreach ($line in $file)
{
    if ($line.Length -gt 1)
    {
        $unixDict.Add($line,"")
    }

    [string]$reverseLine = Reverse-String $line

    if ($reverseLine -notmatch $line -and $unixDict.ContainsKey($reverseLine))
    {
        $semordnilap.Add($line,$reverseLine)
    }
}

$semordnilap

"`nSemordnilap count: {0}" -f ($semordnilap.GetEnumerator() | Measure-Object).Count
Output:
Name                           Value
----                           -----
nil                            lin
regal                          lager
tom                            mot
.
.
.
peek                           keep
soma                           amos
nob                            bon

Semordnilap count: 158

PureBasic

If OpenConsole("")=0 : End 1 : EndIf
If ReadFile(0,"./Data/unixdict.txt")=0 : End 2 : EndIf
NewList dict$()

While Eof(0)=0 : AddElement(dict$()) : dict$()=Trim(ReadString(0)) : Wend : CloseFile(0)

While FirstElement(dict$())  
  buf$=dict$() : DeleteElement(dict$())
  If buf$="" : Continue : EndIf
  xbuf$=ReverseString(buf$)  
  ForEach dict$()
    If xbuf$=dict$()
      res$+buf$+" / "+xbuf$+#LF$
      Break
    EndIf
  Next  
Wend

PrintN("Semordnilap pairs found: "+Str(CountString(res$,#LF$)))
For k=1 To 5
  If k=1 : PrintN(~"\nFirst 5 pairs: "+StringField(res$,k,#LF$)) : Continue : EndIf
  PrintN(Space(15)+StringField(res$,k,#LF$))
Next
Input()
Output:
Semordnilap pairs found: 158

First 5 pairs: able / elba
               abut / tuba
               ac / ca
               ah / ha
               al / la

Python

Idiomatic

>>> with open('unixdict.txt') as f:
	wordset = set(f.read().strip().split())

>>> revlist = (''.join(word[::-1]) for word in wordset)
>>> pairs   = set((word, rev) for word, rev in zip(wordset, revlist) 
                  if word < rev and rev in wordset)
>>> len(pairs)
158
>>> sorted(pairs, key=lambda p: (len(p[0]), p))[-5:]
[('damon', 'nomad'), ('lager', 'regal'), ('leper', 'repel'), ('lever', 'revel'), ('kramer', 'remark')]
>>>

As a fold, using reduce

Works with: Python version 3.7
'''Dictionary words paired by equivalence under reversal'''

from functools import (reduce)
from itertools import (chain)
import urllib.request


# semordnilaps :: [String] -> [String]
def semordnilaps(xs):
    '''The subset of words in a list which
       are paired (by equivalence under reversal)
       with other words in that list.
    '''
    def go(tpl, w):
        (s, ws) = tpl
        if w[::-1] in s:
            return (s, ws + [w])
        else:
            s.add(w)
            return (s, ws)
    return reduce(go, xs, (set(), []))[1]


# TEST ----------------------------------------------------
def main():
    '''Test'''

    url = 'http://wiki.puzzlers.org/pub/wordlists/unixdict.txt'
    ws = semordnilaps(
        urllib.request.urlopen(
            url
        ).read().splitlines()
    )
    print(
        fTable(
            __doc__ + ':\n\n(longest of ' +
            str(len(ws)) + ' in ' + url + ')\n'
        )(snd)(fst)(identity)(
            sorted(
                concatMap(
                    lambda x: (
                        lambda s=x.decode('utf8'): [
                            (s, s[::-1])
                        ] if 4 < len(x) else []
                    )()
                )(ws),
                key=compose(len)(fst),
                reverse=True
            )
        )
    )


# GENERIC -------------------------------------------------

# compose (<<<) :: (b -> c) -> (a -> b) -> a -> c
def compose(g):
    '''Right to left function composition.'''
    return lambda f: lambda x: g(f(x))


# concatMap :: (a -> [b]) -> [a] -> [b]
def concatMap(f):
    '''A concatenated list over which a function has been mapped.
       The list monad can be derived by using a function f which
       wraps its output in a list,
       (using an empty list to represent computational failure).'''
    return lambda xs: list(
        chain.from_iterable(map(f, xs))
    )


