Anagrams/Deranged anagrams

From Rosetta Code
Task
Anagrams/Deranged anagrams
You are encouraged to solve this task according to the task description, using any language you may know.

Two or more words are said to be anagrams if they have the same characters, but in a different order.

By analogy with derangements we define a deranged anagram as two words with the same characters, but in which the same character does not appear in the same position in both words.

Task

Use the word list at unixdict to find and display the longest deranged anagram.


Related


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: Kotlin
F is_not_deranged(s1, s2)
   L(i) 0 .< s1.len
      I s1[i] == s2[i]
         R 1B
   R 0B

Dict[String, Array[String]] anagram
V count = 0
L(word) File(‘unixdict.txt’).read().split("\n")
   V a = sorted(word)
   I a !C anagram
      anagram[a] = [word]
   E
      L(ana) anagram[a]
         I is_not_deranged(ana, word)
            L.break
      L.was_no_break
         anagram[a].append(word)
         count = max(count, word.len)

L(ana) anagram.values()
   I ana.len > 1 & ana[0].len == count
      print(ana)
Output:
[excitation, intoxicate]

AArch64 Assembly

Works with: as version Raspberry Pi 3B version Buster 64 bits
or android 64 bits with application Termux
/* ARM assembly AARCH64 Raspberry PI 3B */
/*  program anaderan64.s   */

/*******************************************/
/* Constantes file                         */
/*******************************************/
/* for this file see task include a file in language AArch64 assembly*/
.include "../includeConstantesARM64.inc"
 
.equ MAXI,         50000
.equ BUFFERSIZE,   300000

/*********************************/
/* Structures                    */
/*********************************/
/* this structure has size multiple de 8 */
/* see task anagram for program not use structure */
    .struct  0
Word_Ptr_buffer:                           // p.quader word buffer
    .struct  Word_Ptr_buffer + 8 
Word_Ptr_sorted:                           // p.quader word sorted letters
    .struct  Word_Ptr_sorted + 8 
Word_length:                               // word length
    .struct  Word_length + 8 
Word_top:                                  // top
    .struct  Word_top + 8 
Word_end:
/*********************************/
/* Initialized data              */
/*********************************/
.data
szFileName:           .asciz "./unixdict.txt"
//szFileName:           .asciz "./listwordT.txt"
szMessErreur:         .asciz "FILE ERROR."
szMessStart:          .asciz "Program 64 bits start.\n"
szCarriageReturn:     .asciz "\n"
szMessSpace:          .asciz " "

ptBuffex1:            .quad sBuffex1
/*********************************/
/* UnInitialized data            */
/*********************************/
.bss
tbStWords:                  .skip Word_end * MAXI
qNBword:                    .skip 8
sBuffer:                    .skip BUFFERSIZE
sBuffex1:                   .skip BUFFERSIZE

/*********************************/
/*  code section                 */
/*********************************/
.text
.global main 
main:                      // entry of program 
    ldr x0,qAdrszMessStart
    bl affichageMess
    mov x4,#0              // loop indice
    mov x0,AT_FDCWD        // current directory
    ldr x1,qAdrszFileName  // file name
    mov x2,#O_RDWR         // flags
    mov x3,#0              // mode
    mov x8,#OPEN           // 
    svc 0 
    cmp x0,#0              // error open
    ble 99f
    mov x9,x0              // FD save Fd
    ldr x1,qAdrsBuffer     // buffer address
    ldr x2,qSizeBuf        // buffersize
    mov x8, #READ
    svc 0 
    cmp x0,#0              // error read ?
    blt 99f
    mov x5,x0              // save size read bytes
    ldr x4,qAdrsBuffer     // buffer address
    ldr x0,qAdrsBuffer     // start word address
    mov x2,#0
    mov x1,#0              // word length
1:
    cmp x2,x5
    bge 2f
    ldrb w3,[x4,x2]
    cmp w3,#0xD            // end word ?
    cinc x1,x1,ne          // increment word length
    cinc x2,x2,ne          // increment indice
    bne 1b                 // and loop
    strb wzr,[x4,x2]        // store final zero
    bl anaWord             // sort word letters
    add x2,x2,#2           // jump OD and 0A 
    add x0,x4,x2           // new address begin word
    mov x1,#0              // init length
    b 1b                   // and loop
    
2:
    strb wzr,[x4,x2]       // last word
    bl anaWord
    
    mov x0,x9              // file Fd
    mov x8, #CLOSE
    svc 0 
    cmp x0,#0              // error close ?
    blt 99f
    
    ldr x0,qAdrtbStWords   // array structure words address
    mov x1,#0              // first indice
    ldr x2,qAdrqNBword
    ldr x2,[x2]            // last indice
    bl triRapide           // quick sort
    ldr x4,qAdrtbStWords   // array structure words address
    mov x0,x4
    mov x9,x2                     // size word array
    mov x8,#0                     // indice first occurence
    ldr x3,[x4,#Word_Ptr_sorted]  // load first value
    mov x2,#1                     // loop indice
    mov x10,#Word_end             // words structure size
    mov x12,#0                    // max word length
3:                                // begin loop 
    madd x7,x2,x10,x4              // compute array index
    ldr x5,[x7,#Word_Ptr_sorted]  // load next value
    mov x0,x3
    mov x1,x5
    bl comparStrings
    cmp x0,#0                     // sorted strings equal ?
    bne 4f
    madd x11,x8,x10,x4
    ldr x0,[x11,#Word_Ptr_buffer] // address word 1
    ldr x1,[x7,#Word_Ptr_buffer]  // address word 2
    bl controlLetters
    cmp x0,#0                     // not ok ?
    beq 5f
    mov x0,#1                     // letters ok
    str x0,[x7,#Word_top]         // store top in first occurence
    str x0,[x11,#Word_top]        // store top in current occurence
    ldr x0,[x7,#Word_length]      // word length
    cmp x0,x12                    // compare maxi
    csel x12,x0,x12,gt            // yes  length word -> value max
    b 5f
4:                                // no
    mov x0,x8
    add x8,x8,#1                  // init index  new first occurence
    madd x11,x8,x10,x4             // compute array index
    ldr x3,[x11,#Word_Ptr_sorted] // init value new first occurence
    mov x2,x0                     // reprise au debut de la sequence
5:
    add x2,x2,#1                  // increment indice
    cmp x2,x9                     // end word array ?
    blt 3b                        // no -> loop
    
    mov x2,#0                     // raz indice
    ldr x4,qAdrtbStWords          // array structure words address
6:                                // begin display loop
    madd x11,x2,x10,x4             // compute array index
    ldr x6,[x11,#Word_top]        // load top
    cmp x6,#0                     // top ok ?
    beq 7f
    ldr x6,[x11,#Word_length]     // load length
    cmp x6,x12                    // compare maxi
    bne 7f
    ldr x0,[x11,#Word_Ptr_buffer] // load address first word
    bl affichageMess              // display first word
    add x2,x2,#1                  // increment indice

    madd x11,x2,x10,x4             // compute array index
    ldr x6,[x11,#Word_top]        // load top
    cmp x6,#0                     // top ok ?
    beq 7f
    ldr x0,qAdrszMessSpace
    bl affichageMess
    ldr x0,[x11,#Word_Ptr_buffer] // load address other word
    bl affichageMess              // display second word

    ldr x0,qAdrszCarriageReturn
    bl affichageMess
7:
    add x2,x2,#1                  // increment indice
    cmp x2,x9                     // maxi ?
    blt 6b                        // no -> loop
    
    b 100f
99:                               // display error
    ldr x0,qAdrszMessErreur
    bl affichageMess
    
100:                              // standard end of the program 
    mov x0, #0                    // return code
    mov x8, #EXIT                 // request to exit program
    svc #0                        // perform the system call
qAdrszCarriageReturn:        .quad szCarriageReturn
qAdrszFileName:              .quad szFileName
qAdrszMessErreur:            .quad szMessErreur
qAdrsBuffer:                 .quad sBuffer
qSizeBuf:                    .quad BUFFERSIZE
qAdrszMessSpace:             .quad szMessSpace
qAdrtbStWords:               .quad tbStWords
qAdrszMessStart:             .quad szMessStart
/******************************************************************/
/*     analizing word                                   */ 
/******************************************************************/
/*  x0  word address */
/*  x1 word length   */
anaWord:
    stp x1,lr,[sp,-16]!           // save  registers
    stp x2,x3,[sp,-16]!           // save  registers
    stp x4,x5,[sp,-16]!           // save  registers
    stp x6,x7,[sp,-16]!           // save  registers
    mov x5,x0
    mov x6,x1
    ldr x1,qAdrtbStWords
    ldr x2,qAdrqNBword
    ldr x3,[x2]
    mov x4,#Word_end
    madd x1,x3,x4,x1
    str x0,[x1,#Word_Ptr_buffer]
    mov x0,#0
    str x0,[x1,#Word_top]
    str x6,[x1,#Word_length]
    ldr x4,qAdrptBuffex1
    ldr x0,[x4]
    add x6,x6,x0
    add x6,x6,#1
    str x6,[x4]
    str x0,[x1,#Word_Ptr_sorted]
    
    add x3,x3,#1
    str x3,[x2]
    mov x1,x0
    mov x0,x5
    bl triLetters         // sort word letters
    mov x2,#0
100:
    ldp x6,x7,[sp],16              // restaur  2 registers
    ldp x4,x5,[sp],16              // restaur  2 registers
    ldp x2,x3,[sp],16              // restaur  2 registers
    ldp x1,lr,[sp],16              // restaur  2 registers
    ret                            // return to address lr x30
qAdrqNBword:            .quad qNBword
qAdrptBuffex1:          .quad ptBuffex1
/******************************************************************/
/*     sort word letters                                  */ 
/******************************************************************/
/* x0  address begin word */
/* x1  address recept array */
triLetters:
    stp x1,lr,[sp,-16]!           // save  registers
    stp x2,x3,[sp,-16]!           // save  registers
    stp x4,x5,[sp,-16]!           // save  registers
    stp x6,x7,[sp,-16]!           // save  registers
    mov x2,#0
1:
    ldrb w3,[x0,x2]         // load letter
    cmp w3,#0               // end word ?
    beq 6f
    cmp x2,#0               // first letter ?
    bne 2f
    strb w3,[x1,x2]         // yes store in first position
    add x2,x2,#1            // increment indice
    b 1b                    // and loop
2:
    mov x4,#0
3:                          // begin loop to search insertion position
    ldrb w5,[x1,x4]         // load letter 
    cmp w3,w5               // compare
    blt 4f                  // to low -> insertion
    add x4,x4,#1            // increment indice
    cmp x4,x2               // compare to letters number in place
    blt 3b                  // search loop
    strb w3,[x1,x2]         // else store in last position
    add x2,x2,#1
    b 1b                    // and loop
4:                          // move first letters in one position 
    sub x6,x2,#1            // start indice
5:
    ldrb w5,[x1,x6]         // load letter
    add x7,x6,#1            // store indice - 1
    strb w5,[x1,x7]         // store letter
    sub x6,x6,#1            // decrement indice
    cmp x6,x4               // end ?
    bge 5b                  // no loop
    strb w3,[x1,x4]         // else store letter in free position
    add x2,x2,#1
    b 1b                    // and loop
6:
    strb wzr,[x1,x2]        // final zéro
100:
    ldp x6,x7,[sp],16              // restaur  2 registers
    ldp x4,x5,[sp],16              // restaur  2 registers
    ldp x2,x3,[sp],16              // restaur  2 registers
    ldp x1,lr,[sp],16              // restaur  2 registers
    ret                            // return to address lr x30
/******************************************************************/
/*     control letters                                  */ 
/******************************************************************/
/* x0  address  word 1*/
/* x1  address  word 2 */
controlLetters:
    stp x1,lr,[sp,-16]!           // save  registers
    stp x2,x3,[sp,-16]!           // save  registers
    stp x4,x5,[sp,-16]!           // save  registers
    mov x2,#0
    mov x5,x0
1:
    ldrb w3,[x5,x2]      // load a letter 
    cmp w3,#0            // end word ?
    cset x0,eq
   // moveq x0,#1          //  yes it is OK
    beq 100f
    ldrb w4,[x1,x2]      // load a letter word 2 same position
    cmp w3,w4            // equal ?
    cset x0,ne           // yes -> not good
    //moveq x0,#0          // yes -> not good
    beq 100f
    add x2,x2,#1
    b 1b
100:
    ldp x4,x5,[sp],16              // restaur  2 registers
    ldp x2,x3,[sp],16              // restaur  2 registers
    ldp x1,lr,[sp],16              // restaur  2 registers
    ret                            // return to address lr x30
/***************************************************/
/*   Appel récursif Tri Rapide quicksort           */
/***************************************************/
/* x0 contains the address of table */
/* x1 contains index of first item  */
/* x2 contains the number of elements  > 0  */
triRapide:
    stp x1,lr,[sp,-16]!           // save  registers
    stp x2,x3,[sp,-16]!           // save  registers
    stp x4,x5,[sp,-16]!           // save  registers
    sub x2,x2,#1               // last item index
    cmp x1,x2               // first > last ? 
    bge 100f                // yes -> end
    mov x4,x0               // save x0
    mov x5,x2               // save x2
    bl partition1           // cutting.quado 2 parts
    mov x2,x0               // index partition
    mov x0,x4               // table address
    bl triRapide            // sort lower part
    mov x0,x4               // table address
    add x1,x2,#1            // index begin = index partition + 1
    add x2,x5,#1            // number of elements
    bl triRapide            // sort higter part
   
 100:                       // end function
    ldp x4,x5,[sp],16       // restaur  2 registers
    ldp x2,x3,[sp],16       // restaur  2 registers
    ldp x1,lr,[sp],16       // restaur  2 registers
    ret                     // return to address lr x30

/******************************************************************/
/*      Partition table elements                                */ 
/******************************************************************/
/* x0 contains the address of table */
/* x1 contains index of first item  */
/* x2 contains index of last item   */
partition1:
    stp x1,lr,[sp,-16]!     // save  registers
    stp x2,x3,[sp,-16]!     // save  registers
    stp x4,x5,[sp,-16]!     // save  registers
    stp x6,x7,[sp,-16]!     // save  registers
    stp x8,x9,[sp,-16]!     // save  registers
    mov x8,x0               // save address table 2
    mov x7,x2
    mov x9,#Word_end
    madd x3,x7,x9,x8
    ldr x6,[x3,#Word_Ptr_sorted]  // load string address last index
    mov x4,x1               // init with first index
    mov x5,x1               // init with first index
1:                          // begin loop
    madd x3,x5,x9,x8
    ldr x0,[x3,#Word_Ptr_sorted]   // load current string address
    mov x1,x6                      // first string address
    bl comparStrings        
    cmp x0,#0
    bge 2f
    mov x0,x8                // current string < first string
    mov x1,x4                // swap array
    mov x2,x5
    bl swapWord
    add x4,x4,#1             // and increment index 1
2:
    add x5,x5,#1             // increment index 2
    cmp x5,x7                // end ?
    blt 1b                   // no -> loop
    
    mov x0,x8                // and swap array 
    mov x1,x4
    mov x2,x7
    bl swapWord

    mov x0,x4                // return index partition
100:
    ldp x8,x9,[sp],16       // restaur  2 registers
    ldp x6,x7,[sp],16       // restaur  2 registers
    ldp x4,x5,[sp],16       // restaur  2 registers
    ldp x2,x3,[sp],16       // restaur  2 registers
    ldp x1,lr,[sp],16       // restaur  2 registers
    ret                     // return to address lr x30
/******************************************************************/
/*      Swap table elements                                */ 
/******************************************************************/
/* x0 contains the address of table */
/* x1 contains index 1  */
/* x2 contains index  2   */
swapWord:
    stp x1,lr,[sp,-16]!           // save  registers
    stp x2,x3,[sp,-16]!           // save  registers
    stp x4,x5,[sp,-16]!           // save  registers
    stp x6,x7,[sp,-16]!           // save  registers
    mov x3,#Word_end
    madd x4,x1,x3,x0        // compute array index
    madd x5,x2,x3,x0
    mov x6,#0
1:
    ldr x2,[x4,x6]         // load 4 bytes
    ldr x3,[x5,x6]
    str x2,[x5,x6]         // store 4 bytes
    str x3,[x4,x6]
    add x6,x6,#8           // increment 4 bytes
    cmp x6,#Word_end       // structure size is multiple to 4 
    blt 1b
100:
    ldp x6,x7,[sp],16              // restaur  2 registers
    ldp x4,x5,[sp],16              // restaur  2 registers
    ldp x2,x3,[sp],16              // restaur  2 registers
    ldp x1,lr,[sp],16              // restaur  2 registers
    ret                            // return to address lr x30

/************************************/       
/* Strings case sensitive comparisons  */
/************************************/      
/* x0 et x1 contains the address of strings */
/* return 0 in x0 if equals */
/* return -1 if string x0 < string x1 */
/* return 1  if string x0 > string x1 */
comparStrings:
    stp x1,lr,[sp,-16]! // save  registers
    stp x2,x3,[sp,-16]! // save  registers
    stp x4,x5,[sp,-16]! // save  registers
    mov x2,#0           // counter
1:    
    ldrb w3,[x0,x2]     // byte string 1
    ldrb w4,[x1,x2]     // byte string 2
    cmp w3,w4
    blt 2f              // small
    bgt 3f              // greather  
    cmp x3,#0           // 0 end string
    beq 4f              // end string
    add x2,x2,#1        // else add 1 in counter
    b 1b                // and loop
2:
    mov x0,#-1          // small
    b 100f
3:
    mov x0,#1           // greather  
    b 100f
4:
   mov x0,#0            // equal
100:
    ldp x4,x5,[sp],16   // restaur  2 registers
    ldp x2,x3,[sp],16   // restaur  2 registers
    ldp x1,lr,[sp],16   // restaur  2 registers
    ret                 // return to address lr x30

/********************************************************/
/*        File Include fonctions                        */
/********************************************************/
/* for this file see task include a file in language AArch64 assembly */
.include "../includeARM64.inc"
Program 64 bits start.
excitation intoxicate

Ada

Works with: Ada 2005
with Ada.Text_IO; use Ada.Text_IO;
with Ada.Containers.Generic_Array_Sort;
with Ada.Containers.Indefinite_Vectors;
procedure Danagrams is
   package StringVector is new Ada.Containers.Indefinite_Vectors
      (Positive, String);
   procedure StrSort is new Ada.Containers.Generic_Array_Sort
      (Index_Type => Positive,
      Element_Type => Character,
      Array_Type => String);
   function Derange (s1 : String; s2 : String) return Boolean is begin
      for i in s1'Range loop
         if (s1 (i) = s2 (i)) then return False; end if;
      end loop;
      return True;
   end Derange;
   File : File_Type;
   len, foundlen : Positive := 1;
   Vect, SVect : StringVector.Vector;
   index, p1, p2 : StringVector.Extended_Index := 0;
begin
   Open (File, In_File, "unixdict.txt");
   while not End_Of_File (File) loop
      declare str : String := Get_Line (File);
      begin
         len := str'Length;
         if len > foundlen then
            Vect.Append (str);
            StrSort (str);
            index := 0;
            loop --  Loop through anagrams by index in vector of sorted strings
               index := SVect.Find_Index (str, index + 1);
               exit when index = StringVector.No_Index;
               if Derange (Vect.Last_Element, Vect.Element (index)) then
                     p1 := Vect.Last_Index; p2 := index;
                     foundlen := len;
               end if;
            end loop;
            SVect.Append (str);
         end if;
      end;
   end loop;
   Close (File);
   Put_Line (Vect.Element (p1) & " " & Vect.Element (p2));
end Danagrams;
Output:
intoxicate excitation

ALGOL 68

Works with: ALGOL 68G version Any - tested with release 2.8.3.win32
Uses the "read" PRAGMA of Algol 68 G to include the associative array code from the Associative_array/Iteration task.
# find the largest deranged anagrams in a list of words              #
# use the associative array in the Associate array/iteration task    #
PR read "aArray.a68" PR

# returns the length of str                                          #
OP LENGTH = ( STRING str )INT: 1 + ( UPB str - LWB str );

# returns TRUE if a and b are the same length and have no            #
# identical characters at any position,                              #
# FALSE otherwise                                                    #
PRIO ALLDIFFER = 9;
OP   ALLDIFFER = ( STRING a, b )BOOL:
     IF LENGTH a /= LENGTH b
     THEN
        # the two stringa are not the same size                      #
        FALSE
     ELSE
        # the strings are the same length, check the characters      #
        BOOL result := TRUE;
        INT  b pos  := LWB b;
        FOR a pos FROM LWB a TO UPB a WHILE result := ( a[ a pos ] /= b[ b pos ] )
        DO
            b pos +:= 1
        OD;
        result
     FI # ALLDIFFER # ;

# returns text with the characters sorted                            #
OP   SORT = ( STRING text )STRING:
     BEGIN
        STRING sorted := text;
        FOR end pos FROM UPB sorted - 1 BY -1 TO LWB sorted
        WHILE
            BOOL swapped := FALSE;
            FOR pos FROM LWB sorted TO end pos DO
                IF sorted[ pos ] > sorted[ pos + 1 ]
                THEN              
                    CHAR  t           := sorted[ pos     ];
                    sorted[ pos     ] := sorted[ pos + 1 ];
                    sorted[ pos + 1 ] := t;
                    swapped           := TRUE
                FI
            OD;
            swapped
        DO SKIP OD;
        sorted
     END # SORTED # ;