# FORMATTING ----------------------------------------------

# fTable :: String -> (a -> String) ->
#                     (b -> String) -> (a -> b) -> [a] -> String
def fTable(s):
    '''Heading -> x display function -> fx display function ->
                     f -> xs -> tabular string.
    '''
    def go(xShow, fxShow, f, xs):
        ys = [xShow(x) for x in xs]
        w = max(map(len, ys))
        return s + '\n' + '\n'.join(map(
            lambda x, y: y.rjust(w, ' ') + ' -> ' + fxShow(f(x)),
            xs, ys
        ))
    return lambda xShow: lambda fxShow: lambda f: lambda xs: go(
        xShow, fxShow, f, xs
    )


# fst :: (a, b) -> a
def fst(tpl):
    '''First member of a pair.'''
    return tpl[0]


# identity :: a -> a
def identity(x):
    '''The identity function.'''
    return x


# snd :: (a, b) -> b
def snd(tpl):
    '''Second member of a pair.'''
    return tpl[1]


if __name__ == '__main__':
    main()
Output:
Dictionary words paired by equivalence under reversal:

(longest of 158 in http://wiki.puzzlers.org/pub/wordlists/unixdict.txt)

kramer -> remark
 damon -> nomad
 lager -> regal
 leper -> repel
 lever -> revel

Lazy generator

Requires the requests library.

import sys
import random
import requests
 
URL = 'http://wiki.puzzlers.org/pub/wordlists/unixdict.txt'
 
 
def find_semordnilaps(word_generator):
 
  # Keys in this dict are the words seen so far, reversed.
  # Values are booleans determining whether we have seen (and yielded)
  # the key, so that we don't yield the same word twice.
  seen = {}
 
  for word in word_generator:
    if word not in seen:
      reversed_word = word[::-1]
      seen[reversed_word] = False  # not yielded yet
    else:
      yielded_already = seen[word]
      if not yielded_already:
        yield word
        seen[word] = True  # the word has been yielded
 
 
def url_lines(url):
  with requests.get(url, stream=True) as req:
    yield from req.iter_lines(decode_unicode=True)
 
 
def main(url=URL, num_of_examples=5):
 
  semordnilaps_generator = find_semordnilaps(url_lines(url))
 
  semordnilaps = list(semordnilaps_generator)
 
  example_words = random.choices(semordnilaps, k=int(num_of_examples))
  example_pairs = ((word, word[::-1]) for word in example_words)
 
  print(
    f'found {len(semordnilaps)} semordnilap words:',
    * ['%s %s' % p for p in example_pairs]+['...'],
    sep='\n'
  )
 
  return semordnilaps
 
 
if __name__ == '__main__':
  main(*sys.argv[1:])
Output:
found 158 semordnilap words:
mug gum
revel lever
gab bag
lime emil
lag gal
...

Quackery

  []
  $ "rosetta/unixdict.txt" sharefile drop 
  nest$
  [ behead reverse swap
    2dup find over found iff
      [ dip [ nested join ] ]
    else nip
    dup [] = until ] 
  drop
  say "Number of semordnilaps: "
  dup size echo cr
  sortwith [ size swap size > ]
  5 split drop
  say "Five longest: "
  witheach
    [ dup echo$ say "<->" 
      reverse echo$ sp ]
Output:
Number of semordnilaps: 158
Five longest: remark<->kramer nomad<->damon regal<->lager repel<->leper revel<->lever 


Racket

#lang racket
(define seen (make-hash))
(define semordnilaps '())
(call-with-input-file "/usr/share/dict/words"
  (λ(i) (for ([l (in-lines i)])
          (define r (list->string (reverse (string->list l))))
          (unless (equal? r l)
            (hash-set! seen l #t)
            (when (hash-ref seen r #f)
              (set! semordnilaps (cons (list r l) semordnilaps)))))))
(printf "Total semordnilaps found: ~s\n" (length semordnilaps))
(printf "The five longest ones:\n")
(for ([s (take (sort semordnilaps > #:key (compose1 string-length car)) 5)])
  (apply printf "  ~s ~s\n" s))
Output:
Total semordnilaps found: 1961
The five longest ones:
  "desserts" "stressed"
  "dioramas" "samaroid"
  "redrawer" "rewarder"
  "dessert" "tressed"
  "pat-pat" "tap-tap"

Raku

(formerly Perl 6)

Works with: rakudo version 2015-10-26
my $words = set slurp("unixdict.txt").lines;

my @sems = gather for $words.flat -> $word {
    my $drow = $word.key.flip;
    take $drow if $drow$words and $drow lt $word;
}

say $_ ~ ' ' ~ $_.flip for @sems.pick(5);
Output:
abut tuba
avid diva
bard drab
loot tool
part trap

REXX

version 1

/* REXX ***************************************************************
* 07.09.2012 Walter Pachl
**********************************************************************/
fid='unixdict.txt'                     /* the test dictionary        */
have.=''                               /* words encountered          */
pi=0                                   /* number of palindromes      */
Do li=1 By 1 While lines(fid)>0        /* as long there is input     */
  w=linein(fid)                        /* read a word                */
  If w>'' Then Do                      /* not a blank line           */
    r=reverse(w)                       /* reverse it                 */
    If have.r>'' Then Do               /* was already encountered    */
      pi=pi+1                          /* increment number of pal's  */
      If pi<=5 Then                    /* the first 5 ale listed     */
        Say have.r w
      End
    have.w=w                           /* remember the word          */
    End
  End
Say pi 'words in' fid 'have a palindrome' /* total number found      */
Output:
ac ca
bad dab
avid diva
and dna
bard drab
158 words in unixdict.txt have a palindrome  

version 2

This REXX version makes use of sparse (stemmed) arrays.

The dictionary file wasn't assumed to be in any particular case (upper/lower/mixed).

For instance,   DNA   &   and   would be considered semordnilaps,   even though their   case   is different.

The UNIXDICT dictionary specified to be used is all lowercase, however, but the REXX program assumes that the
words may be in any   case   (lowercase/uppercase/mixedcased).

The order of the words in the dictionary isn't important.

Any blank lines or duplicate words in the dictionary are ignored (as duplicate words wouldn't make them unique).

Any leading, trailing, or imbedded blanks are also ignored (as well as tab characters or other whitespace).

The semordnilap word pairs are shown with a comma delimiter in case there are phrases   (words with imbedded
blanks like   Sing Sing).

The (first five) semordnilap pairs are shown as they are specified/defined (respective to case) in the dictionary.

/*REXX program finds  N  semordnilap pairs using a specified dictionary  (UNIXDICT.TXT).*/
parse arg n iFID .                               /*obtain optional argument from the CL.*/
if    n=='' |    n=="," then    n= 5             /*Not specified?  Then use the default.*/
if iFID=='' | iFID=="," then iFID='UNIXDICT.TXT' /* "      "         "   "   "     "    */
#= 0                                             /*number of semordnilaps  (so far).    */
@.=                                              /*caseless non─duplicated dict. words. */
    do while lines(iFID)\==0;  _= linein(iFID);  u= space(_, 0);  upper u  /*get a word.*/
    if length(u)<2 | @.u\==''  then iterate      /*word can't be a unique semordnilap.  */
    r= reverse(u)                                /*obtain reverse of the dictionary word*/
    if @.r\==''  then do;   #= # + 1             /*found a semordnilap word; bump count.*/
                            if #<=n  then say right(@.r,  max(32, length(@.r) ) )','   u
                      end
    @.u= _                                       /*define reverse of the dictionary word*/
    end   /*while*/                              /*stick a fork in it,  we're all done. */
say
say "There're "     #     ' unique semordnilap pairs in the dictionary file: '    iFID
output   when using the default inputs:
                              ac, ca
                             bad, dab
                            avid, diva
                             and, dna
                            bard, drab

There're  158  unique semordnilap pairs in the dictionary file:  UNIXDICT.TXT

Ring

# Project : Semordnilap

load "stdlib.ring"
nr = 0
num = 0
aList = file2list("C:\Ring\CalmoSoft\unixdict.txt")
for n = 1 to len(aList)
     bool = semordnilap(aList[n])
     if (bool > 0 and nr > n)
        num = num + 1 
        if num % 31 = 0
           see aList[n] + " " + aList[nr] + nl 
        ok
     ok
next
see "Total number of unique pairs = " + num + nl

func semordnilap(aString)
       bString = ""
       for i=len(aString) to 1 step -1
            bString = bString + aString[i]
       next
       nr = find(aList,bString)
       return nr
Output:
brag garb
edit tide
it ti
mit tim
suez zeus
Total number of unique pairs = 158