# read the list of words and find the longest deranged anagrams      #

CHAR separator = "|"; # character that will separate the anagrams    #

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    longest derangement := 0;
    STRING longest word        := "<none>";
    STRING longest anagram     := "<none>";
    WHILE NOT at eof
    DO
        STRING word;
        get( input file, ( word, newline ) );
        INT    word length = LENGTH word;
        IF  word length >= longest derangement
        THEN
            # this word is at least long as the longest derangement   #
            # found so far - test it                                  #
            STRING sorted word = SORT word;
            IF ( words // sorted word ) /= ""
            THEN
                # we already have this sorted word - test for         #
                # deranged anagrams                                   #
                # the word list will have a leading separator         #
                # and be followed by one or more words separated by   #
                # the separator                                       #
                STRING word list   := words // sorted word;
                INT    list pos    := LWB word list + 1;
                INT    list max     = UPB word list;
                BOOL   is deranged := FALSE;
                WHILE  list pos < list max
                  AND  NOT is deranged
                DO
                    STRING anagram = word list[ list pos : ( list pos + word length ) - 1 ];
                    IF  is deranged := word ALLDIFFER anagram
                    THEN
                        # have a deranged anagram                     #
                        longest derangement := word length;
                        longest word        := word;
                        longest anagram     := anagram
                    FI;            
                    list pos +:= word length + 1
                OD
            FI;
            # add the word to the anagram list                        #
            words // sorted word +:= separator + word
        FI
    OD;
    close( input file );
    print( ( "Longest deranged anagrams: "
           , longest word
           , " and "
           , longest anagram
           , newline
           )
         )
FI
Output:
Longest deranged anagrams: intoxicate and excitation

AppleScript

use AppleScript version "2.3.1" -- OS X 10.9 (Mavericks) or later.
use sorter : script ¬
    "Custom Iterative Ternary Merge Sort" -- <www.macscripter.net/t/timsort-and-nigsort/71383/3>
use scripting additions

on join(lst, delim)
    set astid to AppleScript's text item delimiters
    set AppleScript's text item delimiters to delim
    set txt to lst as text
    set AppleScript's text item delimiters to astid
    return txt
end join

on longestDerangedAnagrams(listOfWords)
    script o
        property wordList : listOfWords
        property groupingTexts : wordList's items
        property derangementLength : 0
        property output : {}
        
        -- Test for any deranged pairs amongst the words of an anagram group.
        on testPairs(a, b)
            set anagramGroup to my wordList's items a thru b
            set groupSize to b - a + 1
            set wordLength to (count beginning of anagramGroup)
            repeat with i from 1 to (groupSize - 1)
                set w1 to anagramGroup's item i
                repeat with j from (i + 1) to groupSize
                    set w2 to anagramGroup's item j
                    set areDeranged to true
                    repeat with c from 1 to wordLength
                        if (w1's character c = w2's character c) then
                            set areDeranged to false
                            exit repeat
                        end if
                    end repeat
                    -- Append any deranged pairs found to the output and note the words' length.
                    if (areDeranged) then
                        set end of output to {w1, w2}
                        set derangementLength to wordLength
                    end if
                end repeat
            end repeat
        end testPairs
        
        -- Custom comparison handler for the sort. Text a should go after text b if
        -- it's the same length and has a greater lexical value or it's shorter than b.
        -- (The lexical sort direction isn't really relevant. It's just to group equal texts.)
        on isGreater(a, b)
            set aLen to a's length
            set bLen to b's length
            if (aLen = bLen) then return (a > b) -- or (b < a)!
            return (aLen < bLen)
        end isGreater
    end script
    
    set wordCount to (count o's wordList)
    ignoring case
        -- Replace the words in the groupingTexts list with sorted-character versions.
        repeat with i from 1 to wordCount
            set chrs to o's groupingTexts's item i's characters
            tell sorter to sort(chrs, 1, -1, {})
            set o's groupingTexts's item i to join(chrs, "")
        end repeat
        -- Sort the list descending by text length and ascending (say) by value
        -- within lengths. Echo the moves in the original word list.
        tell sorter to sort(o's groupingTexts, 1, wordCount, {comparer:o, slave:{o's wordList}})
        
        -- Work through the runs of grouping texts, starting with the longest texts.
        set i to 1
        set currentText to beginning of o's groupingTexts
        repeat with j from 2 to (wordCount)
            set thisText to o's groupingTexts's item j
            if (thisText is not currentText) then
                if (j - i > 1) then tell o to testPairs(i, j - 1)
                set currentText to thisText
                set i to j
            end if
            -- Stop on reaching a text that's shorter than any derangement(s) found.
            if ((count thisText) < o's derangementLength) then exit repeat
        end repeat
        if (j > i) then tell o to testPairs(i, j)
    end ignoring
    
    return o's output
end longestDerangedAnagrams

local wordFile, wordList
set wordFile to ((path to desktop as text) & "www.rosettacode.org:unixdict.txt") as «class furl»
set wordList to paragraphs of (read wordFile as «class utf8»)
return longestDerangedAnagrams(wordList)
Output:
{{"excitation", "intoxicate"}}

ARM Assembly

Works with: as version Raspberry Pi
or android 32 bits with application Termux
/* ARM assembly Raspberry PI  */
/*  program anaderan.s   */

 /* REMARK 1 : this program use routines in a include file 
   see task Include a file language arm assembly 
   for the routine affichageMess conversion10 
   see at end of this program the instruction include */
/* for constantes see task include a file in arm assembly */
/************************************/
/* Constantes                       */
/************************************/
.include "../constantes.inc"
 
.equ MAXI,         50000
.equ BUFFERSIZE,   300000
.equ READ, 3                            @ system call 
.equ OPEN, 5                            @ system call
.equ CLOSE, 6                           @ system call
.equ O_RDWR,  0x0002                    @ open for reading and writing

/*********************************/
/* Structures                    */
/*********************************/
/* this structure has size multiple de 4 */
    .struct  0
Word_Ptr_buffer:                           @ pointer word buffer
    .struct  Word_Ptr_buffer + 4 
Word_Ptr_sorted:                           @ pointer word sorted letters
    .struct  Word_Ptr_sorted + 4 
Word_length:                               @ word length
    .struct  Word_length + 4 
Word_top:                                  @ top
    .struct  Word_top + 4 
Word_end:
/*********************************/
/* Initialized data              */
/*********************************/
.data
szFileName:           .asciz "./unixdict.txt"
//szFileName:           .asciz "./listwordT.txt"
szMessErreur:         .asciz "FILE ERROR."
szMessStart:          .asciz "Program 32 bits start.\n"
szCarriageReturn:     .asciz "\n"
szMessSpace:          .asciz " "

ptBuffer1:            .int sBuffer1
/*********************************/
/* UnInitialized data            */
/*********************************/
.bss
tbStWords:                  .skip Word_end * MAXI
iNBword:                    .skip 4
sBuffer:                    .skip BUFFERSIZE
sBuffer1:                   .skip BUFFERSIZE

/*********************************/
/*  code section                 */
/*********************************/
.text
.global main 
main:                      @ entry of program 
    ldr r0,iAdrszMessStart
    bl affichageMess
    mov r4,#0              @ loop indice
    ldr r0,iAdrszFileName  @ file name
    mov r1,#O_RDWR         @ flags
    mov r2,#0              @ mode
    mov r7,#OPEN           @ 
    svc 0 
    cmp r0,#0              @ error open
    ble 99f
    mov r8,r0              @ FD save Fd
    ldr r1,iAdrsBuffer     @ buffer address
    ldr r2,iSizeBuf        @ buffersize
    mov r7, #READ
    svc 0 
    cmp r0,#0              @ error read ?
    blt 99f
    mov r5,r0              @ save size read bytes
    ldr r4,iAdrsBuffer     @ buffer address
    ldr r0,iAdrsBuffer     @ start word address
    mov r2,#0
    mov r1,#0              @ word length
1:
    cmp r2,r5
    bge 2f
    ldrb r3,[r4,r2]
    cmp r3,#0xD            @ end word ?
    addne r1,r1,#1         @ increment word length
    addne r2,r2,#1         @ increment indice
    bne 1b                 @ and loop
    mov r3,#0
    strb r3,[r4,r2]        @ store final zero
    bl anaWord             @ sort word letters
    add r2,r2,#2           @ jump OD and 0A 
    add r0,r4,r2           @ new address begin word
    mov r1,#0              @ init length
    b 1b                   @ and loop
    
2:
    mov r3,#0              @ last word
    strb r3,[r4,r2]
    bl anaWord
    
    mov r0,r8              @ file Fd
    mov r7, #CLOSE
    svc 0 
    cmp r0,#0              @ error close ?
    blt 99f
    
    ldr r0,iAdrtbStWords   @ array structure words address
    mov r1,#0              @ first indice
    ldr r2,iAdriNBword
    ldr r2,[r2]            @ last indice
    bl triRapide           @ quick sort
    ldr r4,iAdrtbStWords   @ array structure words address

    mov r0,r4
    mov r9,r2                     @ size word array
    mov r8,#0                     @ indice first occurence
    ldr r3,[r4,#Word_Ptr_sorted]  @ load first value
    mov r2,#1                     @ loop indice
    mov r10,#Word_end             @ words structure size
    mov r12,#0                    @ max word length
3:                                @ begin loop 
    mla r7,r2,r10,r4              @ compute array index
    ldr r5,[r7,#Word_Ptr_sorted]  @ load next value
    mov r0,r3
    mov r1,r5
    bl comparStrings
    cmp r0,#0                     @ sorted strings equal ?
    bne 4f
    mla r11,r8,r10,r4
    ldr r0,[r11,#Word_Ptr_buffer] @ address word 1
    ldr r1,[r7,#Word_Ptr_buffer]  @ address word 2
    bl controlLetters
    cmp r0,#0                     @ not ok ?
    beq 5f
    mov r0,#1                     @ letters ok
    str r0,[r7,#Word_top]         @ store top in first occurence
    str r0,[r11,#Word_top]        @ store top in current occurence
    ldr r0,[r7,#Word_length]      @ word length
    cmp r0,r12                    @ compare maxi
    movgt r12,r0                  @ yes  length word -> value max
    b 5f
4:                                @ no
    mov r0,r8
    add r8,r8,#1                  @ init index  new first occurence
    mla r11,r8,r10,r4             @ compute array index
    ldr r3,[r11,#Word_Ptr_sorted] @ init value new first occurence
    mov r2,r0                     @ reprise au debut de la sequence
5:
    add r2,r2,#1                  @ increment indice
    cmp r2,r9                     @ end word array ?
    blt 3b                        @ no -> loop
    
    mov r2,#0                     @ raz indice
    ldr r4,iAdrtbStWords          @ array structure words address
6:                                @ begin display loop
    mla r11,r2,r10,r4             @ compute array index
    ldr r6,[r11,#Word_top]        @ load top
    cmp r6,#0                     @ top ok ?
    beq 7f
    ldr r6,[r11,#Word_length]     @ load length
    cmp r6,r12                    @ compare maxi
    bne 7f
    ldr r0,[r11,#Word_Ptr_buffer] @ load address first word
    bl affichageMess              @ display first word
    add r2,r2,#1                  @ increment indice

    mla r11,r2,r10,r4             @ compute array index
    ldr r6,[r11,#Word_top]        @ load top
    cmp r6,#0                     @ top ok ?
    beq 7f
    ldr r0,iAdrszMessSpace
    bl affichageMess
    ldr r0,[r11,#Word_Ptr_buffer] @ load address other word
    bl affichageMess              @ display second word

    ldr r0,iAdrszCarriageReturn
    bl affichageMess
7:
    add r2,r2,#1                  @ increment indice
    cmp r2,r9                     @ maxi ?
    blt 6b                        @ no -> loop
    
    b 100f
99:                               @ display error
    ldr r1,iAdrszMessErreur
    bl displayError
    
100:                              @ standard end of the program 
    mov r0, #0                    @ return code
    mov r7, #EXIT                 @ request to exit program
    svc #0                        @ perform the system call
iAdrszCarriageReturn:        .int szCarriageReturn
iAdrszFileName:              .int szFileName
iAdrszMessErreur:            .int szMessErreur
iAdrsBuffer:                 .int sBuffer
iSizeBuf:                    .int BUFFERSIZE
iAdrszMessSpace:             .int szMessSpace
iAdrtbStWords:               .int tbStWords
iAdrszMessStart:             .int szMessStart
/******************************************************************/
/*     analizing word                                   */ 
/******************************************************************/
/*  r0  word address */
/*  r1 word length   */
anaWord:
    push {r1-r6,lr}
    mov r5,r0
    mov r6,r1
    //ldr r1,iAdrptTabBuffer
    ldr r1,iAdrtbStWords
    ldr r2,iAdriNBword
    ldr r3,[r2]
    mov r4,#Word_end
    mla r1,r3,r4,r1
    str r0,[r1,#Word_Ptr_buffer]
    mov r0,#0
    str r0,[r1,#Word_top]
    str r6,[r1,#Word_length]
    ldr r4,iAdrptBuffer1
    ldr r0,[r4]
    add r6,r6,r0
    add r6,r6,#1
    str r6,[r4]
    str r0,[r1,#Word_Ptr_sorted]
    
    add r3,r3,#1
    str r3,[r2]
    mov r1,r0
    mov r0,r5
    bl triLetters         @ sort word letters
    mov r2,#0
100:
    pop {r1-r6,pc}
iAdriNBword:            .int iNBword
iAdrptBuffer1:          .int ptBuffer1
/******************************************************************/
/*     sort word letters                                  */ 
/******************************************************************/
/* r0  address begin word */
/* r1  address recept array */
triLetters:
    push {r1-r7,lr}
    mov r2,#0
1:
    ldrb r3,[r0,r2]         @ load letter
    cmp r3,#0               @ end word ?
    beq 6f
    cmp r2,#0               @ first letter ?
    bne 2f
    strb r3,[r1,r2]         @ yes store in first position
    add r2,r2,#1            @ increment indice
    b 1b                    @ and loop
2:
    mov r4,#0
3:                          @ begin loop to search insertion position
    ldrb r5,[r1,r4]         @ load letter 
    cmp r3,r5               @ compare
    blt 4f                  @ to low -> insertion
    add r4,r4,#1            @ increment indice
    cmp r4,r2               @ compare to letters number in place
    blt 3b                  @ search loop
    strb r3,[r1,r2]         @ else store in last position
    add r2,r2,#1
    b 1b                    @ and loop
4:                          @ move first letters in one position 
    sub r6,r2,#1            @ start indice
5:
    ldrb r5,[r1,r6]         @ load letter
    add r7,r6,#1            @ store indice - 1
    strb r5,[r1,r7]         @ store letter
    sub r6,r6,#1            @ decrement indice
    cmp r6,r4               @ end ?
    bge 5b                  @ no loop
    strb r3,[r1,r4]         @ else store letter in free position
    add r2,r2,#1
    b 1b                    @ and loop
6: 
    mov r3,#0               @ final zéro
    strb r3,[r1,r2]
100:
    pop {r1-r7,pc}
/******************************************************************/
/*     control letters                                  */ 
/******************************************************************/
/* r0  address  word 1*/
/* r1  address  word 2 */
controlLetters:
    push {r1-r4,lr}
    mov r2,#0
1:
    ldrb r3,[r0,r2]      @ load a letter 
    cmp r3,#0            @ end word ?
    moveq r0,#1          @  yes it is OK
    beq 100f
    ldrb r4,[r1,r2]      @ load a letter word 2 same position
    cmp r3,r4            @ equal ?
    moveq r0,#0          @ yes -> not good
    beq 100f
    add r2,r2,#1
    b 1b
100:
    pop {r1-r4,pc}
/***************************************************/
/*   Appel récursif Tri Rapide quicksort           */
/***************************************************/
/* r0 contains the address of table */
/* r1 contains index of first item  */
/* r2 contains the number of elements  > 0  */
triRapide:
    push {r2-r5,lr}         @ save registers
    sub r2,#1               @ last item index
    cmp r1,r2               @ first > last ? 
    bge 100f                @ yes -> end
    mov r4,r0               @ save r0
    mov r5,r2               @ save r2
    bl partition1           @ cutting into 2 parts
    mov r2,r0               @ index partition
    mov r0,r4               @ table address
    bl triRapide            @ sort lower part
    mov r0,r4               @ table address
    add r1,r2,#1            @ index begin = index partition + 1
    add r2,r5,#1            @ number of elements
    bl triRapide            @ sort higter part
   
 100:                       @ end function
    pop {r2-r5,lr}          @ restaur  registers 
    bx lr                   @ return

/******************************************************************/
/*      Partition table elements                                */ 
/******************************************************************/
/* r0 contains the address of table */
/* r1 contains index of first item  */
/* r2 contains index of last item   */
partition1:
    push {r1-r9,lr}        @ save registers
    mov r8,r0               @ save address table 2
    mov r7,r2
    mov r9,#Word_end
    mla r3,r7,r9,r8
    ldr r6,[r3,#Word_Ptr_sorted]  @ load string address last index
    mov r4,r1               @ init with first index
    mov r5,r1               @ init with first index
1:                          @ begin loop
    mla r3,r5,r9,r8
    ldr r0,[r3,#Word_Ptr_sorted]   @ load current string address
    mov r1,r6                      @ first string address
    bl comparStrings        
    cmp r0,#0
    bge 2f
    mov r0,r8                @ current string < first string
    mov r1,r4                @ swap array
    mov r2,r5
    bl swapWord
    add r4,r4,#1             @ and increment index 1
2:
    add r5,r5,#1             @ increment index 2
    cmp r5,r7                @ end ?
    blt 1b                   @ no -> loop
    
    mov r0,r8                @ and swap array 
    mov r1,r4
    mov r2,r7
    bl swapWord

    mov r0,r4                @ return index partition
100:
    pop {r1-r9,lr}
    bx lr
/******************************************************************/
/*      Swap table elements                                */ 
/******************************************************************/
/* r0 contains the address of table */
/* r1 contains index 1  */
/* r2 contains index  2   */
swapWord:
    push {r1-r6,lr}        @ save registers
    mov r3,#Word_end
    mla r4,r1,r3,r0        @ compute array index
    mla r5,r2,r3,r0
    mov r6,#0
1:
    ldr r2,[r4,r6]         @ load 4 bytes
    ldr r3,[r5,r6]
    str r2,[r5,r6]         @ store 4 bytes
    str r3,[r4,r6]
    add r6,r6,#4           @ increment 4 bytes
    cmp r6,#Word_end       @ structure size is multiple to 4 
    blt 1b
100:
    pop {r1-r6,pc}
/************************************/       
/* Strings case sensitive comparisons  */
/************************************/      
/* r0 et r1 contains the address of strings */
/* return 0 in r0 if equals */
/* return -1 if string r0 < string r1 */
/* return 1  if string r0 > string r1 */
comparStrings:
    push {r1-r4}      @ save des registres
    mov r2,#0         @ counter
1:    
    ldrb r3,[r0,r2]   @ byte string 1
    ldrb r4,[r1,r2]   @ byte string 2
    cmp r3,r4
    movlt r0,#-1      @ small
    movgt r0,#1       @ greather  
    bne 100f          @ not equals
    cmp r3,#0         @ 0 end string
    moveq r0,#0       @ equals
    beq 100f          @ end string
    add r2,r2,#1      @ else add 1 in counter
    b 1b              @ and loop
100:
    pop {r1-r4}
    bx lr   

/***************************************************/
/*      ROUTINES INCLUDE                           */
/***************************************************/
.include "../affichage.inc"
Program 32 bits start.
excitation intoxicate

Arturo

isDeranged?: function [p][
    [a,b]: p
    loop 0..dec size a 'i [
        if a\[i] = b\[i] [return false]
    ]
    return true
]

wordset: map read.lines relative "unixdict.txt" => strip

anagrams: #[]

loop wordset 'word [
    anagram: sort to [:char] word
    unless key? anagrams anagram ->
        anagrams\[anagram]: new []

    anagrams\[anagram]: anagrams\[anagram] ++ word
]

deranged: select values anagrams 'anagram [ 2 = size anagram]

maxDeranged: ["" ""]

loop deranged 'd [
    if (size first d) > size first maxDeranged [
        pair: @[first d, last d]
        if isDeranged? pair [
            maxDeranged: pair
        ]
    ]
]

print maxDeranged
Output:
excitation intoxicate

AutoHotkey

Time := A_TickCount
SetWorkingDir %A_ScriptDir%  ; Ensures a consistent starting directory.
SetBatchLines   -1
Loop, Read, unixdict.txt
	StrOut .= StrLen(A_LoopReadLine) - 2 . "," . A_LoopReadLine . "`n"
Sort StrOut, N R
Loop, Parse, StrOut, `n, `r
{
	StringSplit, No_Let, A_Loopfield, `,
	if ( old1 = no_let1 )
		string .= old2 "`n"
	if ( old1 != no_let1 )
	{
		string := trim(string old2)
		if ( old2 != "" )
			Loop, Parse, string, `n, `r  ; Specifying `n prior to `r allows both Windows and Unix files to be Parsed.
				line_number := A_Index
		if ( line_number > 1 )
		{
			Loop, Parse, string, `n, `r
			{
				StringSplit, newstr, A_Loopfield, `, ; break the string based on Comma
				Loop, Parse, newstr2
					k .= A_LoopField " "
				Sort k, D%A_Space%
				k := RegExReplace( k, "\s", "" )
				file .= "`r`n" k . "," . newstr1 . "," . newstr2
				k =
			}
			Sort File
			Loop, Parse, File, `n, `r
			{
				if ( A_Loopfield != "" )
				{
					StringSplit, T_C, A_Loopfield, `,
					if ( old = T_C1 )
					{
						Loop, 1
						{
							Loop % T_C2
								if (SubStr(T_C3, A_Index, 1) = SubStr(old3, A_Index, 1))
									break 2
							Time := (A_tickcount - Time)/1000
							MsgBox % T_C3 " " old3 " in " Time . " seconds."
							ExitApp
						}
					}
					old := T_C1, old3 := T_C3
				}
			}
			file =
		}
		string =
	}
	old1 := no_let1, old2 := A_Loopfield
}
Output:
intoxicate excitation in 0.844000 seconds.

AWK

#!/bin/gawk -f
BEGIN{
FS=""
wordcount = 0
maxlength = 0

}
# hash generates the sorted sequence of characters in a word,
# so that the hashes for a pair of anagrams will be the same.
# Example: hash meat = aemt and hash team = aemt
function hash(myword, i,letters,myhash){
  split(myword,letters,"")
  asort(letters)
  for (i=1;i<=length(myword);i++) myhash=myhash letters[i]
  return myhash
}
# deranged checks two anagrems for derangement
function deranged(worda, wordb,  a,b,i,n,len){
  n=0
  len=split(worda,a,"")
  split(wordb,b,"")
  for (i=len; i>=1; i--){
      if (a[i] == b[i]) n = n+1
  }
  return n==0
}

# field separator null makes gawk split input record character by character.
# the split function works the same way
{ 
  wordcount = wordcount + 1
  fullword[wordcount]=$0
  bylength[length($0)]=bylength[length($0)]  wordcount "|"
  if (length($0) > maxlength) maxlength = length($0)
}

END{
 for (len=maxlength; len>1; len--){ 
  numwords=split(bylength[len],words,"|")
  split("",hashed)
 split("",anagrams)
  for (i=1;i<=numwords;i++){
#   make lists of anagrams in hashed
      myword = fullword[words[i]]
      myhash = hash(myword)
      hashed[myhash] = hashed[myhash] myword " "
  }
# check anagrams for derangement
 for (myhash in hashed){
     n = split(hashed[myhash],anagrams," ")
     for (i=1; i< n; i++)
         for (j=i+1; j<=n; j++){
             if(deranged(anagrams[i],anagrams[j])) found = found anagrams[i] " " anagrams[j] " "
          }
  }
 if (length(found) > 0 ) print "deranged: " found
  if (length(found) > 0) exit
 }
}

On my system, this awk-file is located at /usr/local/bin/deranged, so it can be invoked with:

deranged /tmp/unixdict.txt 

Regular invocation would be:

gawk -f deranged.awk /tmp/unixdict.txt
Output:
deranged: excitation intoxicate

BASIC

BaCon

DECLARE idx$ ASSOC STRING

FUNCTION Deranged(a$, b$)
    FOR i = 1 TO LEN(a$)
        IF MID$(a$, i, 1) = MID$(b$, i, 1) THEN RETURN FALSE
    NEXT
    RETURN TRUE
END FUNCTION

FOR w$ IN LOAD$(DIRNAME$(ME$) & "/unixdict.txt") STEP NL$
    set$ = EXTRACT$(SORT$(EXPLODE$(w$, 1)), " ")
    idx$(set$) = APPEND$(idx$(set$), 0, w$)
NEXT

FOR w$ IN OBTAIN$(idx$)
    FOR x = 1 TO AMOUNT(idx$(w$))
        FOR y = x+1 TO AMOUNT(idx$(w$))
            IF Deranged(TOKEN$(idx$(w$), x), TOKEN$(idx$(w$), y)) AND LEN(TOKEN$(idx$(w$), x)) > current THEN
                current = LEN(TOKEN$(idx$(w$), x))
                an1$ = TOKEN$(idx$(w$), x)
                an2$ = TOKEN$(idx$(w$), y)
            END IF
        NEXT
    NEXT
NEXT

PRINT "Maximum deranged anagrams: ", an1$, " and ", an2$

PRINT NL$, "Total time: ", TIMER, " msecs.", NL$
Output:
Maximum deranged anagrams: excitation and intoxicate

Total time: 75 msecs.

BBC BASIC

      INSTALL @lib$+"SORTLIB"
      Sort% = FN_sortinit(0,0)
      
      DIM dict$(26000), sort$(26000), indx%(26000)
      
      REM Load the dictionary:
      dict% = OPENIN("C:\unixdict.txt")
      IF dict%=0 ERROR 100, "No dictionary file"
      index% = 0
      REPEAT
        index% += 1
        dict$(index%) = GET$#dict%
        indx%(index%) = index%
      UNTIL EOF#dict%
      CLOSE #dict%
      Total% = index%
      
      TIME = 0
      REM Sort the letters in each word:
      FOR index% = 1 TO Total%
        sort$(index%) = FNsortstring(dict$(index%))
      NEXT
      
      REM Sort the sorted words:
      C% = Total%
      CALL Sort%, sort$(1), indx%(1)
      
      REM Find anagrams and deranged anagrams:
      maxlen% = 0
      maxidx% = 0
      FOR index% = 1 TO Total%-1
        IF sort$(index%) = sort$(index%+1) THEN
          One$ = dict$(indx%(index%))
          Two$ = dict$(indx%(index%+1))
          FOR c% = 1 TO LEN(One$)
            IF MID$(One$,c%,1) = MID$(Two$,c%,1) EXIT FOR
          NEXT
          IF c%>LEN(One$) IF c%>maxlen% maxlen% = c% : maxidx% = index%
        ENDIF
      NEXT
      
      PRINT "The longest deranged anagrams are '" dict$(indx%(maxidx%));
      PRINT "' and '" dict$(indx%(maxidx%+1)) "'"
      PRINT "(taking " ; TIME/100 " seconds)"
      END
      
      DEF FNsortstring(A$)
      LOCAL C%, a&()
      C% = LEN(A$)
      DIM a&(C%)
      $$^a&(0) = A$
      CALL Sort%, a&(0)
      = $$^a&(0)
Output:
The longest deranged anagrams are 'excitation' and 'intoxicate'
(taking 0.95 seconds)

Bracmat

The file is read into a single string, wordList. Then, in a while loop, each line is read and, in a nested loop, atomised into single letters. The letters are added together to create a sorted list that is the letter sum, the 'anagram fingerprint', of the word. To make sure that even single letter words create a sum of at least two terms, the sum is initialised with the empty string rather than zero. (Otherwise the words a and aaa later on would get the same fingerprint, the factors 1 and 3 being factored out.)

For the word bizarre the letter sum is (+a+b+e+2*r+z+i). The letter sum, with the word as the exponent ((+a+b+e+2*r+z+i)^bizarre) is prepended to a list unsorted. Somewhat later the word brazier also is prepended to the unsorted list. This word happens to have the same letter sum as bizarre, so these two words must be anagrams of each other. The program brings these two elements together by merge sorting the unsorted list, using Bracmat's Computer Algebra powers to normalise sums and products by sorting and combining like terms or factors. During the sort, all elements in the the unsorted list are multiplied together, combining factors with the same letter sums by adding their exponents together. So at some stage during sorting, the two elements (+a+b+e+2*r+z+i)^bizarre and (+a+b+e+2*r+z+i)^brazier are brought together in a product (+a+b+e+2*r+z+i)^bizarre*(+a+b+e+2*r+z+i)^brazier which immediately is transformed to the single factor (+a+b+e+2*r+z+i)^(bizarre+brazier). In the product of all elements the anagrams are to be found in the exponents consisting of a sum of at least two terms. To find the longest deranged anagrams, we traverse the product list to find all exponents with multiple words, check that the length of the first word is at least as long as the length of the longest deranged anagram up to now, and check each pair of words for being deranged. If a pair of deranged anagrams is found with more letters than previously found deranged anagrams, the earlier finds are forgotten. If the new anagrams are the same length, however, they are added to the output.

The Bracmat solution to the similar task anagrams skips the explicit merge sort and instead prepends new factors directly to the product one by one. Bracmat shuffles each new factor into place to keep the growing product normalized before continuing with the next word from the list. The result is exactly the same, but the running time becomes much longer.

  get$("unixdict.txt",STR):?wordList
& 1:?product
& :?unsorted
&   whl
  ' ( @(!wordList:(%?word:?letterString) \n ?wordList)
    & :?letterSum
    &   whl
      ' ( @(!letterString:%?letter ?letterString)
        &   (!letter:~#|str$(N !letter))+!letterSum
          : ?letterSum
        )
    & !letterSum^!word !unsorted:?unsorted
    )
& ( mergeSort
  =   newL L first second
    .   !arg:?L
      &   whl
        ' ( !L:% %
          & :?newL
          &   whl
            ' ( !L:%?first %?second ?L
              & !first*!second !newL:?newL
              )
          & !L !newL:?L
          )
      & !L
  )
& mergeSort$!unsorted:?product
& 0:?maxLength:?oldMaxLength
& :?derangedAnagrams
& ( deranged
  =   nextLetter Atail Btail
    .   !arg
      : ( (.)
        |   ( @(?:%@?nextLetter ?Atail)
            . @(?:(%@:~!nextLetter) ?Btail)
            )
          & deranged$(!Atail.!Btail)
        )
  )
& (   !product
    :   ?
      *   ?
        ^ ( %+%
          : @(%:? ([~<!maxLength:[?maxLength))+?
          :   ?
            + %@?anagramA
            + ?
            + %@?anagramB
            + ( ?
              & deranged$(!anagramA.!anagramB)
              &     (!anagramA.!anagramB)
                    (   !maxLength:>!oldMaxLength:?oldMaxLength
                      & 
                    | !derangedAnagrams
                    )
                : ?derangedAnagrams
              & ~
              )
          )
      * ?
  | out$!derangedAnagrams
  );
Output:
excitation.intoxicate

C

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <sys/types.h>
#include <fcntl.h>
#include <sys/stat.h>

// Letter lookup by frequency.  This is to reduce word insertion time.
const char *freq = "zqxjkvbpygfwmucldrhsnioate";
int char_to_idx[128];

// Trie structure of sorts
struct word {
	const char *w;
	struct word *next;
};

union node {
	union node *down[10];
	struct word *list[10];
};

int deranged(const char *s1, const char *s2)
{
	int i;
	for (i = 0; s1[i]; i++)
		if (s1[i] == s2[i]) return 0;
	return 1;
}

int count_letters(const char *s, unsigned char *c)
{
	int i, len;
	memset(c, 0, 26);
	for (len = i = 0; s[i]; i++) {
		if (s[i] < 'a' || s[i] > 'z')
			return 0;
		len++, c[char_to_idx[(unsigned char)s[i]]]++;
	}
	return len;
}

const char * insert(union node *root, const char *s, unsigned char *cnt)
{
	int i;
	union node *n;
	struct word *v, *w = 0;

	for (i = 0; i < 25; i++, root = n) {
		if (!(n = root->down[cnt[i]]))
			root->down[cnt[i]] = n = calloc(1, sizeof(union node));
	}

	w = malloc(sizeof(struct word));
	w->w = s;
	w->next = root->list[cnt[25]];
	root->list[cnt[25]] = w;

	for (v = w->next; v; v = v->next) {
		if (deranged(w->w, v->w))
			return v->w;
	}
	return 0;
}

int main(int c, char **v)
{
	int i, j = 0;
	char *words;
	struct stat st;

	int fd = open(c < 2 ? "unixdict.txt" : v[1], O_RDONLY);
	if (fstat(fd, &st) < 0) return 1;

	words = malloc(st.st_size);
	read(fd, words, st.st_size);
	close(fd);

	union node root = {{0}};
	unsigned char cnt[26];
	int best_len = 0;
	const char *b1, *b2;

	for (i = 0; freq[i]; i++)
		char_to_idx[(unsigned char)freq[i]] = i;

	/* count words, change newline to null */
	for (i = j = 0; i < st.st_size; i++) {
		if (words[i] != '\n') continue;
		words[i] = '\0';

		if (i - j > best_len) {
			count_letters(words + j, cnt);
			const char *match = insert(&root, words + j, cnt);

			if (match) {
				best_len = i - j;
				b1 = words + j;
				b2 = match;
			}
		}

		j = ++i;
	}

	if (best_len) printf("longest derangement: %s %s\n", b1, b2);

	return 0;
}
Output:
longest derangement: intoxicate excitation

C#

Library: System
Library: System.Linq
Library: System.IO
Works with: C sharp version 6
public static void Main()
{
    var lookupTable = File.ReadLines("unixdict.txt").ToLookup(line => AnagramKey(line));
    var query = from a in lookupTable
        orderby a.Key.Length descending
        let deranged = FindDeranged(a)
        where deranged != null
        select deranged[0] + " " + deranged[1];
    Console.WriteLine(query.FirstOrDefault());
}
	
static string AnagramKey(string word) => new string(word.OrderBy(c => c).ToArray());
	
static string[] FindDeranged(IEnumerable<string> anagrams) => (
    from first in anagrams
    from second in anagrams
    where !second.Equals(first)
        && Enumerable.Range(0, first.Length).All(i => first[i] != second[i])
    select new [] { first, second })
    .FirstOrDefault();
Output:
excitation intoxicate

C++

#include <algorithm>
#include <fstream>
#include <functional>
#include <iostream>
#include <map>
#include <numeric>
#include <set>
#include <string>

bool is_deranged(const std::string& left, const std::string& right)
{
    return (left.size() == right.size()) &&
        (std::inner_product(left.begin(), left.end(), right.begin(), 0, std::plus<int>(), std::equal_to<char>()) == 0);
}

int main()
{
    std::ifstream input("unixdict.txt");
    if (!input) {
        std::cerr << "can't open input file\n";
        return EXIT_FAILURE;
    }

    typedef std::set<std::string> WordList;
    typedef std::map<std::string, WordList> AnagraMap;
    AnagraMap anagrams;

    std::pair<std::string, std::string> result;
    size_t longest = 0;

    for (std::string value; input >> value; /**/) {
        std::string key(value);
        std::sort(key.begin(), key.end());

        if (longest < value.length()) { // is it a long candidate?
            if (0 < anagrams.count(key)) { // is it an anagram?
                for (const auto& prior : anagrams[key]) {
                    if (is_deranged(prior, value)) { // are they deranged?
                        result = std::make_pair(prior, value);
                        longest = value.length();
                    }
                }
            }
        }
        anagrams[key].insert(value);
    }

    std::cout << result.first << ' ' << result.second << '\n';
    return EXIT_SUCCESS;
}
Output:
excitation intoxicate

Clojure

(->> (slurp "unixdict.txt") ; words 
     (re-seq #"\w+")        ; |
     (group-by sort)        ; anagrams
     vals                   ; |
     (filter second)        ; |
     (remove #(some true? (apply map = %))) ; deranged
     (sort-by #(count (first %))) 
     last
     prn)
Output:
$ lein exec deranged.clj
["excitation" "intoxicate"]

COBOL

Works with: X/Open COBOL
******************************************************************
* COBOL solution to Anagrams Deranged challange
* The program was run on OpenCobolIDE
* Input data is stored in file 'Anagrams.txt' on my PC
******************************************************************
IDENTIFICATION DIVISION.
PROGRAM-ID. DERANGED.

ENVIRONMENT DIVISION.
INPUT-OUTPUT SECTION.
FILE-CONTROL.
    SELECT IN-FILE ASSIGN TO 'C:\Both\Rosetta\Anagrams.txt'
        ORGANIZATION IS LINE SEQUENTIAL.

DATA DIVISION.
FILE SECTION.
FD IN-FILE.
01 IN-RECORD               PIC X(22).

WORKING-STORAGE SECTION.
01  SWITCHES.
    05 WS-EOF              PIC X VALUE 'N'.
    05 WS-FND              PIC X VALUE 'N'.
    05 WS-EXIT             PIC X VALUE 'N'.

01  COUNTERS.
    05 WS-TOT-RECS         PIC 9(5) USAGE PACKED-DECIMAL VALUE 0.
    05 WS-SEL-RECS         PIC 9(5) USAGE PACKED-DECIMAL VALUE 0.
    05 WT-REC-NBR          PIC 9(5) USAGE PACKED-DECIMAL VALUE 0.

* Extra byte to guarentee a space at end - needed in sort logic.
01  WS-WORD-TEMP           PIC X(23).
01  FILLER REDEFINES WS-WORD-TEMP.
    05 WS-LETTER OCCURS 23 TIMES PIC X.
77  WS-LETTER-HLD          PIC X.

77  WS-WORD-IN             PIC X(22).
77  WS-WORD-KEY            PIC X(22).

01  WS-WORD-TABLE.
    05 WT-RECORD OCCURS 0 to 24000 TIMES
            DEPENDING ON WT-REC-NBR
            DESCENDING KEY IS WT-WORD-LEN
            INDEXED BY WT-IDX.
        10 WT-WORD-KEY     PIC X(22).
        10 WT-WORD-LEN     PIC 9(2).
        10 WT-ANAGRAM-CNT  PIC 9(5) USAGE PACKED-DECIMAL.
        10 WT-ANAGRAMS OCCURS 6 TIMES.
            15 WT-ANAGRAM  PIC X(22).

01  WS-WORD-TEMP1          PIC X(22).
01  FILLER REDEFINES WS-WORD-TEMP1.
    05 WS-LETTER1          PIC X OCCURS 22 TIMES.

01  WS-WORD-TEMP2          PIC X(22).
01  FILLER REDEFINES WS-WORD-TEMP2.
    05 WS-LETTER2 OCCURS 22 TIMES PIC X.

77  WS-I                   PIC 9(5) USAGE PACKED-DECIMAL.
77  WS-J                   PIC 9(5) USAGE PACKED-DECIMAL.
77  WS-K                   PIC 9(5) USAGE PACKED-DECIMAL.
77  WS-L                   PIC 9(5) USAGE PACKED-DECIMAL.
77  WS-BEG                 PIC 9(5) USAGE PACKED-DECIMAL.
77  WS-MAX                 PIC 9(5) USAGE PACKED-DECIMAL.

PROCEDURE DIVISION.
000-MAIN.
    PERFORM 100-INITIALIZE.
    PERFORM 200-PROCESS-RECORD UNTIL WS-EOF = 'Y'.
    SORT WT-RECORD ON DESCENDING KEY WT-WORD-LEN.
    PERFORM 500-FIND-DERANGED.
    PERFORM 900-TERMINATE.
    STOP RUN.

100-INITIALIZE.
    OPEN INPUT IN-FILE.
    PERFORM 150-READ-RECORD.

150-READ-RECORD.
    READ IN-FILE INTO WS-WORD-IN
        AT END
            MOVE 'Y' TO WS-EOF
        NOT AT END
            COMPUTE WS-TOT-RECS = WS-TOT-RECS + 1
    END-READ.

200-PROCESS-RECORD.
    IF WS-WORD-IN IS ALPHABETIC
        COMPUTE WS-SEL-RECS = WS-SEL-RECS + 1 END-COMPUTE
        MOVE WS-WORD-IN TO WS-WORD-TEMP
        PERFORM 300-SORT-WORD
        MOVE WS-WORD-TEMP TO WS-WORD-KEY
        PERFORM 400-ADD-TO-TABLE
    END-IF.
    PERFORM 150-READ-RECORD.

* bubble sort:
300-SORT-WORD.
    PERFORM VARYING WS-MAX FROM 1 BY 1
        UNTIL WS-LETTER(WS-MAX) = SPACE
    END-PERFORM.
    PERFORM VARYING WS-I FROM 1 BY 1 UNTIL WS-I = WS-MAX
        PERFORM VARYING WS-J FROM WS-I BY 1
          UNTIL WS-J > WS-MAX - 1
            IF WS-LETTER(WS-J) < WS-LETTER(WS-I) THEN
                MOVE WS-LETTER(WS-I) TO WS-LETTER-HLD
                MOVE WS-LETTER(WS-J) TO WS-LETTER(WS-I)
                MOVE WS-LETTER-HLD   TO WS-LETTER(WS-J)
            END-IF
        END-PERFORM
    END-PERFORM.

400-ADD-TO-TABLE.
    SET WT-IDX TO 1.
    SEARCH WT-RECORD
        AT END
            PERFORM 420-ADD-RECORD
        WHEN WT-WORD-KEY(WT-IDX) = WS-WORD-KEY
            PERFORM 440-UPDATE-RECORD
    END-SEARCH.

420-ADD-RECORD.
    ADD 1 To WT-REC-NBR.
    MOVE WS-WORD-KEY TO WT-WORD-KEY(WT-REC-NBR).
    COMPUTE WT-WORD-LEN(WT-REC-NBR) = WS-MAX - 1 END-COMPUTE.
    MOVE 1 TO WT-ANAGRAM-CNT(WT-REC-NBR).
    MOVE WS-WORD-IN TO
        WT-ANAGRAM(WT-REC-NBR, WT-ANAGRAM-CNT(WT-REC-NBR)).

440-UPDATE-RECORD.
    ADD 1 TO WT-ANAGRAM-CNT(WT-IDX).
    MOVE WS-WORD-IN TO
        WT-ANAGRAM(WT-IDX, WT-ANAGRAM-CNT(WT-IDX)).

500-FIND-DERANGED.
    PERFORM VARYING WS-I FROM 1 BY 1
      UNTIL WS-I > WT-REC-NBR OR WS-FND = 'Y'
        PERFORM VARYING WS-J FROM 1 BY 1
          UNTIL WS-J > WT-ANAGRAM-CNT(WS-I) - 1 OR WS-FND = 'Y'
            COMPUTE WS-BEG = WS-J + 1 END-COMPUTE
            PERFORM VARYING WS-K FROM WS-BEG BY 1
              UNTIL WS-K > WT-ANAGRAM-CNT(WS-I) OR WS-FND = 'Y'
                MOVE WT-ANAGRAM(WS-I, WS-J) TO WS-WORD-TEMP1
                MOVE WT-ANAGRAM(WS-I, WS-K) To WS-WORD-TEMP2
                PERFORM 650-CHECK-DERANGED
            END-PERFORM
        END-PERFORM
    END-PERFORM.

650-CHECK-DERANGED.
    MOVE 'N' TO WS-EXIT.
    PERFORM VARYING WS-L FROM 1 BY 1
        UNTIL WS-L > WT-WORD-LEN(WS-I) OR WS-EXIT = 'Y'
            IF WS-LETTER1(WS-L) = WS-LETTER2(WS-L)
                MOVE 'Y' TO WS-EXIT
            END-IF
    END-PERFORM.
    IF WS-EXIT = 'N'
        DISPLAY
            WS-WORD-TEMP1(1:WT-WORD-LEN(WS-I)) ' ' WS-WORD-TEMP2
        END-DISPLAY
        MOVE 'Y' TO WS-FND
    END-IF.

900-TERMINATE.
    DISPLAY 'RECORDS READ: ' WS-TOT-RECS.
    DISPLAY 'RECORDS SELECTED ' WS-SEL-RECS.
    DISPLAY 'RECORD KEYS: ' WT-REC-NBR.
    CLOSE IN-FILE.

END PROGRAM DERANGED.