Ruby

Note: An alternative (old fashioned) method of solving this task (not using a Set as done by other solutions) is to produce 2 sorted files and walk through them. This can be done entirly on disk if required, when done in memory it is faster than a set for large samples.--Nigel Galloway 11:12, 17 September 2012 (UTC)

dict = File.readlines("unixdict.txt").collect(&:strip)
i = 0
res = dict.collect(&:reverse).sort.select do |z| 
  i += 1  while z > dict[i] and i < dict.length-1
  z == dict[i] and z < z.reverse
end
puts "There are #{res.length} semordnilaps, of which the following are 5:"
res.take(5).each {|z| puts "#{z}   #{z.reverse}"}
Output:
There are 158 semordnilaps, of which the following are 5:
able   elba
abut   tuba
ac   ca
ah   ha
al   la

Another way

words = File.readlines("unixdict.txt")
            .group_by{|x| [x.strip!, x.reverse].min}
            .values
            .select{|v| v.size==2}
puts "There are #{words.size} semordnilaps, of which the following are 5:"
words.take(5).each {|a,b| puts "#{a}   #{b}"}

output is the same above.

Rust

Translation of: Java
use std::collections::HashSet;
use std::fs::File;
use std::io::{self, BufRead};
use std::iter::FromIterator;

fn semordnilap(filename: &str) -> std::io::Result<()> {
    let file = File::open(filename)?;
    let mut seen = HashSet::new();
    let mut count = 0;
    for line in io::BufReader::new(file).lines() {
        let mut word = line?;
        word.make_ascii_lowercase();
        let rev = String::from_iter(word.chars().rev());
        if seen.contains(&rev) {
            if count < 5 {
                println!("{}\t{}", word, rev);
            }
            count += 1;
        } else {
            seen.insert(word);
        }
    }
    println!("\nSemordnilap pairs found: {}", count);
    Ok(())
}

fn main() {
    match semordnilap("unixdict.txt") {
        Ok(()) => {}
        Err(error) => eprintln!("{}", error),
    }
}
Output:
ca	ac
dab	bad
diva	avid
dna	and
drab	bard

Semordnilap pairs found: 158

A Shortened Version

use std::{collections::HashSet, fs};

fn reverse(x: &str) -> String {
    x.chars().rev().collect::<String>()
}

fn main() {
    let content = fs::read_to_string("unixdict.txt").expect("No file found!");
    let work: HashSet<&str> = content.lines().collect();

    let mut candidats: Vec<&str> = work.clone().into_iter().collect();
    candidats.retain(|&x| work.contains(&reverse(x).as_str()) && x < reverse(x).as_str());

    println!("Numbers of pairs found: {}", candidats.len());
    for ind in 0..5 {
        println!("{}, {}", candidats[ind], reverse(candidats[ind]));
    }
}
Output:
Numbers of pairs found: 158
mn, nm
map, pam
saw, was
deep, peed
door, rood

Scala

val wordsAll =
  scala.io.Source.
    fromURL("http://wiki.puzzlers.org/pub/wordlists/unixdict.txt").
    getLines().map(_.toLowerCase).toIndexedSeq


/**
 * Given a sequence of lower-case words return a sub-sequence 
 * of matches containing the word and its reverse if the two
 * words are different.
 */
def semordnilap( words:IndexedSeq[String] ) : IndexedSeq[(String,String)] = {

  words.
    zipWithIndex.                        // index will be needed to eliminate duplicate
    filter {
      case (w,i) =>
        val j = words.indexOf(w.reverse) // eg. (able,62) and (elba,7519)
        i < j && w != w.reverse          // save the matches which are not palindromes
    }.
    map {
      case (w,_) => (w,w.reverse)        // drop the index
    }
}

val ss = semordnilap(wordsAll)

{
println( s"${ss.size} matches, including: \n" )
println( ss.take(5).mkString( "\n" ) )
}
Output:
158 matches, including:

(able,elba)
(abut,tuba)
(ac,ca)
(ah,ha)
(al,la)

Seed7

Note that the Seed7 program downloads unixdict.txt from the net.