*>   OUTPUT:

*>      excitation intoxicate
*>      RECORDS READ: 25104
*>      RECORDS SELECTED 24978
*>      RECORD KEYS: 23441

*>   BUBBLE SORT REFERENCE:
*>   https://mainframegeek.wordpress.com/tag/bubble-sort-in-cobol

CoffeeScript

This example was tested with node.js.

http = require 'http'

is_derangement = (word1, word2) ->
  for c, i in word1
    return false if c == word2[i]
  true

show_longest_derangement = (word_lst) ->
  anagrams = {}
  max_len = 0
  
  for word in word_lst
    continue if word.length < max_len
    key = word.split('').sort().join('')
    if anagrams[key]
      for prior in anagrams[key]
        if is_derangement(prior, word)
          max_len = word.length
          result = [prior, word]
    else
      anagrams[key] = []
    anagrams[key].push word
    
  console.log "Longest derangement: #{result.join ' '}"

get_word_list = (process) ->
  options =
    host: "www.puzzlers.org"
    path: "/pub/wordlists/unixdict.txt"
  
  req = http.request options, (res) ->
    s = ''
    res.on 'data', (chunk) ->
      s += chunk
    res.on 'end', ->
      process s.split '\n'
  req.end()
  
get_word_list show_longest_derangement
Output:
> coffee anagrams.coffee 
Longest derangement: excitation intoxicate

Common Lisp

(defun read-words (file)
  (with-open-file (stream file)
    (loop with w = "" while w collect (setf w (read-line stream nil)))))
 
(defun deranged (a b)
  (loop for ac across a for bc across b always (char/= ac bc)))

(defun longest-deranged (file)
  (let ((h (make-hash-table :test #'equal))
	(wordlist (sort (read-words file)
			#'(lambda (x y) (> (length x) (length y))))))
    (loop for w in wordlist do
	  (let* ((ws (sort (copy-seq w) #'char<))
		 (l (gethash ws h)))
	    (loop for w1 in l do
		  (if (deranged w w1)
		    (return-from longest-deranged (list w w1))))
	    (setf (gethash ws h) (cons w l))))))

(format t "~{~A~%~^~}" (longest-deranged "unixdict.txt"))
Output:
intoxicate
excitation

D

Short Version

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

    string[][dstring] anags;
    foreach (const w; "unixdict.txt".readText.split)
        anags[w.array.sort().release.idup] ~= w;

    anags
    .byValue
    .map!(words => words.cartesianProduct(words)
                   .filter!q{ a[].equal!q{ a != b }})
    .join
    .minPos!q{ a[0].length > b[0].length }[0]
    .writeln;
}
Output:
Tuple!(string, string)("intoxicate", "excitation")

Runtime: about 0.11 seconds with LDC2 compiler.

Using const(ubytes)[] instead of dstrings gives a runtime of about 0.07 seconds:

    string[][ubyte[]] anags;
    foreach (const w; "unixdict.txt".readText.split)
        anags[w.dup.representation.sort().release.assumeUnique] ~= w;

Faster Version

import std.stdio, std.file, std.algorithm, std.string, std.array,
       std.functional, std.exception;

string[2][] findDeranged(in string[] words) pure nothrow /*@safe*/ {
    // return words
    //        .map!representation
    //        .pairwise
    //        .filter!(ww => ww[].equal!q{ a != b });
    typeof(return) result;
    foreach (immutable i, immutable w1; words)
        foreach (immutable w2; words[i + 1 .. $])
            if (w1.representation.equal!q{ a != b }(w2.representation))
                result ~= [w1, w2];
    return result;
}

void main() /*@safe*/ {
    Appender!(string[])[30] wClasses;
    foreach (const w; "unixdict.txt".readText.splitter)
        wClasses[$ - w.length] ~= w;

    foreach (const ws; wClasses[].map!q{ a.data }.filter!(not!empty)) {
        string[][const ubyte[]] anags; // Assume ASCII input.
        foreach (immutable w; ws)
            anags[w.dup.representation.sort().release.assumeUnique] ~= w;
        auto pairs = anags.byValue.map!findDeranged.joiner;
        if (!pairs.empty)
            return writefln("Longest deranged: %-(%s %)", pairs.front);
    }
}
Output:
Longest deranged: excitation intoxicate

Runtime: about 0.03 seconds.

Delphi

program Anagrams_Deranged;

{$APPTYPE CONSOLE}

{$R *.res}

uses
  System.SysUtils,
  System.Classes,
  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 IsDerangement(word1, word2: string): Boolean;
var
  i: Integer;
begin
  for i := 1 to word1.Length do
    if word1[i] = word2[i] then
      exit(False);
  Result := True;
end;

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

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

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

  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 IsDerangement(Dict[Index], Dict[Index - Count]) then
      begin
        words := Dict[Index] + ' - ' + Dict[Index - Count];
        Break;
      end;
      Inc(Count);
    end
    else
    begin
      Dec(Index, Count);
      Count := 1;
    end;
  end;

  StopWatch.Stop;

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

  writeln(#10'Longest derangement words are:'#10#10, words);

  Dict.Free;
  Readln;
end.
Output:
Time pass: 455 ms [i7-4500U Windows 7]

Longest derangement words are:

intoxicate - excitation

EchoLisp

For a change, we use the french dictionary included in EchoLisp package.

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


(define H (make-hash))

(define (deranged w1 w2)
	(for ((a w1) (b w2))
		#:break (string=? a b) => #f
		#t))

(define (anagrams (normal) (name) (twins))
(for ((w *words*))
	(set! name (word-name w))
	(set! normal (list->string (list-sort string<? (string->list name))))
	(set! twins (or (hash-ref H normal) null))
	#:continue (member name twins)
	#:when  (or (null? twins)  (for/or ((anagram twins)) (deranged name anagram)))
	(hash-set H normal (cons name twins))))
		

(define (task (lmin 8))
(anagrams)
(for ((lw (hash-values H))) ;; lw = list of words
	#:continue (= (length lw) 1)
	#:continue (< (string-length (first lw)) lmin)
	(set! lmin (string-length (first lw)))
	(write lmin) (for-each write lw)
	(writeln)))
Output:
(lib 'dico.fr.no-accent) ;; 209315 words into *words* table
(task)

[...]
13 tractionnaire contrariaient
13 ressourcement contremesures
13 saintsimonien inseminations
14 tergiversation interrogatives
14 suralimenterai mineralisateur
14 transoceaniens reconnaissante 

(lib 'dico.en ) ;;  235886 words
(task) 

[...]

12 reaccomplish accomplisher
12 chromatician achromatinic
12 unaccumulate acutenaculum
14 charlatanistic antarchistical
15 megachiropteran cinematographer
17 misconstitutional constitutionalism

Eiffel

class
	ANAGRAMS_DERANGED

create
	make

feature

	make
			-- Longest deranged anagram.
		local
			deranged_anagrams: LINKED_LIST [STRING]
			count: INTEGER
		do
			read_wordlist
			across
				words as wo
			loop
				deranged_anagrams := check_list_for_deranged (wo.item)
				if not deranged_anagrams.is_empty and deranged_anagrams [1].count > count then
					count := deranged_anagrams [1].count
				end
				wo.item.wipe_out
				wo.item.append (deranged_anagrams)
			end
			across
				words as wo
			loop
				across
					wo.item as w
				loop
					if w.item.count = count then
						io.put_string (w.item + "%T")
						io.new_line
					end
				end
			end
		end

	original_list: STRING = "unixdict.txt"

feature {NONE}

	check_list_for_deranged (list: LINKED_LIST [STRING]): LINKED_LIST [STRING]
			-- Deranged anagrams in 'list'.
		do
			create Result.make
			across
				1 |..| list.count as i
			loop
				across
					(i.item + 1) |..| list.count as j
				loop
					if check_for_deranged (list [i.item], list [j.item]) then
						Result.extend (list [i.item])
						Result.extend (list [j.item])
					end
				end
			end
		end

	check_for_deranged (a, b: STRING): BOOLEAN
			-- Are 'a' and 'b' deranged anagrams?
		local
			n: INTEGER
		do
			across
				1 |..| a.count as i
			loop
				if a [i.item] = b [i.item] then
					n := n + 1
				end
			end
			Result := n = 0
		end

	read_wordlist
			-- Hashtable 'words' with alphabetically sorted Strings used as key.
		local
			l_file: PLAIN_TEXT_FILE
			sorted: STRING
			empty_list: LINKED_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 words.make (wordlist.count)
			across
				wordlist as w
			loop
				create empty_list.make
				sorted := sort_letters (w.item)
				words.put (empty_list, sorted)
				if attached words.at (sorted) as ana then
					ana.extend (w.item)
				end
			end
		end

	wordlist: LIST [STRING]

	sort_letters (word: STRING): STRING
			--Alphabetically sorted.
		local
			letters: SORTED_TWO_WAY_LIST [STRING]
		do
			create letters.make
			create Result.make_empty
			across
				1 |..| word.count as i
			loop
				letters.extend (word.at (i.item).out)
			end
			across
				letters as s
			loop
				Result.append (s.item)
			end
		end

	words: HASH_TABLE [LINKED_LIST [STRING], STRING]

end
Output:
excitation
intoxicate

Elixir

Translation of: Ruby
defmodule Anagrams do
  def deranged(fname) do
    File.read!(fname)
    |> String.split
    |> Enum.map(fn word -> to_charlist(word) end)
    |> Enum.group_by(fn word -> Enum.sort(word) end)
    |> Enum.filter(fn {_,words} -> length(words) > 1 end)
    |> Enum.sort_by(fn {key,_} -> -length(key) end)
    |> Enum.find(fn {_,words} -> find_derangements(words) end)
  end
  
  defp find_derangements(words) do
    comb(words,2) |> Enum.find(fn [a,b] -> deranged?(a,b) end)
  end
  
  defp deranged?(a,b) do
    Enum.zip(a, b) |> Enum.all?(fn {chr_a,chr_b} -> chr_a != chr_b end)
  end
  
  defp comb(_, 0), do: [[]]
  defp comb([], _), do: []
  defp comb([h|t], m) do
    (for l <- comb(t, m-1), do: [h|l]) ++ comb(t, m)
  end
end

case Anagrams.deranged("/work/unixdict.txt") do
  {_, words} -> IO.puts "Longest derangement anagram: #{inspect words}"
  _          -> IO.puts "derangement anagram: nothing"
end
Output:
Longest derangement anagram: ["intoxicate", "excitation"]

Erlang

Using anagrams:fetch/2 from Anagrams and init_http/0 from Rosetta_Code/Find_unimplemented_tasks. Exporting words_from_url/1 to Ordered_words.

-module( anagrams_deranged ).
-export( [task/0, words_from_url/1] ).

task() ->
       find_unimplemented_tasks:init_http(),
       Words = words_from_url( "http://www.puzzlers.org/pub/wordlists/unixdict.txt" ),
       Anagram_dict = anagrams:fetch( Words, dict:new() ),
       Deranged_anagrams = deranged_anagrams( Anagram_dict ),
       {_Length, Longest_anagrams} = dict:fold( fun keep_longest/3, {0, []}, Deranged_anagrams ),
       Longest_anagrams.

words_from_url( URL ) ->
	{ok, {{_HTTP, 200, "OK"}, _Headers, Body}} = httpc:request( URL ),
	string:tokens( Body, "\n" ).


deranged_anagrams( Dict ) ->
        Deranged_dict = dict:map( fun deranged_words/2, Dict ),
        dict:filter( fun is_anagram/2, Deranged_dict ).

deranged_words( _Key, [H | T] ) ->
        [{H, X} || X <- T, is_deranged_word(H, X)].

keep_longest( _Key, [{One, _} | _]=New, {Length, Acc} ) ->
        keep_longest_new( erlang:length(One), Length, New, Acc ).

keep_longest_new( New_length, Acc_length, New, _Acc ) when New_length > Acc_length ->
        {New_length, New};
keep_longest_new( New_length, Acc_length, New, Acc ) when New_length =:= Acc_length ->
        {Acc_length, Acc ++ New};
keep_longest_new( _New_length, Acc_length, _New, Acc ) ->
        {Acc_length, Acc}.

is_anagram( _Key, [] ) -> false;
is_anagram( _Key, _Value ) -> true.

is_deranged_word( Word1, Word2 ) ->
        lists:all( fun is_deranged_char/1, lists:zip(Word1, Word2) ).

is_deranged_char( {One, Two} ) -> One =/= Two.
Output:
8> anagrams_deranged:task().
[{"excitation","intoxicate"}]

F#

open System;

let keyIsSortedWord = Seq.sort >> Seq.toArray >> String
let isDeranged = Seq.forall2 (<>)

let rec pairs acc l = function
| [] -> acc
| x::rtail ->
    pairs (acc @ List.fold (fun acc y -> (y, x)::acc) [] l) (x::l) rtail

 
[<EntryPoint>]
let main args =
    System.IO.File.ReadAllLines("unixdict.txt")
    |> Seq.groupBy keyIsSortedWord
    |> Seq.fold (fun (len, found) (key, words) ->
        if String.length key < len || Seq.length words < 2 then (len, found)
        else
            let d = List.filter (fun (a, b) -> isDeranged a b) (pairs [] [] (List.ofSeq words))
            if List.length d = 0 then (len, found)
            elif String.length key = len then (len, found @ d)
            else (String.length key, d)
    ) (0, [])
    |> snd
    |> printfn "%A"
    0
Output:
[("excitation", "intoxicate")]

Factor

USING: assocs fry io.encodings.utf8 io.files kernel math
math.combinatorics sequences sorting strings ;
IN: rosettacode.deranged-anagrams

: derangement? ( str1 str2 -- ? ) [ = not ] 2all? ;
: derangements ( seq -- seq )
    2 [ first2 derangement? ] filter-combinations ;

: parse-dict-file ( path -- hash )
    utf8 file-lines 
    H{ } clone [
        '[ 
            [ natural-sort >string ] keep
            _ [ swap suffix  ] with change-at
        ] each
    ] keep ;

: anagrams ( hash -- seq ) [ nip length 1 > ] assoc-filter values ;

: deranged-anagrams ( path -- seq )
    parse-dict-file anagrams [ derangements ] map concat ;

: longest-deranged-anagrams ( path -- anagrams )
    deranged-anagrams [ first length ] sort-with last ;
   "unixdict.txt" longest-deranged-anagrams .
   { "excitation" "intoxicate" }

FreeBASIC

' FB 1.05.0 Win64

Type IndexedWord
  As String word
  As Integer index
End Type

' selection sort, quick enough for sorting small number of letters
Sub sortWord(s As String)
  Dim As Integer i, j, m, n = Len(s)
  For i = 0 To n - 2
    m = i
    For j = i + 1 To n - 1
      If s[j] < s[m] Then m = j
    Next j
    If m <> i Then Swap s[i], s[m]
  Next i
End Sub

' quicksort for sorting whole dictionary of IndexedWord instances by sorted word
Sub quicksort(a() As IndexedWord, first As Integer, last As Integer)
  Dim As Integer length = last - first + 1
  If length < 2 Then Return 
  Dim pivot As String = a(first + length\ 2).word
  Dim lft As Integer = first 
  Dim rgt As Integer = last 
  While lft <= rgt
    While a(lft).word < pivot
      lft +=1
    Wend
    While a(rgt).word > pivot
      rgt -= 1
    Wend
    If lft <= rgt Then
       Swap a(lft), a(rgt)
       lft += 1
       rgt -= 1
    End If 
  Wend
  quicksort(a(), first, rgt)
  quicksort(a(), lft, last)
End Sub

Function isDeranged(s1 As String, s2 As String) As Boolean
  For i As Integer = 0 To Len(s1) - 1
    If s1[i] = s2[i] Then Return False
  Next
  Return True
End Function 

Dim t As Double = timer
Dim As String w()  '' array to hold actual words
Open "undict.txt" For Input As #1
Dim count As Integer = 0
While Not Eof(1)
  count +=1
  Redim Preserve w(1 To count)
  Line Input #1, w(count)
Wend
Close #1

Dim As IndexedWord iw(1 To count) '' array to hold sorted words and their index into w() 
Dim word As String
For i As Integer = 1 To count
  word = w(i)
  sortWord(word)
  iw(i).word = word
  iw(i).index = i 
Next
quickSort iw(), 1, count  '' sort the IndexedWord array by sorted word

Dim As Integer startIndex = 1, maxLength, ub
Dim As Integer maxIndex()
Dim As IndexedWord iWord = iw(1)
maxLength = 0

For i As Integer = 2 To count
  If iWord.word = iw(i).word Then
    If isDeranged(w(iword.index), w(iw(i).index)) Then       
      If startIndex + 1 < i Then Swap iw(i), iw(startIndex + 1)
      If Len(iWord.word) > maxLength Then      
        maxLength = Len(iWord.word)
        Erase maxIndex
        ub = 1
        Redim maxIndex(1 To ub)
        maxIndex(ub) = startIndex
        startIndex += 2 
        i = startIndex
        iWord = iw(i)
      ElseIf Len(iWord.word) = maxLength Then
        ub += 1
        Redim Preserve maxIndex(1 To ub)
        maxIndex(ub) = startIndex 
        startIndex += 2
        i = startIndex
        iWord = iw(i)
      End If      
    End If   
  ElseIf i = count Then
    Exit For
  Else 
    For j As Integer = i To count - 1
      iWord = iw(j)
      If Len(iWord.word) >= maxLength Then
        startIndex = j 
        i = startIndex
        Exit For
      End If
    Next      
  End If
Next

Print Str(count); " words in the dictionary"
Print "The deranged anagram pair(s) with the greatest length (namely"; maxLength; ") is:"
Print
Dim iws(1 To maxLength) As IndexedWord  '' array to hold each deranged anagram pair
For i As Integer = 1 To UBound(maxIndex)
  For j As Integer = maxIndex(i) To maxIndex(i) + 1
    iws(j - maxIndex(i) + 1) = iw(j)
  Next j
  If iws(1).index > iws(2).index Then swap iws(1), iws(2) '' ensure pair is in correct order 
  For j As Integer = 1 To 2 
    Print w(iws(j).index); " ";
  Next j
  Print
Next i

Print
Print "Took "; 
Print Using "#.###"; timer - t; 
Print " seconds on i3 @ 2.13 GHz"

Print
Print "Press any key to quit"
Sleep
Output:
25104 words in the dictionary
The deranged anagram pair(s) with the greatest length (namely 10) is:

excitation intoxicate

Took 0.089 seconds on i3 @ 2.13 GHz

FutureBasic

While there is nothing time sensitive about this task, fast code is often efficient code. Several of the entries in this category show their computation times. This FutureBasic entry is designed to outrace them all.

The other entries examined have started by sorting the letters in each word. Here we take a different approach by creating an "avatar" for each word. All anagrams of a word have the same avatar—-without any sorting. Here's how it works:
An 8-byte variable can hold a lot of information. We create a 64-bit avatar that starts at the high end with 8 bits for the length of the word, so that longer words will be sorted first. The remaining 56 bits contain 2-bit fields for each letter of the alphabet. A 2-bit field can record from 0 to 3 occurrences of the letter, but even if there were 4 or more occurrences (think "Mississippi"), bleeding into the next field, the only matching avatar would still be an exact anagram. Here's how the bits would be set for the word "Anagrams":

Anagrams
  length   ZzYyXx WwVvUuTt SsRrQqPp OoNnMmLl KkJjIiHh GgFfEeDd CcBbAa  
00001000 00000000 00000000 01010000 00010100 00000000 01000000 00001100

Bit shifts and 8-byte comparisons are fast operations, which contribute to the speed. As each avatar is generated, it is saved, along with the offset to its word, and an index to it inserted in a sorted list, guaranteeing that longest words occur first, and all matching anagrams are adjacent.

When words have the same avatars, they are anagrams, but for this task we still need to check for letters occurring in the same location in both words. That is a quick check that only has to be done for otherwise qualified candidates.

On a 1.2 GHz Quad-Core Intel Core i7 MacBook Pro, this code runs in ~6 ms, which is several times faster than times claimed by other entries. In that time, it finds not just the longest, but all 486 deranged anagrams in unixdict.txt. (Yes, there is an option to view all of them.)

FWIW, this code can easily be amended to show all 1800+ anagram pairs.

#plist NSAppTransportSecurity @{NSAllowsArbitraryLoads:YES}
defstr long
begin globals
xref   xwords( 210000 ) as char
long   gAvatars( 26000 )
uint32 gwordNum, gfilen, gcount = 0, gOffset( 26000 )
uint16 gndx( 26000 ), deranged( 600, 1 )
long   sh : sh = system( _scrnHeight ) -100
long   sw : sw = (system( _scrnWidth ) -360 ) / 2
CFTimeInterval t
_len = 56
end globals

local fn loadDictionary
  CFURLRef url = fn URLWithString( @"http://wiki.puzzlers.org/pub/wordlists/unixdict.txt" )
  CFStringRef dictStr = fn StringWithContentsOfURL( url, NSUTF8StringEncoding, NULL )
  dictStr = fn StringByAppendingString( @" ", dictStr )
  xwords  = fn StringUTF8String(  dictstr )
  gfilen  = len(dictstr)
end fn

local fn deranagrams
  uint64 ch, p, wordStart = 0
  long   avatar = 0
  uint32 med, bot, top
  byte   chk, L
  
  for p = 1 to gfilen
    ch = xwords(p)                              //build avatar
    if ch > _" " then avatar += (long) 1 << ( ch and 31 ) * 2: continue
    
    avatar += (long)(p - wordStart - 1) << _len //complete avatar by adding word length
    gAvatars(gWordNum) = avatar                 //store the avatar in list
    gOffset( gWordNum) = wordStart              //store offset to the word
    
    //Insert into ordered list of avatars
    bot = 0 : top = gwordNum                    //quick search for place to insert
    while (top - bot) > 1
      med = ( top + bot ) >> 1
      if avatar > gAvatars(gndx(med)) then bot = med else top = med
    wend
    blockmove( @gndx( top ), @gndx( top + 1 ), ( gwordNum - top ) * 2 )
    gndx(top) = gWordNum
    
    gwordNum++ : wordStart = p : avatar = 0     //ready for new word
  next p
  
  //Check for matching avatars
  for p = gWordNum to 1 step -1
    chk = 1  //to make sure each word is compared with all matching avatars
    while gAvatars( gndx( p ) ) == gAvatars( gndx( p - chk ) )
      
      // found anagram; now check for chars in same position
      L = ( gAvatars( gndx( p ) ) >> _len )  //get word length
      while L
        if xwords(gOffset(gndx(p)) +L) == xwords(gOffset(gndx(p-chk)) +L) then break
        L--
      wend
      if L == 0
        
        //no matching chars: found Deranged Anagram!
        deranged( gcount, 0 ) = gndx( p       )
        deranged( gcount, 1 ) = gndx( p - chk )
        gcount++
      end if
      chk++
    wend
  next
end fn

local fn printPair( ndx as uint32, chrsToCntr as byte )
  ptr   p : str255 pair : pair = ""
  short n = ( gAvatars( deranged( ndx, 0 ) ) >> _len )
  if n < chrsToCntr then print string$( chrsToCntr - n, " " );
  p     = xwords + gOffset( deranged( ndx, 0 ) )
  p.0`` = n : print p.0$; "   ";
  p     = xwords + gOffset( deranged( ndx, 1 ) )
  p.0`` = n : print p.0$
end fn

local fn doDialog(evt as long)
  if evt == _btnclick
    long r
    button -1 : window 1,,(sw,50,335,sh-50)
    for r = 1 to gcount-1
      fn printPair( r, 21 )
    next
  end if
end fn

fn loadDictionary : t = fn CACurrentMediaTime
fn deranagrams    : t = fn CACurrentMediaTime - t

window 1, @"Deranged Anagrams in FutureBasic",(sw,sh-130,335,130)
printf @"\n     %u deranged anagrams found among \n   %u words ¬
in %.2f ms.\n", gcount, gWordNum, t * 1000
print  "  Longest:";: fn printPair( 0, 11 )
button 1,,,fn StringWithFormat(@"Show remaining %u deranged anagrams.",gcount-1),(24,20,285,34)
on dialog fn doDialog
handleevents
Output:

File:FB output for Deranged Anagrams.png

GAP

Using function Anagrams.

IsDeranged := function(a, b)
	local i, n;
	for i in [1 .. Size(a)] do
		if a[i] = b[i] then
			return false;
		fi;
	od;
	return true;
end;

# This solution will find all deranged pairs of any length.
Deranged := function(name)
	local sol, ana, u, v;
	sol := [ ];
	ana := Anagrams(name);
	for u in ana do
		for v in Combinations(u, 2) do
			if IsDeranged(v[1], v[2]) then
				Add(sol, v);
			fi;
		od;
	od;
	return sol;
end;

# Now we find all deranged pairs of maximum length
a := Deranged("unixdict.txt");;
n := Maximum(List(a, x -> Size(x[1])));
Filtered(a, x -> Size(x[1]) = n);
# [ [ "excitation", "intoxicate" ] ]

Go

package main
import (
	"fmt"
	"io/ioutil"
	"strings"
	"sort"
)

func deranged(a, b string) bool {
	if len(a) != len(b) {
		return false
	}
	for i := range(a) {
		if a[i] == b[i] { return false }
	}
	return true
}

func main() {
	/* read the whole thing in. how big can it be? */
	buf, _ := ioutil.ReadFile("unixdict.txt")
	words := strings.Split(string(buf), "\n")

	m := make(map[string] []string)
	best_len, w1, w2 := 0, "", ""

	for _, w := range(words) {
		// don't bother: too short to beat current record
		if len(w) <= best_len { continue }

		// save strings in map, with sorted string as key
		letters := strings.Split(w, "")
		sort.Strings(letters)
		k := strings.Join(letters, "")

		if _, ok := m[k]; !ok {
			m[k] = []string { w }
			continue
		}

		for _, c := range(m[k]) {
			if deranged(w, c) {
				best_len, w1, w2 = len(w), c, w
				break
			}
		}

		m[k] = append(m[k], w)
	}

	fmt.Println(w1, w2, ": Length", best_len)
}
Output:
excitation intoxicate : Length 10

Groovy

Solution:

def map = new TreeMap<Integer,Map<String,List<String>>>()

new URL('http://www.puzzlers.org/pub/wordlists/unixdict.txt').eachLine { word ->
    def size = - word.size()
    map[size] = map[size] ?: new TreeMap<String,List<String>>()
    def norm = word.toList().sort().sum()
    map[size][norm] = map[size][norm] ?: []
    map[size][norm] << word
}

def result = map.findResult { negasize, normMap ->
    def size = - negasize
    normMap.findResults { x, anagrams ->
        def n = anagrams.size()
        (0..<(n-1)).findResults { i ->
            ((i+1)..<n).findResult { j ->
                (0..<size).every { k -> anagrams[i][k] != anagrams[j][k] } \
                    ? anagrams[i,j]
                    : null
            }
        }?.flatten() ?: null
    }?.flatten() ?: null
}

if (result) {
    println "Longest deranged anagram pair: ${result}"    
} else {
    println 'Deranged anagrams are a MYTH!'
}
Output:
Longest deranged anagram pair: [excitation, intoxicate]

Haskell

If the longest deranged anagram includes three or more words we'll only print two of them. We also correctly handle duplicate words in the input.

{-# LANGUAGE TupleSections #-}

import Data.List (maximumBy, sort, unfoldr)
import Data.Ord (comparing)
import qualified Data.Map as M
import qualified Data.Set as S

-- Lists of words grouped by their "signatures".  A signature is a sorted
-- list of characters.  Duplicate words stored in sets.
groupBySig :: [String] -> [(String, S.Set String)]
groupBySig = map ((,) . sort <*> S.singleton)

-- Groups as lists of equivalent words.
equivs :: [(String, S.Set String)] -> [[String]]
equivs = map (S.toList . snd) . M.toList . M.fromListWith S.union

-- True if a pair of words differs in all character positions.
isDerangement :: (String, String) -> Bool
isDerangement (a, b) = and $ zipWith (/=) a b

-- All pairs of elements, ignoring order.
pairs :: [t] -> [(t, t)]
pairs = concat . unfoldr step
  where
    step (x:xs) = Just (map (x, ) xs, xs)
    step [] = Nothing

-- All anagram pairs in the input string.
anagrams :: [String] -> [(String, String)]
anagrams = concatMap pairs . equivs . groupBySig

-- The pair of words forming the longest deranged anagram.
maxDerangedAnagram :: [String] -> Maybe (String, String)
maxDerangedAnagram = maxByLen . filter isDerangement . anagrams
  where
    maxByLen [] = Nothing
    maxByLen xs = Just $ maximumBy (comparing (length . fst)) xs

main :: IO ()
main = do
  input <- readFile "unixdict.txt"
  case maxDerangedAnagram $ words input of
    Nothing -> putStrLn "No deranged anagrams were found."
    Just (a, b) -> putStrLn $ "Longest deranged anagrams: " <> a <> " and " <> b
Output:
Longest deranged anagrams: excitation and intoxicate

and a variant:

import Control.Monad ((<=<))
import Data.Function (on)
import Data.List (find, groupBy, sort, sortOn)
import Data.Ord (Down (Down))

-------------------- DERANGED ANAGRAMS -------------------

longestDeranged :: [String] -> String
longestDeranged xs =
  case find deranged (longestAnagramPairs xs) of
    Nothing -> "No deranged anagrams found."
    Just (a, b) -> a <> " -> " <> b

deranged :: (String, String) -> Bool
deranged (a, b) = and (zipWith (/=) a b)

longestAnagramPairs :: [String] -> [(String, String)]
longestAnagramPairs = ((<*>) =<< fmap (,)) <=< 
  (sortOn (Down . length . head) . anagramGroups)

anagramGroups :: [String] -> [[String]]
anagramGroups xs =
  groupBy
    (on (==) fst)
    (sortOn fst (((,) =<< sort) <$> xs))
    >>= (\g -> [snd <$> g | 1 < length g])


--------------------------- TEST -------------------------
main :: IO ()
main =
  readFile "unixdict.txt"
    >>= (putStrLn . longestDeranged . lines)
Output:
excitation -> intoxicate

Icon and Unicon

This solution (which works in both languages) does a strict interpretation of the problem and ignores the fact that there may be multiple derangements that are the same length (including ignoring multiple derangements arising from the same set of words that are all anagrams).

link strings     # for csort() procedure

procedure main()
    anagrams := table()                     # build lists of anagrams
    every *(word := !&input) > 1 do {
        canon := csort(word)
        /anagrams[canon] := []
        put(anagrams[canon], word)
        }

    longest := 1                            # find a longest derangement
    every *(aList := !anagrams) > 1 do     
        if derangement := derange(aList) then 
            if longest <:= *derangement[1] then long := derangement

    every writes((!\long||" ")|"\n")        # show longest
end

procedure derange(aList)    # Return a single derangement from this list
    while aWord := get(aList) do return noMatch(aWord, !aList)
end

procedure noMatch(s1,s2)    # Produce pair only if s1,s2 are deranged.
    every i := 1 to *s1 do if s1[i] == s2[i] then fail
    return [s1,s2]
end
Sample run:
->dra <unixdict.txt
excitation intoxicate 
->

J

This assumes that unixdict.txt has been saved in the current directory.

   #words=: 'b' freads 'unixdict.txt'
25104
   #anagrams=: (#~ 1 < #@>) (</.~ /:~&>) words
1303
   #maybederanged=: (#~ (1 -.@e. #@~."1)@|:@:>&>) anagrams
432
   #longest=: (#~ [: (= >./) #@>@{.@>) maybederanged
1
   longest
┌───────────────────────┐
│┌──────────┬──────────┐│
││excitationintoxicate││
│└──────────┴──────────┘│
└───────────────────────┘

Note that anagram sets with more than two members might, hypothetically, have made things more complicated. By lucky coincidence, this was not an issue. We could have taken advantage of that coincidence to achieve slight further simplifications. Perhaps maybederanged=: (#~ (-: ~."1)@|:@:>&>) anagrams

In other words, if we had had to consider whether ascertain/cartesian/sectarian contained a deranged pair, we would have had to break it out into the three pairs it contains. However, since 'excitation' is a longer word than 'ascertain', we know that this triple cannot contain the longest deranged anagram pair. And since there are no anagrams longer than 'excitation' which involve more than a single pair, we know that we can ignore this issue entirely.

Java

Works with: Java version 8
import java.io.File;
import java.io.IOException;
import java.nio.file.Files;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Comparator;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
 
public class DerangedAnagrams {
 
    public static void main(String[] args) throws IOException {
        List<String> words = Files.readAllLines(new File("unixdict.txt").toPath());
        printLongestDerangedAnagram(words);
    }
 
    private static void printLongestDerangedAnagram(List<String> words) {
        words.sort(Comparator.comparingInt(String::length).reversed().thenComparing(String::toString));

        Map<String, ArrayList<String>> map = new HashMap<>();
        for (String word : words) {
            char[] chars = word.toCharArray();
            Arrays.sort(chars);
            String key = String.valueOf(chars);

            List<String> anagrams = map.computeIfAbsent(key, k -> new ArrayList<>());
            for (String anagram : anagrams) {
                if (isDeranged(word, anagram)) {
                    System.out.printf("%s %s%n", anagram, word);
                    return;
                }
            }
            anagrams.add(word);
        }
        System.out.println("no result");
    }

    private static boolean isDeranged(String word1, String word2) {
        for (int i = 0; i < word1.length(); i++) {
            if (word1.charAt(i) == word2.charAt(i)) {
                return false;
            }
        }
        return true;
    }
}
Output:
excitation intoxicate

JavaScript

Spidermonkey

This example is a little long because it tries to emphasize generality and clarity over brevity.

#!/usr/bin/env js

function main() {
    var wordList = read('unixdict.txt').split(/\s+/);
    var anagrams = findAnagrams(wordList);
    var derangedAnagrams = findDerangedAnagrams(anagrams);
    var longestPair = findLongestDerangedPair(derangedAnagrams);
    print(longestPair.join(' '));
    
}

function findLongestDerangedPair(danas) {
    var longestLen = danas[0][0].length;
    var longestPair = danas[0];
    for (var i in danas) {
        if (danas[i][0].length > longestLen) {
            longestLen = danas[i][0].length;
            longestPair = danas[i];
        }
    }
    return longestPair;
}

function findDerangedAnagrams(anagrams) {
    var deranged = [];
    
    function isDeranged(w1, w2) {
        for (var c = 0; c < w1.length; c++) {
            if (w1[c] == w2[c]) {
                return false;
            }
        }
        return true;
    }

    function findDeranged(anas) {
        for (var a = 0; a < anas.length; a++) {
            for (var b = a + 1; b < anas.length; b++) {
                if (isDeranged(anas[a], anas[b])) {
                    deranged.push([anas[a], anas[b]]);
                }   
            }
        }
    }
    
    for (var a in anagrams) {
        var anas = anagrams[a];
        findDeranged(anas);
    }
    
    return deranged;
}
    
function findAnagrams(wordList) {
    var anagrams = {};

    for (var wordNum in wordList) {
        var word = wordList[wordNum];
        var key = word.split('').sort().join('');
        if (!(key in anagrams)) {
            anagrams[key] = [];
        }
        anagrams[key].push(word);
    }

    for (var a in anagrams) {
        if (anagrams[a].length < 2) {
            delete(anagrams[a]);
        }
    }

    return anagrams;
}

main();
Output:
excitation intoxicate

Gecko

Word file is saved locally because browser won't fetch it cross-site. Tested on Gecko.

<html><head><title>Intoxication</title></head>
<body><pre id='x'></pre>
<script type="application/javascript">

function show(t) {
	var l = document.createTextNode(t + '\n');
	document.getElementById('x').appendChild(l);
}

// get words; be ware of cross-site restrictions on XMLHttpRequest
var words = null;
var req = new XMLHttpRequest();
req.open('GET', 'file:///tmp/unixdict.txt', false);
req.send(null);
words = req.responseText.split('\n');

var idx = {};
for (var i = 0; i < words.length; i++) {
	var t = words[i].split('').sort().join('');
	if (idx[t]) idx[t].push(words[i]);
	else	    idx[t] = [words[i]];
}

var best = '';
var best_pair;
for (var i in idx) {
	if (i.length <= best.length) continue;
	if (idx[i].length == 1) continue;

	var a = idx[i], got = null;
	for (var j = 0, l1 = a[j]; j < a.length && !got; j++) {
		for (var k = j + 1, l2 = a[k]; k < a.length && !got; k++)
			for (var m = 0; m < l1.length || !(got = [l2]); m++)
				if (l1[m] == l2[m]) break;
		if (got) got.push(l1);
	}

	if (got) {
		best_pair = got;
		best = got[0];
	}
}

show(best_pair);
</script></body></html>
Output (in a browser window):
intoxicate,excitation

jq

Works with: jq version 1.5

This solution allows for the possibility of more than one answer.

# Input: an array of strings
# Output: a stream of arrays
def anagrams:
  reduce .[] as $word (
    {table: {}, max: 0};   # state
    ($word | explode | sort | implode) as $hash
    | .table[$hash] += [ $word ]
    | .max   = ([ .max, ( .table[$hash] | length) ] | max ) )
  | .table | .[]  | select(length>1);

# Check whether the input and y are deranged,
# on the assumption that they are anagrams:
def deranged(y):
  explode as $x                             # explode is fast
  | (y | explode) as $y
  | all( range(0;length); $x[.] != $y[.] );
  
# The task: loop through the anagrams,
# retaining only the best set of deranged anagrams so far. 
split("\n") | select(length>0)              # read all the words as an array          
| reduce anagrams as $words ([];            # loop through all the anagrams
    reduce $words[] as $v (.;   
      reduce ($words - [$v])[] as $w (.;    # $v and $w are distinct members of $words
        if $v|deranged($w)
        then if length == 0 then [$v,$w]
             elif ($v|length) == (.[0]|length) then . + [$v,$w]
             elif ($v|length) >  (.[0]|length) then [$v,$w]
	     else .
	     end
        else .
        end) ) )
| unique

Invocation and output

    $ jq -M -s -c -R -f program.jq unixdict.txt 
    ["excitation","intoxicate"]

Julia

using Base.isless
# Let's define the less than operator for any two vectors that have the same type:
# This does lexicographic comparison, we use it on vectors of chars in this task.
function Base.isless(t1, t2)
    for (a, b) in zip(t1, t2) # zip only to the shorter length
        if !isequal(a, b)
            return isless(a, b)
        end
    end
    return length(t1) < length(t2)
end

# The sort function of Julia doesn't work on strings, so we write one:
# This returns a sorted vector of the chars of the given string
sortchars(s::AbstractString) = sort(collect(Char, s))

# Custom comparator function for sorting the loaded wordlist
sortanagr(s1::AbstractString, s2::AbstractString) =
    if length(s1) != length(s2) length(s1) < length(s2) else sortchars(s1) < sortchars(s2) end

# Tests if two strings are deranged anagrams, returns a bool:
# in our case s2 is never longer than s1
function deranged(s1::AbstractString, s2::AbstractString)
    # Tests for derangement first
    for (a, b) in zip(s1, s2)
        if a == b return false end
    end
    # s1 and s2 are deranged, but are they anagrams at all?
    return sortchars(s1) == sortchars(s2)
end

# Task starts here, we load the wordlist line by line, strip eol char, and sort the wordlist
# in a way that ensures that longer words come first and anagrams go next to each other
words = sort(open(readlines, "./data/unixdict.txt"), rev = true, lt = sortanagr)

# Now we just look for deranged anagrams in the neighbouring words of the sorted wordlist
for i in 1:length(words)-1
    if deranged(words[i], words[i+1])
        # The first match is guaranteed to be the longest due to the custom sorting
        println("The longest deranged anagrams are $(words[i]) and $(words[i+1])")
        break
    end
end
Output:
The longest deranged anagrams are excitation and intoxicate

K

   / anagram clusters
   a:{x g@&1<#:'g:={x@<x}'x}@0:"unixdict.txt";
 
   / derangements in these clusters
   b@&c=|/c:{#x[0]}'b:a@&{0=+//{x=y}':x}'a
("excitation"
 "intoxicate")

Kotlin

// version 1.0.6

import java.io.BufferedReader
import java.io.InputStreamReader
import java.net.URL

fun isDeranged(s1: String, s2: String): Boolean {
    return (0 until s1.length).none { s1[it] == s2[it] }
}

fun main(args: Array<String>) {
    val url = URL("http://www.puzzlers.org/pub/wordlists/unixdict.txt")
    val isr = InputStreamReader(url.openStream())
    val reader = BufferedReader(isr)
    val anagrams = mutableMapOf<String, MutableList<String>>()
    var count = 0
    var word = reader.readLine()
    while (word != null) {
        val chars = word.toCharArray()
        chars.sort()
        val key = chars.joinToString("")
        if (!anagrams.containsKey(key)) {
            anagrams.put(key, mutableListOf<String>())
            anagrams[key]!!.add(word)
        }
        else {
            val deranged = anagrams[key]!!.any { isDeranged(it, word) }
            if (deranged) {
                anagrams[key]!!.add(word)
                count = Math.max(count, word.length)
            }
        }
        word = reader.readLine()
    }
    reader.close()
    anagrams.values
        .filter { it.size > 1 && it[0].length == count }
        .forEach { println(it) }
}
Output:
[excitation, intoxicate]

Lasso

local(
	anagrams	= map,
	words		= include_url('http://www.puzzlers.org/pub/wordlists/unixdict.txt') -> split('\n'),
	key,
	max		= 0,
	wordsize,
	findings	= array,
	derangedtest	= { // this code snippet is not executed until the variable is invoked. It will return true if the compared words are a deranged anagram
		local(
			w1		= #1,
			w2		= #2,
			testresult	= true
		)

		loop(#w1 -> size) => {
			#w1 -> get(loop_count) == #w2 -> get(loop_count) ? #testresult = false
		}
		return #testresult
	}
)

// find all anagrams
with word in #words do {
	#key = #word -> split('') -> sort& -> join('')
	not(#anagrams >> #key) ? #anagrams -> insert(#key = array)
	#anagrams -> find(#key) -> insert(#word)
}

// step thru each set of anagrams to find deranged ones
with ana in #anagrams
let ana_size = #ana -> size
where #ana_size > 1
do {
	#wordsize = #ana -> first -> size

	if(#wordsize >= #max) => {

		loop(#ana_size - 1) => {
			if(#derangedtest -> detach & invoke(#ana -> get(loop_count), #ana -> get(loop_count + 1))) => {
				// we only care to save the found deranged anagram if it is longer than the previous longest one
				if(#wordsize > #max) => {
					#findings = array(#ana -> get(loop_count) + ', ' + #ana -> get(loop_count + 1))
				else
					#findings -> insert(#ana -> get(loop_count) + ', ' + #ana -> get(loop_count + 1))
				}
				#max = #wordsize
			}
		}

	}
}

#findings -> join('<br />\n')

Result -> excitation, intoxicate

Liberty BASIC

print "Loading dictionary file."
open "unixdict.txt" for input as #1
a$=input$(#1,lof(#1))
close #1

dim theWord$(30000)
dim ssWord$(30000)

c10$ = chr$(10)
i = 1
print "Creating array of words."
while instr(a$,c10$,i) <> 0
  j     = instr(a$,c10$,i)
  ln    = j - i
  again = 1
  sWord$ = mid$(a$,i,j-i)
  n = n + 1
 theWord$(n) = sWord$

 while again = 1
   again  = 0
   for kk = 1 to len(sWord$) - 1
   if mid$(sWord$,kk,1) > mid$(sWord$,kk +1,1) then
     sWord$ = left$(sWord$,kk-1);mid$(sWord$,kk+1,1);mid$(sWord$,kk,1);mid$(sWord$,kk+2)
     again  = 1
   end if
   next kk
 wend
 ssWord$(n) = sWord$
 i = j + 1
wend
print "Checking for deranged anagrams."
for i = 1 to n
  if len(theWord$(i)) > maxLen then
    for j = 1 to n
      if ssWord$(i) = ssWord$(j) and i <> j then
        cnt = 0
    for k = 1 to len(theWord$(i))
      if mid$(theWord$(i),k,1) = mid$(theWord$(j),k,1) then cnt = cnt + 1
    next k
    if cnt = 0 then
      maxLen = len(theWord$(i))
      maxPtrI = i
      maxPtrJ = j
    end if
      end if
    next j
  end if
next i

print theWord$(maxPtrI);" => ";theWord$(maxPtrJ)
end
Output:

excitation => intoxicate

Lua

string.tacnoc = function(str) -- 'inverse' of table.concat
  local arr={}
  for ch in str:gmatch(".") do arr[#arr+1]=ch end
  return arr
end

local function deranged(s1, s2)
  if s1==s2 then return false end
  local t1, t2 = s1:tacnoc(), s2:tacnoc()
  for i,v in ipairs(t1) do if t2[i]==v then return false end end
  return true
end

local dict = {}
local f = io.open("unixdict.txt", "r")
for word in f:lines() do
  local ltrs = word:tacnoc()
  table.sort(ltrs)
  local hash = table.concat(ltrs)
  dict[hash] = dict[hash] or {}
  table.insert(dict[hash], word)
end

local answer = { word="", anag="", len=0 }
for _,list in pairs(dict) do
  if #list>1 and #list[1]>answer.len then
    for _,word in ipairs(list) do
      for _,anag in ipairs(list) do
        if deranged(word, anag) then
          answer.word, answer.anag, answer.len = word, anag, #word
        end
      end
    end
  end
end
print(answer.word, answer.anag, answer.len)
Output:
excitation      intoxicate      10

Maple

with(StringTools):
dict:=Split([HTTP:-Get("www.puzzlers.org/pub/wordlists/unixdict.txt")][2]):
L:=[seq(select(t->HammingDistance(t,w)=length(w),[Anagrams(w,dict)])[],w=dict)]:
len:=length(ListTools:-FindMaximalElement(L,(a,b)->length(a)<length(b))):
select(w->length(w)=len,L)[];
Output:
                        "intoxicate", "excitation"

Mathematica / Wolfram Language

words=First/@Import["http://www.puzzlers.org/pub/wordlists/unixdict.txt","Table"];
anagramDegrangement=Function[{w1,w2},
	Module[{c1=ToCharacterCode@w1,c2=ToCharacterCode@w2},
	Sort@c1==Sort@c2&&Select[c1-c2,#==0&,1]==={}]];
gs=Select[GatherBy[words,{StringLength@#,Union@ToCharacterCode@#}&],Length@#>=2&];
First@Flatten[Function[ws,Select[Join@@Outer[List,ws,ws,1],anagramDegrangement@@#&]]/@SortBy[gs,-StringLength@First@#&],1]
Output:
{"excitation", "intoxicate"}

A similar approach using Mathematica 10:

list = Import["http://www.puzzlers.org/pub/wordlists/unixdict.txt","Lines"];
MaximalBy[
 Select[GatherBy[list, Sort@*Characters], 
  Length@# > 1 && And @@ MapThread[UnsameQ, Characters /@ #] &], 
 StringLength@*First]
Output:
{{"excitation", "intoxicate"}}

Nim

import algorithm
import tables
import times

var anagrams: Table[seq[char], seq[string]]  # Mapping sorted_list_of chars -> list of anagrams.