$ include "seed7_05.s7i";
  include "gethttp.s7i";

const proc: main is func
  local
    var array string: wordList is 0 times "";
    var set of string: words is (set of string).value;
    var string: word is "";
    var string: drow is "";
    var integer: count is 0;
  begin
    wordList := split(lower(getHttp("wiki.puzzlers.org/pub/wordlists/unixdict.txt")), "\n");
    for word range wordList do
      drow := reverse(word);
      if drow not in words then
        incl(words, word);
      else
        if count < 5 then
          writeln(word <& " " <& drow);
        end if;
        incr(count);
      end if;
    end for;
    writeln;
    writeln("Semordnilap pairs: " <& count);
  end func;
Output:
ca ac
dab bad
diva avid
dna and
drab bard

Semordnilap pairs: 158

Sidef

Translation of: Perl
var c = 0
var seen = Hash()

ARGF.each { |line|
    line.chomp!
    var r = line.reverse
    ((seen{r} := 0 ++) && (c++ < 5) && say "#{line} #{r}") ->
        || (seen{line} := 0 ++)
}

say c
Output:
$ sidef semordnilap.sf < unixdict.txt
ca ac
dab bad
diva avid
dna and
drab bard
158

SNOBOL4

* Program: semordnilap.sbl
* To run: sbl semordnilap.sbl < unixdict.txt
* Description: A semordnilap is a word (or phrase) that spells
*              a different word (or phrase) backward. "Semordnilap"
*              is a word that itself is a semordnilap.
*              Example: lager and regal
*              Reads file unixdict.txt
* Comment: Tested using the Spitbol for Linux version of SNOBOL4

	output = "Some Semordnilap Pairs from File unixdict.txt"
	atable = table(25200,,-1)
	ntable = table(25200,,-2)

* Read dictionary file into memory
in1
	word = input :f(p1)
	count = count + 1
	atable[word] = word
	ntable[count] = word
	:(in1)

* Process dictionary to find unique semordnilaps
p1
	i = lt(i,count) i + 1 :f(p2)
 	newword = atable[reverse(ntable[i])]
	leq(newword,-1) :s(p1)
	ident(ntable[i],newword) :s(p1)
	output = lt(outcount,5) ntable[i] ', ' newword
	atable[ntable[i]] = atable[newword] = -1
	outcount = outcount + 1
	:(p1)
p2
	output = 'The number of semordnilap pairs is: ' outcount
END
Output:
Some Semordnilap Pairs from File unixdict.txt
able, elba
abut, tuba
ac, ca
ah, ha
al, la
The number of semordnilap pairs is: 158

Stata

set seed 17760704
import delimited http://www.puzzlers.org/pub/wordlists/unixdict.txt, clear
save temp, replace
replace v1=strreverse(v1)
merge 1:1 v1 using temp, nogen keep(3)
drop if v1>=strreverse(v1)
count
  158
sample 5, count
gen v2=strreverse(v1)
list, noheader noobs
  +-------------+
  | evil   live |
  |  pat    tap |
  |   at     ta |
  |  nit    tin |
  |   ku     uk |
  +-------------+


SuperCollider

Works with: SuperCollider version 3.13.0

Submitted to Rosetta Code 2024-06-07 by: MusicCoder.

// ==========================================================================
// START-SuperCollider solution to Rosetta Code TASK: Semordnilap					  ## BY: MusicCoder : 2024-06-07 ##
// ==========================================================================
(
/*
https://rosettacode.org/wiki/Semordnilap
A semordnilap is a word (or phrase) that spells a **different word** (or phrase) backward.
"Semordnilap" is a word that itself is a semordnilap, i.e. palindromes. Example: lager and regal

This task does not consider semordnilap phrases, only single words.
Using only words from this list: unixdict.txt,
report the total number of unique semordnilap pairs, and print 5 examples.
Two matching semordnilaps, such as lager and regal, should be counted as ***one unique pair***.
*/

var text, words, revs, candidates, pairs, selection;
// open file and read all contents into a single string variable called text
File.use("~/rosetta/data/unixdict.txt".standardizePath, "r", { |f| text = f.readAllString });
// to get words - split on new-line, to get distinct convert to SET
words = text.split(Char.nl).as(Set);
// reverse order of all characters then split & convert as above
revs  = text.reverse.split(Char.nl).as(Set);
// get the intersection of the two sets: forward and reversed words
candidates = (words & revs).as(Array);
// use a list comprehension to build the pairs AND to drop any palindromes & revs
pairs =  all {: [word, rev], word <- candidates, var rev = word.reverse, word < rev };