#---------------------------------------------------------------------------------------------------

func deranged(s1, s2: string): bool =
  ## Return true if "s1" and "s2" are deranged anagrams.

  for i, c in s1:
    if s2[i] == c:
      return false
  result = true

#---------------------------------------------------------------------------------------------------

let t0 = getTime()

# Build the anagrams table.
for word in lines("unixdict.txt"):
  anagrams.mgetOrPut(sorted(word), @[]).add(word)

# Find the longest deranged anagrams.
var bestLen = 0
var best1, best2: string
for (key, list) in anagrams.pairs:
  if key.len > bestLen:
    var s1 = list[0]
    for i in 1..list.high:
      let s2 = list[i]
      if deranged(s1, s2):
        # Found a better pair.
        best1 = s1
        best2 = s2
        bestLen = s1.len
        break

echo "Longest deranged anagram pair: ", best1, " ", best2
echo "Processing time: ", (getTime() - t0).inMilliseconds, " ms."
Output:
Longest deranged anagram pair: excitation intoxicate
Processing time: 57 ms.

OCaml

let sort_chars s =
  let r = String.copy s in
  for i = 0 to (String.length r) - 2 do
    for j = i + 1 to (String.length r) - 1 do
      if r.[i] > r.[j] then begin
        let tmp = r.[i] in
        r.[i] <- r.[j];
        r.[j] <- tmp;
      end
    done
  done;
  (r)

let deranged (s1, s2) =
  let len1 = String.length s1
  and len2 = String.length s2 in
  if len1 <> len2 then false else
  try
    for i = 0 to pred len1 do
      if s1.[i] = s2.[i] then raise Exit
    done;
    true
  with Exit -> false

let pairs_of_list lst =
  let rec aux acc = function
    | [] -> acc
    | x::xs ->
        aux (List.fold_left (fun acc y -> (x,y)::acc) acc xs) xs
  in
  aux [] lst

let () =
  let h = Hashtbl.create 3571 in
  let ic = open_in "unixdict.txt" in
  try while true do
    let word = input_line ic in
    let key = sort_chars word in
    let l =
      try Hashtbl.find h key
      with Not_found -> [] 
    in
    Hashtbl.replace h key (word::l);
  done with End_of_file ->
    close_in ic;
    let lst =
      Hashtbl.fold (fun _ lw acc ->
        if List.length lw < 2 then acc else lw::acc) h []
    in
    let lst =
      List.fold_left (fun acc anagrams ->
        let pairs = pairs_of_list anagrams in
        (List.filter deranged pairs) @ acc
      ) [] lst
    in
    let res, _ =
      List.fold_left (fun (res, n) (w1, w2) ->
        let len = String.length w1 in
        match Pervasives.compare len n with
        | 0 -> ((w1, w2)::res, n)
        | 1 -> ([w1, w2], len)
        | _ -> (res, n)
      ) ([], 0) lst
    in
    List.iter (fun (w1, w2) -> Printf.printf "%s, %s\n" w1 w2) res
Output:
$ ocaml deranged_anagram.ml
intoxicate, excitation

ooRexx

-- This assumes you've already downloaded the following file and placed it
-- in the current directory: http://www.puzzlers.org/pub/wordlists/unixdict.txt

-- There are several different ways of reading the file.  I chose the
-- supplier method just because I haven't used it yet in any other examples.
source = .stream~new('unixdict.txt')~supplier
-- this holds our mappings of the anagrams.  This is good use for the
-- relation class
anagrams = .relation~new
count = 0    -- this is used to keep track of the maximums

loop while source~available
    word = source~item
    -- this produces a string consisting of the characters in sorted order
    -- Note: the ~~ used to invoke sort makes that message return value be
    -- the target array.  The sort method does not normally have a return value.
    key = word~makearray('')~~sort~tostring("l", "")
    -- add this to our mapping.  This creates multiple entries for each
    -- word that uses the same key
    anagrams[key] = word
    source~next
end

-- now get the set of unique keys
keys = .set~new~~putall(anagrams~allIndexes)
-- the longest count tracker
longest = 0
-- our list of the longest pairs
pairs = .array~new

loop key over keys
    -- don't even bother doing the deranged checks for any key
    -- shorter than our current longest
    if key~length < longest then iterate

    words = anagrams~allAt(key)
    -- singletons aren't anagrams at all
    newCount = words~items
    loop i = 1 to newCount - 1
        word1 = words[i]
        loop j = 1 to newCount
            word2 = words[j]
            -- bitxor will have '00'x in every position where these
            -- strings match.  If found, go around and check the
            -- next one
            if word1~bitxor(word2)~pos('00'x) > 0 then iterate
            -- we have a match
            else do
                if word1~length > longest then do
                    -- throw away anything we've gathered so far
                    pairs~empty
                    longest = word1~length
                end
                pairs~append(.array~of(word1, word2))
            end
        end
    end
end

say "The longest deranged anagrams we found are:"
loop pair over pairs
     say pair[1] pair[2]
end
Output:
The longest deranged anagrams we found are:
intoxicate excitation

PARI/GP

dict=readstr("unixdict.txt");
len=apply(s->#s, dict);
getLen(L)=my(v=List()); for(i=1,#dict, if(len[i]==L, listput(v, dict[i]))); Vec(v);
letters(s)=vecsort(Vec(s));
getAnagrams(v)=my(u=List(),L=apply(letters,v),t,w); for(i=1,#v-1, w=List(); t=L[i]; for(j=i+1,#v, if(L[j]==t, listput(w, v[j]))); if(#w, listput(u, concat([v[i]], Vec(w))))); Vec(u);
deranged(s1,s2)=s1=Vec(s1);s2=Vec(s2); for(i=1,#s1, if(s1[i]==s2[i], return(0))); 1
getDeranged(v)=my(u=List(),w); for(i=1,#v-1, for(j=i+1,#v, if(deranged(v[i], v[j]), listput(u, [v[i], v[j]])))); Vec(u);
f(n)=my(t=getAnagrams(getLen(n))); if(#t, concat(apply(getDeranged, t)), []);
forstep(n=vecmax(len),1,-1, t=f(n); if(#t, return(t)))
Output:
%1 = [["excitation", "intoxicate"]]

Pascal

Works with: Free Pascal
Translation of: Delphi

Using extra Stringlist for sorted by character words and insertion sort.
Runtime 153 ms -> 35 ms (Free Pascal Compiler version 3.3.1-r20:47268 [2020/11/02] for x86_64)

program Anagrams_Deranged;
{$IFDEF FPC}
  {$MODE Delphi}
  {$Optimization ON,ALL}
uses
  SysUtils,
  Classes;
{$ELSE}
  {$APPTYPE CONSOLE}
uses
  System.SysUtils,
  System.Classes,
  {$R *.res}
{$ENDIF}

function Sort(const s: string):string;
//insertion sort
var
  pRes : pchar;
  i, j, aLength: NativeInt;
  tmpc: Char;
begin
  aLength := s.Length;

  if aLength = 0 then
    exit('');

  Result := s;
  //without it, s will be sorted
  UniqueString(Result);
  //insertion sort
  pRes := pChar(Result);
  dec(aLength,1);
  for i := 0 to aLength do
  Begin
    tmpc := pRes[i];
    j := i-1;
    while (j>=0) AND (tmpc < pRes[j]) do
    Begin
      pRes[j+1] := pRes[j];
      dec(j);
    end;
    inc(j);
    pRes[j]:= tmpc;
  end;
end;

function CompareLength(List: TStringList; Index1, Index2: longInt): longInt;
begin
  result := List[Index1].Length - List[Index2].Length;
  IF result = 0 then
    result := CompareStr(List[Index1],List[Index2]);
end;

function IsDerangement(const word1, word2: string): Boolean;
var
  i: NativeInt;
begin
  for i := word1.Length downto 1 do
    if word1[i] = word2[i] then
      exit(False);
  Result := True;
end;

var
  Dict,SortDict: TStringList;
  words: string;
  StopWatch: Int64;
  Count, Index: NativeInt;

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

  StopWatch := GettickCount64;
  SortDict:= TStringList.Create();
  SortDict.capacity := Dict.Count;
  For Index := 0 to Dict.Count - 1  do
  Begin
    SortDict.Add(Sort(Dict[Index]));
    //remember the origin in Dict
    SortDict.Objects[Index]:= TObject(Index);
  end;

  SortDict.CustomSort(CompareLength);

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

  while Index - Count >= 0 do
  begin
    if SortDict[Index]= SortDict[Index - Count] then
    begin
      if IsDerangement(Dict[NativeInt(SortDict.Objects[Index])],
         Dict[NativeInt(SortDict.Objects[Index - Count])]) then
      begin
        words := Dict[NativeInt(SortDict.Objects[Index])] + ' - ' +
                 Dict[NativeInt(SortDict.Objects[Index - Count])];
        Break;
      end;
      Inc(Count);
    end
    else
    begin
      Dec(Index, Count);
      Count := 1;
    end;
  end;
  StopWatch := GettickCount64-StopWatch;
  Writeln(Format('Time pass: %d ms [AMD 2200G-Linux Fossa]',[StopWatch]));
  writeln(#10'Longest derangement words are:'#10#10, words);

  SortDict.free;
  Dict.Free;
end.
Output:
Time pass: 33 ms [AMD 2200G-Linux Fossa]

Longest derangement words are:

excitation - intoxicate

Perl

String operations

use strict;
use warnings;

sub deranged {                          # only anagrams ever get here
        my @a = split('', shift);       # split word into letters
        my @b = split('', shift);
        for (0 .. $#a) {
                $a[$_] eq $b[$_] and return;
        }
        return 1
}

sub find_deranged {     
        for my $i ( 0 .. $#_ ) {
                for my $j ( $i+1 .. $#_ ) {
                        next unless deranged $_[$i], $_[$j];

                        print "length ", length($_[$i]), ": $_[$i] => $_[$j]\n";
                        return 1;
                }
        }
}

my %letter_list;
open my $in, 'unixdict.txt';

local $/ = undef;

for (split(' ', <$in>)) {
        # store anagrams in hash table by letters they contain
        push @{ $letter_list{ join('', sort split('', $_)) } }, $_
}

for (   sort { length($b) <=> length($a) }      # sort by length, descending
        grep { @{ $letter_list{$_} } > 1 }      # take only ones with anagrams
        keys %letter_list               )
{
        # if we find a pair, they are the longest due to the sort before
        last if find_deranged(@{ $letter_list{$_} });
}
Output:
length 10: excitation => intoxicate

Bitwise operations

use strict;
use warnings;
use feature 'bitwise';

local (@ARGV, $/) = 'unixdict.txt';
my %anagrams;

for my $word ( sort { length $b <=> length $a } split ' ', <> )
  {
  my $key = join '', sort +split //, $word;
  ($_ ^. $word) =~ /\0/ or exit !print "$_ $word\n" for @{ $anagrams{$key} };
  push @{ $anagrams{$key} }, $word;
  }
Output:
excitation intoxicate

Phix

function deranged(string word1, word2)
    return sum(sq_eq(word1,word2))=0
end function
 
integer fn = open("demo/unixdict.txt","r")
sequence words = {}, anagrams = {}, last="", letters
object word
integer maxlen = 1
 
while 1 do
    word = trim(gets(fn))
    if atom(word) then exit end if
    if length(word) then
        letters = sort(word)
        words = append(words, {letters, word})
    end if
end while
close(fn)

words = sort(words)
for i=1 to length(words) do
    {letters,word} = words[i]
    if letters=last then
        anagrams[$] = append(anagrams[$],word)
        anagrams[$][1] = length(word)
    else
        last = letters
        anagrams = append(anagrams,{0,word})
    end if
end for

anagrams = sort(anagrams)
puts(1,"\nLongest deranged anagrams:\n")
for i=length(anagrams) to 1 by -1 do
    last = anagrams[i]
    if last[1]<maxlen then exit end if
    for j=2 to length(last) do
        for k=j+1 to length(last) do
            if deranged(last[j],last[k]) then
                puts(1,last[j]&", "&last[k]&"\n")
                maxlen = last[1]
            end if
        end for
    end for
end for
Output:
Longest deranged anagrams:
excitation, intoxicate

Phixmonti

/# Rosetta Code problem: http://rosettacode.org/wiki/Anagrams/Deranged_anagrams
by Galileo, 06/2022 #/

include ..\Utilitys.pmt

"unixdict.txt" "r" fopen var f

( )

true while
    f fgets
    dup -1 == if
        drop
        f fclose
        false
    else
        -1 del
        dup sort swap 2 tolist 
        0 put 
        true
    endif   
endwhile

sort

0 var largest
( ) var candidate

( len 2 swap ) for var i
    ( i 1 ) sget >ps
    ( i 1 - 1 ) sget ps> == if
        ( i 2 ) sget >ps
        ( i 1 - 2 ) sget ps> len >ps
        true var test
        tps for var j
            j get rot j get rot == if false var test exitfor endif          
        endfor
        test tps largest > and if
            ps> var largest
            2 tolist var candidate
        else
            ps> drop drop drop
        endif
    endif
endfor

candidate print
Output:
["excitation", "intoxicate"]
=== Press any key to exit ===

PHP

<?php
$words = file(
    'http://www.puzzlers.org/pub/wordlists/unixdict.txt',
    FILE_IGNORE_NEW_LINES
);
$length = 0;

foreach ($words as $word) {
    $chars = str_split($word);
    sort($chars);
    $chars = implode("", $chars);
    $length = strlen($chars);
    $anagrams[$length][$chars][] = $word;
}

krsort($anagrams);

foreach ($anagrams as $anagram) {
    $final_words = array();
    foreach ($anagram as $words) {
        if (count($words) >= 2) {
            $counts = array();
            foreach ($words as $word) {
                $counts[$word] = array($word);
                foreach ($words as $second_word) {
                    for ($i = 0, $length = strlen($word); $i < $length; $i++) {
                        if ($word[$i] === $second_word[$i]) continue 2;
                    }
                    $counts[$word][] = $second_word;
                }
            }
            $max = 0;
            $max_key = '';
            foreach ($counts as $name => $count) {
                if (count($count) > $max) {
                    $max = count($count);
                    $max_key = $name;
                }
            }
            if ($max > 1) {
                $final_words[] = $counts[$max_key];
            }
        }
    }
    if ($final_words) break;
}

foreach ($final_words as $final_word) {
    echo implode(" ", $final_word), "\n";
}
?>
Output:
excitation intoxicate

Picat

go =>
  M = [W:W in read_file_lines("unixdict.txt")].group(sort),
  Deranged = [Value : _Key=Value in M, Value.length > 1, allderanged(Value)],
  MaxLen = max([V[1].length : V in Deranged]),
  println([V : V in Deranged, V[1].length==MaxLen]),
  nl.

% A and B are deranged: i.e. there is no
% position with the same character.
deranged(A,B) => 
   foreach(I in 1..A.length)
       A[I] != B[I]
   end.

% All words in list Value are deranged anagrams of each other.
allderanged(Value) => 
    IsDeranged = 1,
    foreach(V1 in Value, V2 in Value, V1 @< V2, IsDeranged = 1)
       if not deranged(V1,V2) then
          IsDeranged := 0
       end
    end,
    IsDeranged == 1.

% Groups the element in List according to the function F
group(List, F) = P, list(List) =>
   P = new_map(),
   foreach(E in List) 
      V = apply(F,E),
      P.put(V, P.get(V,[]) ++ [E])
   end.
Output:
[[excitation,intoxicate]]

PicoLisp

(let Words NIL
   (in "unixdict.txt"
      (while (line)
         (let (Word @  Key (pack (sort (copy @))))
            (if (idx 'Words Key T)
               (push (car @) Word)
               (set Key (list Word)) ) ) ) )
   (maxi '((X) (length (car X)))
      (extract
         '((Key)
            (pick
               '((Lst)
                  (and
                     (find
                        '((L) (not (find = L Lst)))
                        (val Key) )
                     (cons (pack @) (pack Lst)) ) )
               (val Key) ) )
         (idx 'Words) ) ) )
Output:
-> ("excitation" . "intoxicate")

PowerShell

function Test-Deranged ([string[]]$Strings)
{
    $array1 = $Strings[0].ToCharArray()

    for ($i = 1; $i -lt $Strings.Count; $i++)
    { 
        $array2 = $Strings[$i].ToCharArray()

        for ($i = 0; $i -lt $array1.Count; $i++)
        { 
            if ($array1[$i] -match $array2[$i])
            {
                return $false
            }
        }
    }

    return $true
}


$words = [System.Collections.ArrayList]@()

Get-Content -Path ".\unixdict.txt" |
    ForEach-Object { [void]$words.Add([PSCustomObject]@{Word=$_; SortedWord=(($_.ToCharArray() | Sort-Object) -join "")}) }

[object[]]$anagrams = $words | Group-Object -Property SortedWord | Where-Object -Property Count -GT 1 | Sort-Object {$_.Name.Length}
[string[]]$deranged = ($anagrams | ForEach-Object { if ((Test-Deranged $_.Group.Word)) {$_} } | Select-Object -Last 1).Group.Word

[PSCustomObject]@{
    Length = $deranged[0].Length
    Words  = $deranged
}
Output:
Length Words                   
------ -----                   
    10 {excitation, intoxicate}

Prolog

Works with: SWI Prolog
longest_deranged_anagram :-
	http_open('http://www.puzzlers.org/pub/wordlists/unixdict.txt',In,[]),
	read_file(In, [], Out),
	close(In),
	msort(Out, MOut),
	group_pairs_by_key(MOut, GPL),
	map_list_to_pairs(compute_len, GPL, NGPL),
	predsort(my_compare, NGPL, GPLSort),
	search_derangement(GPLSort).


% order tuples to have longest words first
my_compare(R, N1-(K1-E1), N2-(K2-E2)) :-
	(   N1 < N2 -> R = > ; N1 > N2 -> R = <;
	length(E1, L1),
	length(E2, L2),
	(   L1 < L2 -> R = <; L1 > L2 -> R = >; compare(R, K1, K2))).


compute_len(_-[H|_], Len) :-
	length(H, Len).


% check derangement of anagrams
derangement([], []).
derangement([H1|T1], [H2 | T2]) :-
	H1 \= H2,
	derangement(T1, T2).


search_derangement([_-(_-L) | T]) :-
	length(L, 1), !,
	search_derangement(T).


search_derangement([_-(_-L) | T]) :-
	(   search(L) -> true; search_derangement(T)).

search([]) :- fail.
search([H | T]) :-
	(   search_one(H, T) -> true; search(T)).


search_one(Word, L) :-
	include(derangement(Word), L, [H|_]),
	atom_codes(W, Word),
	atom_codes(W1, H),
	format('Longest deranged anagrams : ~w ~w ~n', [W, W1]).


read_file(In, L, L1) :-
	read_line_to_codes(In, W),
	(   W == end_of_file ->
	       L1 = L
	       ;
	       msort(W, W1),
	       atom_codes(A, W1),
	       read_file(In, [A-W | L], L1)).
Output:
 ?- longest_deranged_anagram.
Longest deranged anagrams : excitation intoxicate 
true.

PureBasic

Structure anagram
  word.s
  letters.s
EndStructure
 
Structure  wordList
  List words.anagram()
EndStructure
 
#True = 1
#False = 0
 
Procedure.s sortLetters(*word.Character, wordLength)
  ;returns a string with the letters of a word sorted
  Protected Dim letters.c(wordLength)
  Protected *letAdr = @letters()
  CopyMemoryString(*word, @*letAdr)
  SortArray(letters(), #PB_Sort_Ascending, 0, wordLength - 1)
  ProcedureReturn PeekS(@letters(), wordLength)
EndProcedure
 
;Compare a list of anagrams for derangement.
Procedure isDeranged(List anagram.s()) 
  ;If a pair of deranged anagrams is found return #True
  ;and and modify the list to include the pair of deranged anagrams.
  Protected i, length, word.s, *ptrAnagram, isDeranged
  Protected NewList deranged.s()
  FirstElement(anagram())
  length = Len(anagram())
  Repeat 
    word = anagram()
    *ptrAnagram = @anagram()
    
    While NextElement(anagram())
      isDeranged = #True
      For i = 1 To length
        If Mid(word, i, 1) = Mid(anagram(), i, 1)
          isDeranged = #False
          Break ;exit for/next
        EndIf 
      Next 
      
      If isDeranged
        AddElement(deranged())
        deranged() = anagram()
        AddElement(deranged())
        deranged() = word
        CopyList(deranged(), anagram())
        ProcedureReturn #True ;deranged anagram found
      EndIf 
    Wend
    ChangeCurrentElement(anagram(), *ptrAnagram)
  Until Not NextElement(anagram())
  
  ProcedureReturn #False ;deranged anagram not found
EndProcedure
 
If OpenConsole()
  ;word file is assumed to be in the same directory
  If Not ReadFile(0,"unixdict.txt"): End: EndIf 
  
  Define maxWordSize = 0, word.s, length
  Dim wordlists.wordList(maxWordSize)
  
  ;Read word file and create separate lists of anagrams and their original
  ;words by length.
  While Not Eof(0)
    word = ReadString(0)
    length = Len(word)
    If length > maxWordSize
      maxWordSize = length
      Redim wordlists.wordList(maxWordSize)
    EndIf 
    AddElement(wordlists(length)\words())
    wordlists(length)\words()\word = word
    wordlists(length)\words()\letters = sortLetters(@word, length)
  Wend
  CloseFile(0)
  
  Define offset = OffsetOf(anagram\letters), option = #PB_Sort_Ascending
  Define sortType = #PB_Sort_String
  Define letters.s, foundDeranged
  NewList anagram.s()
  ;start search from largest to smallest
  For length = maxWordSize To 2 Step -1
    
    If FirstElement(wordlists(length)\words()) ;only examine lists with words
      ;sort words to place anagrams next to each other
      SortStructuredList(wordlists(length)\words(), option, offset, sortType)
      
      With wordlists(length)\words()
        letters = \letters
        AddElement(anagram()): anagram() = \word
        
        ;Compose sets of anagrams and check for derangement with remaining
        ;words in current list.
        While NextElement(wordlists(length)\words())
          ;Check for end of a set of anagrams?
          If letters <> \letters
            
            ;if more than one word in a set of anagrams check for derangement
            If ListSize(anagram()) > 1
              If isDeranged(anagram())
                foundDeranged = #True ;found deranged anagrams, stop processing
                Break 2 ;exit while/wend and for/next
              EndIf 
            EndIf 
            
            letters = \letters ;setup for next set of anagrams
            ClearList(anagram())
          EndIf 
          
          AddElement(anagram()): anagram() = \word
        Wend
      EndWith
      
    EndIf 
    
    ClearList(anagram())
  Next   
  
  ;report results
  If foundDeranged
    Print("Largest 'Deranged' anagrams found are of length ")
    PrintN(Str(length) + ":" + #CRLF$)
    ForEach anagram()
      PrintN("  " + anagram())
    Next 
  Else
    PrintN("No 'Deranged' anagrams were found." + #CRLF$)
  EndIf 
  
  Print(#CRLF$ + #CRLF$ + "Press ENTER to exit"): Input()
  CloseConsole()
EndIf
Output:
Largest 'Deranged' anagrams found are of length 10:

  intoxicate
  excitation

Python

import urllib.request
from collections import defaultdict
from itertools import combinations

def getwords(url='http://www.puzzlers.org/pub/wordlists/unixdict.txt'):
    return list(set(urllib.request.urlopen(url).read().decode().split()))

def find_anagrams(words):
    anagram = defaultdict(list) # map sorted chars to anagrams
    for word in words:
        anagram[tuple(sorted(word))].append( word )
    return dict((key, words) for key, words in anagram.items()
                if len(words) > 1)

def is_deranged(words):
    'returns pairs of words that have no character in the same position'
    return [ (word1, word2)
             for word1,word2 in combinations(words, 2)
             if all(ch1 != ch2 for ch1, ch2 in zip(word1, word2)) ]

def largest_deranged_ana(anagrams):
    ordered_anagrams = sorted(anagrams.items(),
                              key=lambda x:(-len(x[0]), x[0]))
    for _, words in ordered_anagrams:
        deranged_pairs = is_deranged(words)
        if deranged_pairs:
            return deranged_pairs
    return []

if __name__ == '__main__':
    words = getwords('http://www.puzzlers.org/pub/wordlists/unixdict.txt')
    print("Word count:", len(words))

    anagrams = find_anagrams(words)
    print("Anagram count:", len(anagrams),"\n")

    print("Longest anagrams with no characters in the same position:")
    print('  ' + '\n  '.join(', '.join(pairs)
                             for pairs in largest_deranged_ana(anagrams)))
Output:
Word count: 25104
Anagram count: 1303 

Longest anagrams with no characters in the same position:
  excitation, intoxicate

Python: Anagrams with most derangements

The talk page had a later discussion on groups of anagrams with the most derangements. This computes that.

Append the following to the previous code:

def most_deranged_ana(anagrams):
    ordered_anagrams = sorted(anagrams.items(),
                              key=lambda x:(-len(x[0]), x[0]))
    many_anagrams = [anas for _, anas in ordered_anagrams if len(anas) > 2]
    d_of_anas = [is_deranged(ana_group) for ana_group in many_anagrams]
    d_of_anas = [d_group for d_group in d_of_anas if d_group]
    d_of_anas.sort(key=lambda d_group:(-len(d_group), -len(d_group[0])))
    mx = len(d_of_anas[0])
    most = [sorted(d_group) for d_group in d_of_anas if len(d_group) == mx]
    return most

if __name__ == '__main__':
    most = most_deranged_ana(anagrams)
    print(f"\nThere are {len(most)} groups of anagrams all containing" 
          f" a max {len(most[0])} deranged word-pairs:")
    for pairs in most:
        print()
        print('  ' + '\n  '.join(', '.join(p) for p in pairs))
Output:

Extra output generated:

There are 4 groups of anagrams all containing a max 5 deranged word-pairs:

  angel, glean
  angel, lange
  angle, glean
  galen, angle
  lange, glean

  bale, abel
  bela, abel
  elba, abel
  elba, able
  elba, bale

  evil, live
  evil, vile
  levi, evil
  levi, vile
  veil, live

  emit, item
  emit, time
  mite, emit
  mite, item
  time, item


Python: Faster Version

from collections import defaultdict
from itertools import combinations
from pathlib import Path
from typing import (Callable,
                    Dict,
                    Iterable,
                    Iterator,
                    List,
                    Optional,
                    Tuple,
                    TypeVar)

WORDS_FILE = 'unixdict.txt'

T1 = TypeVar('T1')
T2 = TypeVar('T2')


def main():
    words = read_words(Path(WORDS_FILE))
    anagram = longest_deranged_anagram(words)
    if anagram:
        print('The longest deranged anagram is: {}, {}'.format(*anagram))
    else:
        print('No deranged anagrams were found')


def read_words(path: Path) -> Iterator[str]:
    """Yields words from file at specified path"""
    with path.open() as file:
        for word in file:
            yield word.strip()


def longest_deranged_anagram(words: Iterable[str]
                             ) -> Optional[Tuple[str, str]]:
    """
    Returns the longest pair of words
    that have no character in the same position
    """
    words_by_lengths = mapping_by_function(len, words)
    decreasing_lengths = sorted(words_by_lengths, reverse=True)
    for length in decreasing_lengths:
        words = words_by_lengths[length]
        anagrams_by_letters = mapping_by_function(sort_str, words)
        for anagrams in anagrams_by_letters.values():
            deranged_pair = next(deranged_word_pairs(anagrams), None)
            if deranged_pair is not None:
                return deranged_pair
    return None


def mapping_by_function(function: Callable[..., T2],
                        iterable: Iterable[T1]) -> Dict[T2, List[T1]]:
    """
    Constructs a dictionary with keys
    obtained from applying an input function
    to items of an iterable,
    and the values filled from the same iterable
    """
    mapping = defaultdict(list)
    for item in iterable:
        mapping[function(item)].append(item)
    return mapping


def sort_str(string: str) -> str:
    """Sorts input string alphabetically"""
    return ''.join(sorted(string))


def deranged_word_pairs(words: Iterable[str]) -> Iterator[Tuple[str, str]]:
    """Yields deranged words from an input list of words"""
    pairs = combinations(words, 2)  # type: Iterator[Tuple[str, str]]
    yield from filter(is_deranged, pairs)


def is_deranged(word_pair: Tuple[str, str]) -> bool:
    """
    Checks if all corresponding letters are different,
    assuming that words have the same length
    """
    return all(a != b for a, b in zip(*word_pair))


if __name__ == '__main__':
    main()
Output:
The longest anagram is: excitation, intoxicate

Quackery

  [ over size over size != iff
      [ 2drop false ] done
    over sort over sort != iff
      [ 2drop false ] done
    true unrot witheach 
      [ dip behead = if
          [ dip not conclude ] ] 
    drop ]                       is deranged ( $ $ --> b )

  $ 'rosetta/unixdict.txt' sharefile drop nest$
  [] temp put 
  dup size times
    [ behead over witheach
      [ 2dup deranged iff
          [ over nested swap
            nested join nested 
            temp take join temp put ]
        else drop ]
      drop ] 
  drop 
  temp take
  sortwith [ 0 peek size swap 0 peek size > ] 
  0 peek witheach [ echo$ sp ]
Output:
excitation intoxicate 

R

puzzlers.dict <- readLines("http://www.puzzlers.org/pub/wordlists/unixdict.txt")

longest.deranged.anagram <- function(dict=puzzlers.dict) {
  anagram.groups <- function(word.group) {
    sorted <- sapply(lapply(strsplit(word.group,""),sort),paste, collapse="")
    grouped <- tapply(word.group, sorted, force, simplify=FALSE)
    grouped <- grouped[sapply(grouped, length) > 1]
    grouped[order(-nchar(names(grouped)))]
  }
  
  derangements <- function(anagram.group) {
    pairs <- expand.grid(a = anagram.group, b = anagram.group,
                         stringsAsFactors=FALSE)
    pairs <- subset(pairs, a < b)
    deranged <- with(pairs, mapply(function(a,b) all(a!=b),
                                   strsplit(a,""), strsplit(b,"")))
    pairs[which(deranged),]
  }

  for (anagram.group in anagram.groups(dict)) {
    if (nrow(d <- derangements(anagram.group)) > 0) {
      return(d[1,])
    }
  }
}
Output:
> longest.deranged.anagram()
           a          b
3 excitation intoxicate

Racket

#lang racket
(define word-list-file "data/unixdict.txt")

(define (read-words-into-anagram-keyed-hash)
  (define (anagram-key word) (sort (string->list word) char<?))
  (for/fold ((hsh (hash)))
    ((word (in-lines)))
    (hash-update hsh (anagram-key word) (curry cons word) null)))

(define anagrams-list
  (sort
   (for/list
       ((v (in-hash-values
            (with-input-from-file
                word-list-file
              read-words-into-anagram-keyed-hash)))
        #:when (> (length v) 1)) v)
   > #:key (compose string-length first)))


(define (deranged-anagram-pairs l (acc null))
  (define (deranged-anagram-pair? hd tl)
    (define (first-underanged-char? hd tl)
      (for/first
          (((c h) (in-parallel hd tl))
           #:when (char=? c h)) c))  
    (not (first-underanged-char? hd tl)))
  
  (if (null? l) acc
      (let ((hd (car l)) (tl (cdr l)))
        (deranged-anagram-pairs
         tl
         (append acc (map (lambda (x) (list hd x))
                          (filter (curry deranged-anagram-pair? hd) tl)))))))

;; for*/first give the first set of deranged anagrams (as per the RC problem)
;; for*/list gives a full list of the sets of deranged anagrams (which might be interesting)
(for*/first
    ((anagrams (in-list anagrams-list))
     (daps (in-value (deranged-anagram-pairs anagrams)))
     #:unless (null? daps))
  daps)
Output:
'(("intoxicate" "excitation"))

Raku

(formerly Perl 6)

Works with: Rakudo version 2016.08
my @anagrams = 'unixdict.txt'.IO.words
    .map(*.comb.cache)             # explode words into lists of characters
    .classify(*.sort.join).values  # group words with the same characters
    .grep(* > 1)                   # only take groups with more than one word
    .sort(-*[0])                   # sort by length of the first word
;

for @anagrams -> @group {
    for @group.combinations(2) -> [@a, @b] {
        if none @a Zeq @b {
            say "{@a.join}   {@b.join}";
            exit;
        }
    }
}
Output:
excitation   intoxicate

REXX

/*REXX program finds the  largest  deranged word  (within an identified dictionary).    */
iFID= 'unixdict.txt';     words=0                /*input file ID; number of words so far*/
wL.=0                                            /*number of words of length  L.  so far*/
      do while lines(iFID)\==0                   /*read each word in the file  (word=X).*/
      x= strip( linein( iFID) )                  /*pick off a word from the input line. */
      L= length(x);       if L<3 then iterate    /*onesies & twosies can't possible win.*/
      words= words + 1                           /*bump the count of  (usable)  words.  */
      #.words= L                                 /*the length of the word found.        */
      @.words= x                                 /*save the word in an array.           */
      wL.L= wL.L+1;       _= wL.L                /*bump counter of words of length  L.  */
      @@.L._= x                                  /*array   of words of length  L.       */
                                  do i=1  while x\=='';  parse var x !.i +1 x;  end  /*i*/
      call eSort L;       z=;     do j=1  for L;         z= z || !.j;           end  /*j*/
      @@s.L._= z                                 /*store the sorted word (letters).     */
      @s.words= @@s.L._                          /*store the sorted length  L  version. */
      end   /*while*/
a.=                                              /*all the  anagrams  for word  X.      */
say copies('─', 30)   words   'usable words in the dictionary file: '     iFID
m= 0;              n.= 0                         /*# anagrams for word  X;   m=max L.   */
       do j=1  for words                         /*process usable words that were found.*/
       Lx= #.j;   if Lx<m  then iterate          /*get length of word; skip if too short*/
       x= @.j;    xs= @s.j                       /*get some vital statistics for  X     */
           do k=1  for wL.Lx                     /*process all the words of length  L.  */
           if xs\== @@s.Lx.k  then iterate       /*is this not a true anagram of  X ?   */
           if x  ==  @@.Lx.k  then iterate       /*skip of processing anagram on itself.*/
                do c=1  for Lx                   /*ensure no character position shared. */
                if substr(@.j, c, 1) == substr(@@.Lx.k, c, 1)  then iterate k
                end   /*c*/                      /* [+]  now compare the rest of chars. */
           n.j= n.j + 1;     a.j= a.j   @@.Lx.k  /*bump counter;  then add ──► anagrams.*/
           m= max(m, Lx)                         /*M  is the maximum length of a word.  */
           end        /*k*/
       end            /*j*/

   do k=1  for words                             /*now, search all words for the maximum*/
   if #.k==m   then if n.k\==0   then if word(a.k, 1) > @.k  then say  @.k  a.k
   end   /*k*/                                   /* [↑]  REXX has no short-circuits.    */
exit                                             /*stick a fork in it,  we're all done. */
/*──────────────────────────────────────────────────────────────────────────────────────*/
eSort: procedure expose !.;  parse arg ho 1 h    /*obtain number of elements; also copy.*/
         do while h>1;      h=h % 2;                    do i=1  for ho-h;   j= i;   k= h+i
         do while !.k<!.j;  t=!.j;  !.j=!.k;  !.k=t;   if h>=j  then leave;  j=j-h;  k=k-h
         end   /*while !.k···*/;         end  /*i*/;         end  /*while h>1*/;    return
output   when using the default dictionary:
────────────────────────────── 24945 usable words in the dictionary file:  unixdict.txt
excitation  intoxicate

Ring

# Project : Anagrams/Deranged anagrams

load "stdlib.ring"
fn1 = "unixdict.txt"

fp = fopen(fn1,"r")
str = fread(fp, getFileSize(fp))
fclose(fp) 
strlist = str2list(str)
anagram = newlist(len(strlist), 5)
anag = list(len(strlist))
result = list(len(strlist))
for x = 1 to len(result)
     result[x] = 0
next
for x = 1 to len(anag)
     anag[x] = 0
next
for x = 1 to len(anagram)
      for y = 1 to 5
           anagram[x][y] = 0
      next
next

strbig = 1
for n = 1 to len(strlist)
     for m = 1 to len(strlist)
           sum = 0
           if len(strlist[n]) = len(strlist[m]) and n != m
              for p = 1 to len(strlist[m])
                    temp1 = count(strlist[n], strlist[m][p])
                    temp2 = count(strlist[m], strlist[m][p])
                    if temp1 = temp2
                       sum = sum + 1
                    ok
              next
              if sum = len(strlist[n])
                 anag[n] = anag[n] + 1
                 if anag[n] < 6 and result[n] = 0 and result[m] = 0
                    anagram[n][anag[n]] = strlist[m]
                    if len(strlist[m]) > len(strlist[strbig])
                       strbig = n
                    ok
                      result[m] = 1
                  ok
               ok
            ok
      next
      if anag[n] > 0
         result[n] = 1
      ok
next

flag = 0
for m = 1 to 5
     if anagram[strbig][m] != 0
        if m = 1
           see strlist[strbig] +  " "
           flag = 1
        ok
        see anagram[strbig][m] + " "
     ok
next

func getFileSize fp
       c_filestart = 0
       c_fileend = 2
       fseek(fp,0,c_fileend)
       nfilesize = ftell(fp)
       fseek(fp,0,c_filestart)
       return nfilesize

func count(astring,bstring)
       cnt = 0
       while substr(astring,bstring) > 0
                cnt = cnt + 1
                astring = substr(astring,substr(astring,bstring)+len(string(sum)))
       end
       return cnt
Output:
excitation intoxicate 

Ruby

def deranged?(a, b)
  a.chars.zip(b.chars).all? {|char_a, char_b| char_a != char_b}
end

def find_derangements(list)
  list.combination(2) {|a,b| return a,b  if deranged?(a,b)}
  nil
end

require 'open-uri'
anagram = open('http://www.puzzlers.org/pub/wordlists/unixdict.txt') do |f|
  f.read.split.group_by {|s| s.each_char.sort}
end

anagram = anagram.select{|k,list| list.size>1}.sort_by{|k,list| -k.size}

anagram.each do |k,list|
  if derangements = find_derangements(list)
    puts "Longest derangement anagram: #{derangements}"
    break
  end
end
Output:
Longest derangement anagram: ["excitation", "intoxicate"]

Run BASIC

a$ = httpGet$("http://www.puzzlers.org/pub/wordlists/unixdict.txt")
dim theWord$(30000)
dim ssWord$(30000)

c10$ = chr$(10)
i = 1
while instr(a$,c10$,i) <> 0
  j     = instr(a$,c10$,i)
  ln    = j - i
  again = 1
  sWord$ = mid$(a$,i,j-i)
  n = n + 1
 theWord$(n) = sWord$

 while again = 1
   again  = 0
   for kk = 1 to len(sWord$) - 1
   if mid$(sWord$,kk,1) > mid$(sWord$,kk +1,1) then
     sWord$ = left$(sWord$,kk-1);mid$(sWord$,kk+1,1);mid$(sWord$,kk,1);mid$(sWord$,kk+2)
     again  = 1
   end if
   next kk
 wend
 ssWord$(n) = sWord$
 i = j + 1
wend

for i = 1 to n
  if len(theWord$(i)) > maxLen then
    for j = 1 to n
      if ssWord$(i) = ssWord$(j) and i <> j then
        cnt = 0
	for k = 1 to len(theWord$(i))
	  if mid$(theWord$(i),k,1) = mid$(theWord$(j),k,1) then cnt = cnt + 1
	next k
	if cnt = 0 then 
	  maxLen = len(theWord$(i))
	  maxPtrI = i
	  maxPtrJ = j
	end if
      end if
    next j
  end if
next i

print maxLen;" ";theWord$(maxPtrI);" => ";theWord$(maxPtrJ)
end
Output:
10 excitation => intoxicate

Rust

//! Deranged anagrams
use std::cmp::Ordering;
use std::collections::HashMap;
use std::fs::File;
use std::io;
use std::io::BufReader;
use std::io::BufRead;
use std::usize::MAX;

/// Get words from unix dictionary file
pub fn get_words() -> Result<Vec<String>, io::Error> {
    let mut words = vec!();
    // open file
    let f = File::open("data/unixdict.txt")?;
    // read line by line
    let reader = BufReader::new(&f);
    for line in reader.lines() {
        words.push(line?)
    }
    Ok(words)
}

/// Get the longest deranged anagram in the given list of word if any
pub fn longest_deranged(v: &mut Vec<String>) -> Option<(String,String)>{
    // sort by length descending then by alphabetical order
    v.sort_by(|s1, s2| {
        let mut c = s2.len().cmp(&s1.len());
        if c == Ordering::Equal {
            c = s1.cmp(s2);
        }
        c
    });
    // keep all strings keyed by sorted characters (since anagrams have the same list of sorted characters)
    let mut signatures : HashMap<Vec<char>, Vec<&String>> = HashMap::new();
    // save on memory by only keeping in the map strings of the current processed length
    let mut previous_length = MAX;
    for s in v {
        // length change, clear the map
        if s.len()<previous_length {
            signatures.clear();
            previous_length = s.len();
        }
        // generate key as sorted characters
        let mut sorted_chars = s.chars().collect::<Vec<char>>();
        sorted_chars.sort();
        let anagrams = signatures.entry(sorted_chars).or_insert(vec!());
        // find if any anagram (string with the same sorted character vector) is deranged
        if let Some(a) = anagrams.iter().filter(|anagram| is_deranged(anagram, s)).next(){
            return Some(((*a).clone(), s.clone()));
        }
        anagrams.push(s);
    }
    None
}