"There are % semordnilas in unixdict.txt\n".postf(pairs.size);
"For example those, with more than 3 characters:".postln;
// SELECT only those pairs where the word length is >= 4
// SCRAMBLE the resulting array and KEEP only 5 pairs
selection = pairs.select { |each| each[0].size >= 4 }.scramble.keep(5);
// PRINT each example on a new line
selection.do { |each| each.postln; };"\n";
)
// ==========================================================================
// **END-SuperCollider solution to Rosetta Code TASK: Semordnilap				  ## BY: MusicCoder : 2024-06-07 ##
// ==========================================================================
Output:
There are 158 semordnilas in unixdict.txt
For example those, with more than 3 characters:
[ gnaw, wang ]
[ part, trap ]
[ hoop, pooh ]
[ leper, repel ]
[ aryl, lyra ]

Swift

guard let data = try? String(contentsOfFile: "unixdict.txt") else {
  fatalError()
}

let words = Set(data.components(separatedBy: "\n"))
let pairs = words
  .map({ ($0, String($0.reversed())) })
  .filter({ $0.0 < $0.1 && words.contains($0.1) })

print("Found \(pairs.count) pairs")
print("Five examples: \(pairs.prefix(5))")
Output:
Found 158 pairs
Five examples: [("dial", "laid"), ("emil", "lime"), ("burg", "grub"), ("den", "ned"), ("enol", "lone")]

Tcl

package require Tcl 8.5
package require http

# Fetch the words
set t [http::geturl http://www.puzzlers.org/pub/wordlists/unixdict.txt]
set wordlist [split [http::data $t] \n]
http::cleanup $t

# Build hash table for speed
foreach word $wordlist {
    set reversed([string reverse $word]) "dummy"
}

# Find where a reversal exists
foreach word $wordlist {
    if {[info exists reversed($word)] && $word ne [string reverse $word]} {
	# Remove to prevent pairs from being printed twice
	unset reversed([string reverse $word])
	# Add to collection of pairs
	set pairs($word/[string reverse $word]) "dummy"
    }
}
set pairlist [array names pairs] ;# NB: pairs are in *arbitrary* order

# Report what we've found
puts "Found [llength $pairlist] reversed pairs"
foreach pair $pairlist {
    puts "Example: $pair"
    if {[incr i]>=5} break
}
Output:
Found 158 reversed pairs
Example: lap/pal
Example: jar/raj
Example: ix/xi
Example: eros/sore
Example: bard/drab

Transd

#lang transd

MainModule: {
_start: (λ (with fs FileStream() 
    (open-r fs "/mnt/vault/tmp/unixdict.txt") )
    (with v ( -|
        (read-text fs)
        (split)
        (group-by (λ s String() -> String() 
            (ret (min s (reverse (cp s))))))
        (values)
        (filter where: (λ v Vector<String>() (ret (== (size v) 2))))
        (shuffle))

        (lout "Total number of semordnilaps: " (size v))
        (lout "Random five: " Range(in: v 0 5))))
)
}
Output:
Total number of semordnilaps: 158
Random five: [[deer, reed], [nip, pin], [eire, erie], [am, ma], [gem, meg]]

TUSCRIPT

$$ MODE TUSCRIPT,{}
requestdata = REQUEST ("http://www.puzzlers.org/pub/wordlists/unixdict.txt")
DICT semordnilap CREATE 99999
COMPILE
LOOP r=requestdata
rstrings=STRINGS(r," ? ")
rreverse=REVERSE(rstrings)
revstring=EXCHANGE (rreverse,":'':':'::")
group=APPEND (r,revstring)
sort=ALPHA_SORT (group)
DICT semordnilap APPEND/QUIET/COUNT sort,num,cnt,"",""
ENDLOOP
DICT semordnilap UNLOAD wordgroups,num,howmany
get_palins=FILTER_INDEX (howmany,-," 1 ")
size=SIZE(get_palins)
PRINT "unixdict.txt contains ", size, " palindromes"
PRINT " "
palindromes=SELECT (wordgroups,#get_palins)
LOOP n=1,5
take5=SELECT (palindromes,#n)
PRINT n,". ",take5
ENDLOOP
ENDCOMPILE
Output:
unixdict.txt contains 158 palindromes

1. able'elba
2. abut'tuba
3. ac'ca
4. ah'ha
5. al'la

Uiua

For each word check its reverse is in the array, and that it's greater. This removes palindromes and also keeps only first of each pair.