/// check if two strings do NOT have the same character in the same position
pub fn is_deranged(s1: &String, s2: &String) -> bool {
    // we zip the character iterators and check we find no position with the same two characters
    s1.chars().zip(s2.chars()).filter(|(a,b)| a == b).next().is_none()
}

/// an example main method printing the results
fn main() {
    let r = get_words();
    match r {
        Ok(mut v) => {
            let od = longest_deranged(&mut v);
            match od {
                None => println!("No deranged anagrams found!"),
                Some((s1,s2)) => println!("{} {}",s1,s2),
            }
        },
        Err(e) => panic!("Could not read words: {}",e)
    } 
}
Output:
excitation intoxicate

Scala

object DerangedAnagrams {

  /** Returns a map of anagrams keyed by the sorted characters */ 
  def groupAnagrams(words: Iterable[String]): Map[String, Set[String]] =
    words.foldLeft (Map[String, Set[String]]()) { (map, word) =>
      val sorted = word.sorted
      val entry = map.getOrElse(sorted, Set.empty)
      map + (sorted -> (entry + word))
    }
    
  /* Returns true if the pair of strings has no positions with the same
   * characters */
  def isDeranged(ss: (String, String)): Boolean = 
    ss._1 zip ss._2 forall { case (c1, c2) => c1 != c2 }
    
  /* Returns pairwise combination of all Strings in the argument Iterable */
  def pairWords(as: Iterable[String]): Iterable[(String, String)] =
    if (as.size < 2) Seq() else (as.tail map (as.head -> _)) ++ pairWords(as.tail)
    
  /* Returns the contents of the argument URL as an Iterable[String], each
   * String is one line in the file */
  def readLines(url: String): Iterable[String] = 
    io.Source.fromURL(url).getLines().toIterable
    
  val wordsURL = "http://www.puzzlers.org/pub/wordlists/unixdict.txt"
    
  def main(args: Array[String]): Unit = {
    val anagramMap = groupAnagrams(readLines(wordsURL))
    val derangedPairs = anagramMap.values flatMap (pairWords) filter (isDeranged)
    val (w1, w2) = derangedPairs maxBy (pair => pair._1.length)
    println("Longest deranged pair: "+w1+" and "+w2)
  }   

}
Output:
Longest deranged pair: excitation and intoxicate

Scheme

(import (scheme base)
        (scheme char)
        (scheme cxr)
        (scheme file)
        (scheme write)
        (srfi 1)    ; lists
        (srfi 132)) ; sorting library

;; read in word list, and sort into decreasing length
(define (read-ordered-words)
  (with-input-from-file
    "unixdict.txt"
    (lambda ()
      (do ((line (read-line) (read-line))
           (words '() (cons line words)))
        ((eof-object? line) 
         (list-sort (lambda (a b) (> (string-length a) (string-length b)))
                    words))))))

(define (find-deranged-words word-list)
  (define (search words)
    (let loop ((word-chars (let ((chars (map string->list words)))
                             (zip chars 
                                  (map (lambda (word) (list-sort char<? word)) 
                                       chars)))))
      (if (< (length word-chars) 2)
        #f ; failed to find any
        (let ((deranged-word ; seek a deranged version of the first word in word-chars
                (find (lambda (chars)
                        (and (equal? (cadar word-chars) (cadr chars)) ; check it's an anagram?
                             (not (any char=? (caar word-chars) (car chars))))) ; and deranged?
                      word-chars)))
          (if deranged-word ; if we got one, return it with the first word
            (map list->string (list (caar word-chars) (car deranged-word)))
            (loop (cdr word-chars)))))))
  ;
  (let loop ((rem word-list))
    (if (null? rem)
      '()
      (let* ((len (string-length (car rem)))
             (deranged-words (search ; look through group of equal sized words
                               (take-while (lambda (word) (= len (string-length word)))
                                           (cdr rem)))))
        (if deranged-words
          deranged-words
          (loop (drop-while (lambda (word) (= len (string-length word)))
                            (cdr rem))))))))

(display (find-deranged-words (read-ordered-words))) (newline)
Output:
(excitation intoxicate)

Sidef

func find_deranged(Array a) {
    for i in (^a) {
        for j in (i+1 .. a.end) {
            overlaps(a[i], a[j]) || (
                printf("length %d: %s => %s\n", a[i].len, a[i], a[j])
                return true
            )
        }
    }
    return false
}

func main(File file) {

    file.open_r(\var fh, \var err) ->
        || die "Can't open file `#{file}' for reading: #{err}\n"

    var letter_list = Hash()

    # Store anagrams in hash table by letters they contain
    fh.words.each { |word|
        letter_list{word.sort} := [] << word
    }

    letter_list.keys                        \
         .grep {|k| letter_list{k}.len > 1} \     # take only ones with anagrams
         .sort {|a,b| b.len <=> a.len}      \     # sort by length, descending
         .each {|key|

        # If we find a pair, they are the longested due to the sort before
        find_deranged(letter_list{key}) && break
    }
}

main(%f'/tmp/unixdict.txt')
Output:
length 10: excitation => intoxicate

Simula

! cim --memory-pool-size=512 deranged-anagrams.sim;
BEGIN

   CLASS TEXTVECTOR;
   BEGIN

      CLASS TEXTARRAY(N); INTEGER N;
      BEGIN TEXT ARRAY DATA(1:N);
      END TEXTARRAY;

      PROCEDURE EXPAND(N); INTEGER N;
      BEGIN
         INTEGER I;
         REF(TEXTARRAY) TEMP;
         TEMP :- NEW TEXTARRAY(N);
         FOR I := 1 STEP 1 UNTIL SIZE DO
            TEMP.DATA(I) :- ITEMS.DATA(I);
         ITEMS :- TEMP;
      END EXPAND;

      PROCEDURE APPEND(T); TEXT T;
      BEGIN
         IF SIZE + 1 > CAPACITY THEN
         BEGIN
            CAPACITY := 2 * CAPACITY;
            EXPAND(CAPACITY);
         END;
         SIZE := SIZE + 1;
         ITEMS.DATA(SIZE) :- T;
      END APPEND;

      TEXT PROCEDURE ELEMENT(I); INTEGER I;
      BEGIN
         IF I < 1 OR I > SIZE THEN ERROR("ELEMENT: INDEX OUT OF BOUNDS");
         ELEMENT :- ITEMS.DATA(I);
      END ELEMENT;

      INTEGER PROCEDURE FIND_INDEX(STR,INDEX); TEXT STR; INTEGER INDEX;
      BEGIN
         INTEGER I, FOUND;
         FOUND := -1;
         FOR I := INDEX STEP 1 UNTIL SIZE DO
            IF STR = ELEMENT(I) THEN
            BEGIN
               FOUND := I;
               GOTO L;
            END;
      L: FIND_INDEX := FOUND;
      END FIND_INDEX;

      INTEGER CAPACITY;
      INTEGER SIZE;
      REF(TEXTARRAY) ITEMS;

      CAPACITY := 20;
      SIZE := 0;
      EXPAND(CAPACITY);
   END TEXTVECTOR;

   BOOLEAN PROCEDURE DERANGE(S1,S2); TEXT S1,S2;
   BEGIN
      INTEGER I;
      BOOLEAN RESULT;
      RESULT := TRUE;
      I := 1;
      WHILE RESULT AND I <= S1.LENGTH DO
      BEGIN
         CHARACTER C1, C2;
         S1.SETPOS(I); C1 := S1.GETCHAR;
         S2.SETPOS(I); C2 := S2.GETCHAR;
         IF C1 = C2 THEN
            RESULT := FALSE
         ELSE
            I := I+1;
      END;
      DERANGE := RESULT;
   END DERANGE;

   PROCEDURE STRSORT(STR); NAME STR; TEXT STR;
   BEGIN
      INTEGER N, I;
      FOR N := STR.LENGTH STEP -1 UNTIL 2 DO
         FOR I := 1 STEP 1 UNTIL N-1 DO
         BEGIN
            CHARACTER CI1,CI2;
            STR.SETPOS(I); CI1 := STR.GETCHAR; CI2 := STR.GETCHAR;
            IF CI1 > CI2 THEN
            BEGIN
               STR.SETPOS(I); STR.PUTCHAR(CI2); STR.PUTCHAR(CI1);
            END;
         END;
   END STRSORT;

   REF(INFILE) FILE;
   INTEGER LEN, FOUNDLEN;
   REF(TEXTVECTOR) VECT, SVECT;
   INTEGER INDEX, P1, P2;
   TEXT STR;

   VECT :- NEW TEXTVECTOR;
   SVECT :- NEW TEXTVECTOR;
   FOUNDLEN := 1;
   FILE :- NEW INFILE("unixdict.txt");
   FILE.OPEN(BLANKS(132));
   WHILE NOT FILE.LASTITEM DO
   BEGIN
      STR :- FILE.INTEXT(132).STRIP;
      LEN := STR.LENGTH;
      IF LEN > FOUNDLEN THEN
      BEGIN
         VECT.APPEND(COPY(STR));
         STRSORT(STR);
         INDEX := 0;
         COMMENT Loop through anagrams by index in vector of sorted strings;
         INDEX := SVECT.FIND_INDEX(STR, INDEX + 1);
         WHILE INDEX > 0 DO
         BEGIN
            IF DERANGE(VECT.ELEMENT(VECT.SIZE), VECT.ELEMENT(INDEX)) THEN
            BEGIN
               P1 := VECT.SIZE;
               P2 := INDEX;
               FOUNDLEN := LEN;
            END IF;
            INDEX := SVECT.FIND_INDEX(STR, INDEX + 1);
         END WHILE;
         SVECT.APPEND(STR);
      END IF;
   END WHILE;
   FILE.CLOSE;
   OUTTEXT(VECT.ELEMENT(P1) & " " & VECT.ELEMENT(P2));
   OUTIMAGE;
END
Output:
intoxicate excitation

3 garbage collection(s) in 2.9 seconds.

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

# Group by characters in word
foreach word $wordlist {
    dict lappend w [lsort [split $word ""]] [split $word ""]
}

# Deranged test
proc deranged? {l1 l2} {
    foreach c1 $l1 c2 $l2 {
	if {$c1 eq $c2} {return 0}
    }
    return 1
}

# Get a deranged pair from an anagram set, if one exists
proc getDeranged {words} {
    foreach l1 [lrange $words 0 end-1] {
	foreach l2 [lrange $words 1 end] {
	    if {[deranged? $l1 $l2]} {
		return [list $l1 $l2 1]
	    }
	}
    }
    return {{} {} 0}
}

# Find the max-length deranged anagram
set count 0
set candidates {}
set max 0
dict for {k words} $w {
    incr count [expr {[llength $words] > 1}]
    if {[llength $k] > $max && [lassign [getDeranged $words] l1 l2]} {
	set max [llength $l1]
	lappend candidates [join $l1 ""],[join $l2 ""]
    }
}

# Print out what we found
puts "[llength $wordlist] words"
puts "[dict size $w] potential anagram-groups"
puts "$count real anagram-groups"
foreach pair $candidates {
    puts "considered candidate pairing: $pair"
}
puts "MAXIMAL DERANGED ANAGRAM: LENGTH $max\n\t[lindex $candidates end]"
Output:
25105 words
23567 potential anagram-groups
1303 real anagram-groups
considered candidate pairing: abc,cab
considered candidate pairing: abed,bade
considered candidate pairing: abort,bator
considered candidate pairing: afresh,shafer
considered candidate pairing: alberto,latrobe
considered candidate pairing: american,cinerama
considered candidate pairing: ancestral,lancaster
considered candidate pairing: excitation,intoxicate
MAXIMAL DERANGED ANAGRAM: LENGTH 10
	excitation,intoxicate

TUSCRIPT

$$ MODE TUSCRIPT,{}
requestdata = REQUEST ("http://www.puzzlers.org/pub/wordlists/unixdict.txt")

DICT anagramm CREATE 99999

COMPILE
 LOOP word=requestdata
  -> ? : any character
  charsInWord=STRINGS (word," ? ")
  charString =ALPHA_SORT (charsInWord)

  DICT anagramm LOOKUP charString,num,freq,wordalt,wlalt
  IF (num==0) THEN
   WL=SIZE (charString)
   DICT anagramm APPEND/QUIET/COUNT charString,num,freq,word,wl;" "
  ELSE
   DICT anagramm APPEND/QUIET/COUNT charString,num,freq,word,"";" "
  ENDIF
 ENDLOOP

DICT anagramm UNLOAD charString,all,freq,anagrams,wl

index        =DIGIT_INDEX (wl)
reverseIndex =REVERSE (index)
wl           =INDEX_SORT (wl,reverseIndex)
freq         =INDEX_SORT (freq,reverseIndex)
anagrams     =INDEX_SORT (anagrams,reverseIndex)
charString   =INDEX_SORT (charString,reverseIndex)

LOOP fr=freq,a=anagrams,w=wl
 IF (fr==1) cycle
 asplit=SPLIT (a,": :")
 a1=SELECT (asplit,1,arest)
 a1split=STRINGS (a1," ? ")
 LOOP r=arest
  rsplit=STRINGS (r," ? ")
   LOOP v1=a1split,v2=rsplit
    IF (v1==v2) EXIT,EXIT
   ENDLOOP
    PRINT "Largest deranged anagram (length: ",w,"):"
    PRINT a
   STOP
 ENDLOOP
ENDLOOP
ENDCOMPILE
Output:
Largest deranged anagram (length: 10):
excitation intoxicate

UNIX Shell

Works with: ksh93
function get_words {
    typeset host=www.puzzlers.org
    typeset page=/pub/wordlists/unixdict.txt
    exec 7<>/dev/tcp/$host/80
    print -e -u7 "GET $page HTTP/1.1\r\nhost: $host\r\nConnection: close\r\n\r\n" 
    # remove the http header and save the word list
    sed 's/\r$//; 1,/^$/d' <&7 >"$1"
    exec 7<&-
}

function is_deranged {
    typeset -i i
    for ((i=0; i<${#1}; i++)); do
        [[ ${1:i:1} == ${2:i:1} ]] && return 1
    done
    return 0
}

function word2key {
    typeset -a chars=( $(
        for ((i=0; i<${#word}; i++)); do 
            echo "${word:i:1}"
        done | sort
    ) )
    typeset IFS=""
    echo "${chars[*]}"
}

[[ -f word.list ]] || get_words word.list

typeset -A words
typeset -i max=0

while IFS= read -r word; do
    key=$(word2key $word)
    if [[ -z "${words["$key"]}" ]]; then
        words["$key"]=$word
    else
        if (( ${#word} > max )); then
            if is_deranged "${words["$key"]}" "$word"; then
                max_deranged=("${words["$key"]}" "$word")
                max=${#word}
            fi
        fi
    fi
done <word.list
echo $max - ${max_deranged[@]}
Output:
10 - excitation intoxicate

Ursala

This solution assumes the file unixdict.txt is passed to the compiler as a command line parameter.

#import std

anagrams = |=tK33lrDSL2SL ~=&& ==+ ~~ -<&

deranged = filter not zip; any ==

#cast %sW

main = leql$^&l deranged anagrams unixdict_dot_txt

The anagrams function is a little slow as defined above, but can be sped up by at least two orders of magnitude by grouping the words into classes of equal length, and sorting each word once in advance instead of each time a comparison is made as shown below.

anagrams = @NSiXSlK2rSS *= ^(-<&,~&)*; |=rSStFtK33lrDSL2SL ~=@br&& ==@bl

We can speed it up by about another factor of 5 by starting from the group of longest words and stopping as soon as a deranged anagram is found instead of generating all anagrams.

#import std

longest_deranged_anagram = 

@NSiXSlK2rSS leql-<x&h; @NiX ~&lZrB->l ^\~&rt @rh -+
   ~&a^& ~&plrEkZ?ah/~&ah ~&fatPR,
   ^(-<&,~&)*; |=rSStFtK33lrDSL2SL ~=@br&& ==@bl+-

#cast %sW

main = longest_deranged_anagram unixdict_dot_txt
Output:
('excitation','intoxicate')

VBA

Sub Main_DerangedAnagrams()
Dim ListeWords() As String, Book As String, i As Long, j As Long, tempLen As Integer, MaxLen As Integer, tempStr As String, IsDeranged As Boolean, count As Integer, bAnag As Boolean
Dim t As Single
t = Timer
    Book = Read_File("C:\Users\" & Environ("Username") & "\Desktop\unixdict.txt")
    ListeWords = Split(Book, vbNewLine)
    For i = LBound(ListeWords) To UBound(ListeWords) - 1
        For j = i + 1 To UBound(ListeWords)
            If Len(ListeWords(i)) = Len(ListeWords(j)) Then
                tempLen = 0
                IsDeranged = False
                bAnag = IsAnagram(ListeWords(i), ListeWords(j), IsDeranged, tempLen)
                If IsDeranged Then
                    count = count + 1
                    If tempLen > MaxLen Then
                        MaxLen = tempLen
                        tempStr = ListeWords(i) & ", " & ListeWords(j)
                    End If
                End If
            End If
        Next j
    Next i
    Debug.Print "There is : " & count & " deranged anagram, in unixdict.txt."
    Debug.Print "The longest is : " & tempStr
    Debug.Print "Lenght  : " & MaxLen
    Debug.Print "Time to compute : " & Timer - t & " sec."
End Sub
Private Function Read_File(Fic As String) As String
Dim Nb As Integer
    Nb = FreeFile
    Open Fic For Input As #Nb
        Read_File = Input(LOF(Nb), #Nb)
    Close #Nb
End Function
Function IsAnagram(str1 As String, str2 As String, DerangedAnagram As Boolean, Lenght As Integer) As Boolean
Dim i As Integer
    str1 = Trim(UCase(str1))
    str2 = Trim(UCase(str2))
    For i = 1 To Len(str1)
        If Len(Replace(str1, Mid$(str1, i, 1), vbNullString)) <> Len(Replace(str2, Mid$(str1, i, 1), vbNullString)) Then
            Exit Function
        End If
        If Mid$(str1, i, 1) = Mid$(str2, i, 1) Then
            Exit Function
        End If
    Next i
    IsAnagram = True
    DerangedAnagram = True
    Lenght = Len(str1)
End Function
Output:
There is : 507 deranged anagram, in unixdict.txt.
The longest is : EXCITATION, INTOXICATE
Lenght  : 10
Time to compute : 97,00781 sec.

V (Vlang)

Translation of: Go
import os

fn deranged(a string, b string) bool {
	if a.len != b.len {
		return false
	}
	for i in 0..a.len {
		if a[i] == b[i] { return false }
	}
	return true
}
fn main(){
    words := os.read_lines('unixdict.txt')?
    
    mut m := map[string][]string{}
    mut best_len, mut w1, mut w2 := 0, '',''

    for w in words {
		// don't bother: too short to beat current record
		if w.len <= best_len { continue }
 
		// save strings in map, with sorted string as key
		mut letters := w.split('')
		letters.sort()
		k := letters.join("")
 
		if k !in m {
			m[k] = [w]
			continue
		}
 
		for c in m[k] {
			if deranged(w, c) {
				best_len, w1, w2 = w.len, c, w
				break
			}
		}
 
		m[k] << w
	}
 
	println('$w1 $w2: Length $best_len')
}
Output:
excitation intoxicate: Length 10

Wren

Library: Wren-sort
import "io" for File
import "./sort" for Sort

// assumes w1 and w2 are anagrams of each other
var isDeranged = Fn.new { |w1, w2|
    for (i in 0...w1.count) {
        if (w1[i] == w2[i]) return false
    }
    return true
}

var words = File.read("unixdict.txt").split("\n").map { |w| w.trim() }
var wordMap = {}
for (word in words) {
    var letters = word.toList
    Sort.insertion(letters)
    var sortedWord = letters.join()
    if (wordMap.containsKey(sortedWord)) {
        wordMap[sortedWord].add(word)
    } else {
        wordMap[sortedWord] = [word]
    }
}

var deranged = []
for (key in wordMap.keys) {
    var ana = wordMap[key]
    var count = ana.count
    if (count > 1) {
        for (i in 0...count-1) {
            for (j in i + 1...count) {
                if (isDeranged.call(ana[i], ana[j])) deranged.add([ana[i], ana[j]])
            }
        }
    }
}

var most = deranged.reduce(0) { |max, words| (words[0].count > max) ? words[0].count : max }
for (words in deranged) {
    if (words[0].count == most) System.print([words[0], words[1]])
}
Output:
[excitation, intoxicate]

zkl

words:=Dictionary(25000);  //-->Dictionary(sorted word:all anagrams, ...)
File("unixdict.txt").read().pump(Void,'wrap(w){
   w=w.strip(); key:=w.sort(); words[key]=words.find(key,T).append(w);
});

nws:=words.values.pump(List,fcn(ws){ //-->( (len,words), ...)
   if(ws.len()>1){ // two or more anagrams
      r:=List(); n:=ws[0].len(); // length of these anagrams
      foreach idx,w in (ws.enumerate()){
	 foreach w2 in (ws[idx+1,*]){
	    if(Utils.zipWith('!=,w,w2).filter().len()==n)
	       r.write(T(w,w2));
	 }
      }
      if(r) return(r.insert(0,n));
   }
   Void.Skip
});

nws.filter(fcn(nws,max){ nws[0]==max },
   nws.reduce(fcn(p,nws){ p.max(nws[0]) },0) )
.println();
Output:
L(L(10,L("excitation","intoxicate")))

Replace the center section with the following for smaller code (3 lines shorter!) that is twice as slow:

nws:=words.values.pump(List,fcn(ws){ //-->( (len,words), ...)
   if(ws.len()>1){ // two or more anagrams
      n:=ws[0].len(); // length of these anagrams
      r:=Utils.Helpers.permute(ws).filter('wrap(ws2){
            n == Utils.zipWith('!=,ws2.xplode()).filter().len();
	 });
      if(r) return(n,r[0]); // L(L("glove","vogel"))-->L(5,L("glove","vogel"))
   }
   Void.Skip
});