&p &pf "Count: " ⧻.▽≡(×>⇌.⟜∊⇌)⟜¤. ⊜□ ≠@\n. &fras "unixdict.txt"
Output:
Count: 158
{"able" "abut" "ac" "ah" "al" "am" "amos" "and" "ape" "aps" "are" 
...
"pot" "pow" "pus" "rat" "raw" "rot" "saw" "suez" "tort" "tv" "way"}

VBScript

Set objFSO = CreateObject("Scripting.FileSystemObject")
Set objInFile = objFSO.OpenTextFile(objFSO.GetParentFolderName(WScript.ScriptFullName) &_
		"\unixdict.txt",1)

Set objUnixDict = CreateObject("Scripting.Dictionary")
Set objSemordnilap = CreateObject("Scripting.Dictionary")

Do Until objInFile.AtEndOfStream
	line = objInFile.ReadLine
	If Len(line) > 1 Then
		objUnixDict.Add line,""
	End If
	reverse_line = StrReverse(line)
	If reverse_line <> line And objUnixDict.Exists(reverse_line) Then
		objSemordnilap.Add line, reverse_line
	End If
Loop	

'Display the first 5 keys.
k = 0
For Each Key In objSemordnilap.Keys
	WScript.StdOut.Write Key & " - " & objSemordnilap.Item(Key)
	WScript.StdOut.WriteLine
	k = k + 1
	If k = 5 Then
		Exit For
	End If
Next

WScript.StdOut.Write "Total Count:  " & objSemordnilap.Count
WScript.StdOut.WriteLine

objInFile.Close
Set objFSO = Nothing
Set objUnixDict = Nothing
Set objSemordnilap = Nothing
Output:
ca - ac
dab - bad
diva - avid
dna - and
drab - bard
Total Count:  158

Wren

import "io" for File

var dict = File.read("unixdict.txt").split("\n")
var wmap = {}
dict.each { |w| wmap[w] = true }
var pairs = []
var used = {}
for (word in dict) {
    if (word != "") {
        var pal = word[-1..0]
        if (word != pal && wmap[pal] && !used[pal]) {
            pairs.add([word, pal])
            used[word] = true
        }
    }
}
System.print("There are %(pairs.count) unique semordnilap pairs in the dictionary.")
System.print("\nIn sorted order, the first five are:")
for (i in 0..4) System.print("  %(pairs[i][0]), %(pairs[i][1])")
System.print("\nand the last five are:")
for (i in -5..-1) System.print("  %(pairs[i][0]), %(pairs[i][1])")
Output:
There are 158 unique semordnilap pairs in the dictionary.

In sorted order, the first five are:
  able, elba
  abut, tuba
  ac, ca
  ah, ha
  al, la

and the last five are:
  saw, was
  suez, zeus
  tort, trot
  tv, vt
  way, yaw

XPL0

include c:\cxpl\codes;  \intrinsic 'code' declarations
string 0;               \use zero-terminated strings
def LF=$0A, CR=$0D, EOF=$1A;

proc RevStr(S);         \Reverse order of characters in a string
char S;
int  I, J, T;
[J:= 0;
while S(J) do J:= J+1;
J:= J-1;
I:= 0;
while I<J do
        [T:= S(I);  S(I):= S(J);  S(J):= T;     \swap
        I:= I+1;  J:= J-1;
        ];
];

func StrEqual(S1, S2);  \Compare strings, return 'true' if equal
char S1, S2;
int  I;
[for I:= 0 to 80-1 do
        [if S1(I) # S2(I) then return false;
         if S1(I) = 0 then return true;
        ];
];

int  C, I, J, SJ, Count;
char Dict, Word(80);
[\Read file on command line redirected as input, i.e: <unixdict.txt
Dict:= GetHp;           \starting address of block of local "heap" memory
I:= 0;                  \ [GetHp does exact same thing as Reserve(0)]
repeat  repeat C:= ChIn(1) until C#LF;  \get chars sans line feeds
        if C = CR then C:= 0;           \replace carriage return with terminator
        Dict(I):= C;  I:= I+1;
until   C = EOF;
SetHp(Dict+I);          \set heap pointer beyond Dict
I:= 0;  Count:= 0;
loop    [J:= 0;         \get word at I
        repeat  C:= Dict(I+J);  Word(J):= C;  J:= J+1;
        until   C=0;
        RevStr(Word);
        J:= J+I;        \set J to following word in Dict
        if Dict(J) = EOF then quit;
        SJ:= J;         \save index to following word
        loop    [if StrEqual(Word, Dict+J) then
                    [Count:= Count+1;
                    if Count <= 5 then
                        [RevStr(Word);  \show some examples
                        Text(0, Word); ChOut(0, ^ ); Text(0, Dict+J); CrLf(0);
                        ];
                    quit;
                    ];
                repeat J:= J+1 until Dict(J) = 0;
                J:= J+1;
                if Dict(J) = EOF then quit;
                ];
        I:= SJ;         \next word
        ];
IntOut(0, Count);  CrLf(0);
]
Output:
able elba
abut tuba
ac ca
ah ha
al la
158

Yabasic

Translation of: FreeBASIC
dim norm$(27000), result$(270, 2)

print "Start reading unixdict.txt"
open "i:\unixdict.txt" for reading as #1

while not(eof(#1))                   // read to end of file
    line input #1 in_str$            // get line = word
    in_str$ = trim$(in_str$)         // we don//t want spaces
    if len(in_str$) > 1 then         // if length > 1 then reverse$
        rev$ = reverse$(in_str$)
        if in_str$ <> rev$ then      // if in_str is not a palingdrome
            count = count + 1        // increase counter
            norm$(count) = in_str$   // store in the array
            big$ = big$ + rev$ + " " // create big string with reverse$d words
        fi
    fi
wend

close #1
print " ... Done"
print 
print "Start looking for semordnilap"

for i = 1 to count
    for j = 1 to amount                          // check to avoid the double
        if result$(j, 2) = norm$(i)  continue
    next j
    j = instr(big$, " " + norm$(i) + " ")
    if j <> 0 then                               // found one
        amount = amount + 1                      // increase counter
        result$(amount, 1) = norm$(i)            // store normal word
        result$(amount, 2) = reverse$(norm$(i))  // store reverse$ word
    fi
next i

print 
print "Found", amount, " unique semordnilap pairs"
print
print "Display 5 semordnilap pairs"
print

count = 0
for i = 1 to amount
    if len(result$(i, 1)) >= 5 then
        count = count + 1
        print result$(i, 1), chr$(9), result$(i, 2)
        if count >= 5  break
    fi
next i
end

sub reverse$(norm$)
    local rev$, i, l
    l = len(norm$) - 1
    rev$ = norm$
    for i = 0 to l
        mid$(rev$, l - i, 1) = mid$(norm$, i, 1)
    next i
    return rev$
end sub
Output:
Start reading unixdict.txt
 ... Done

Start looking for semordnilap

Found70 unique semordnilap pairs

Display 5 semordnilap pairs

diesel  seidel
dirge   ridge
gamma   magma
groan   organ
latus   talus

---Program done, press RETURN---

zkl

var [const] words=  // create hashed unixdict of striped words (word:True, ...)
   File("dict.txt").howza(11).pump(Dictionary().howza(8).add.fp1(True));
ss:=words.pump(List, // push stripped unixdict words through some functions
   fcn(w){ words.holds(w.reverse()) }, Void.Filter,   // filter palindromes
   // create ("word","drow") if "word"<"drow" (ie remove duplicates)
   fcn(w){ r:=w.reverse(); if(w<r) T(w,r) else Void.Skip });

ss.len().println(); //--> 158
ss.shuffle()[0,5].println();
Output:
158
L(L("bog","gob"),L("cup","puc"),L("mart","tram"),L("kay","yak"),L("able","elba"))