Repeating Characters

[Andreik]

I don't know to use very well StringRegExp and I want to know if is a better way to do that:

$TEXT = "aabdabecfdbaebcfdfaaabceeabfeabbcccd"
MsgBox(0,"",Replace($TEXT))

Func Replace($TEXT)
    Local $RESULT = $TEXT
    For $LIMIT = 4 To 2 Step -1
        $RESULT = StringRegExpReplace($RESULT,"a{" & $LIMIT & "}",$LIMIT & "a")
        $RESULT = StringRegExpReplace($RESULT,"b{" & $LIMIT & "}",$LIMIT & "b")
        $RESULT = StringRegExpReplace($RESULT,"c{" & $LIMIT & "}",$LIMIT & "c")
        $RESULT = StringRegExpReplace($RESULT,"d{" & $LIMIT & "}",$LIMIT & "d")
        $RESULT = StringRegExpReplace($RESULT,"e{" & $LIMIT & "}",$LIMIT & "e")
        $RESULT = StringRegExpReplace($RESULT,"f{" & $LIMIT & "}",$LIMIT & "f")
    Next
    Return $RESULT
EndFunc

This function replace any 2 or more recurring sequences of letters in the number of appearances and the character.

Example:

abcbdddeaff

will become

abcb3dea2f

Important: Characters set have just this letters: abcdef and the main deficiency of this function is that I need to set a trivial limit but maybe somewhere in my string sequence I will cross this limit and I will have 5 or more characters recurring. This function can be improved in any way? Thanks.

[Xenobiologist]

Hi,

not much but :-)

$TEXT = "aabdabecfdbaebcfdfaaabceeabfeabbcccd"
MsgBox(0, "", Replace($TEXT))

Func Replace($TEXT)
    Local $RESULT = $TEXT
    For $i = 97 To 102
        For $LIMIT = 4 To 2 Step -1
            $RESULT = StringRegExpReplace($RESULT, Chr($i) & "{" & $LIMIT & "}", $LIMIT & Chr($i))
        Next
    Next
    Return $RESULT
EndFunc   ;==>Replace

[Ascend4nt]

This was an interesting challenge.. I've never really used backreferences inside a pattern itself, but after a few tries, the below seems to work well enough. You'll need to change the '[[:alpha:]]' bit to '[a-f]' (I made it generic for replacing any repeating character):

; <CharacterRepeatFind.au3>

Local $iLoc=1,$iCount,$aPCRE,$cChar,$sPCRE='([[:alpha:]])(\1+)',$sString='jjkklliowerjklasdddf'

While 1
    $aPCRE=StringRegExp($sString,$sPCRE,1,$iLoc)
    $iLoc=@extended
    If @error Then ExitLoop
    $iCount=StringLen($aPCRE[1])+1
    $cChar=$aPCRE[0]
    ConsoleWrite("Found "&$iCount&" occurrences of '"&$cChar&"'"&@CRLF)
    ; Replace occurrences with repeat # & Char:
    $sString=StringReplace($sString,$cChar&$aPCRE[1],$iCount&$cChar,1,2)
    ; Adjust $iLoc now to new position (we just cut off $iCount-1)
    $iLoc-=$iCount-1
WEnd
ConsoleWrite("Resulting string: "&$sString&@CRLF)

[jchd]

Since you only deal with a small charset but with possibly unbounded repetition factors, something like this is possible albeit probably not he most efficient in all possible case:

Local $txt = 'abccccffedbbbbbbbbbbbffdaccbefdeeaaacccccccccccccccccccccccccccccccccccccccccccccccccccccce'
Local $txtin = StringRegExpReplace($txt, '(?i)(a{2,}|b{2,}|c{2,}|d{2,}|e{2,}|f{2,})', '1$1')
Local $txtout, $txtfinal
;~ ConsoleWrite('Input      : ' & $txt & @LF)
;~ ConsoleWrite('First phase : ' & $txtin & @LF)
While 1
    $txtout = StringRegExpReplace($txtin, '(?i)(\d)([a-f])\g2', '$1+1$2')
    If $txtout = $txtin Then ExitLoop
    $txtin = $txtout
;~  ConsoleWrite('Next phase  : ' & $txtin & @LF)
WEnd
$txtexec = '"' & StringRegExpReplace($txtout, '((?:1\+)+1)', '" & $1 & "') & '"'
;~ ConsoleWrite('Final phase : ' & $txtexec & @LF)
$txtfinal = Execute($txtexec)
ConsoleWrite('Result    : ' & $txtfinal & @LF)

I've placed console outputs in order to make following the process easier. Intermediate variables are only there to optionally show progress.

It was a fun exercise to use only regexps and final execute but I strongly suspect a simpler approach can beat this one, just like Ascendant did in the meantime.

Edited March 15, 2011 by jchd

[Bowmore]

One more method for you to consider

$TEXT = "aabdabecfdbaebcfdfaaabceeabfeabbcccd"
MsgBox(0,"",Replace($TEXT))

Func Replace($TEXT)
    Local $RESULT = ""
    Local $aTmp = StringSplit($TEXT,"",2)
    If Not UBound($aTmp) Then Return ""
    $RESULT = $aTmp[0]
    For $i = 1 To UBound($aTmp) - 1
        If $aTmp[$i] <> $aTmp[$i-1] Then
          $RESULT &= $aTmp[$i]
        EndIf
    Next
    Return $RESULT
EndFunc

Edited March 15, 2011 by Bowmore

[GEOSoft]

Is there a reason to place a number in front of the string? I ask because it's a simple regex to replace repeating Characters with a single character.

$sStr = "aabdabecfdbaebcfdfaaabceeabfeabbcccd"
$sStr2 = StringRegExpReplace($sStr, "(\w)\1+", "$1")
MsgBox(0, "Result", $sStr & @CRLF & "becomes" & @CRLF & $sStr2)

EDIT:

You might be better off using

$sStr2 = StringRegExpReplace($sStr, ([[:alpha:]])\1+", "$1")

Edited March 16, 2011 by GEOSoft

[Malkey]

For better or for worse, here is the obligatory method that does not use regular expressions.

This example shows character "b" is not selected for repeat counting. But is easy to add to the repeating characters' string if required.

Local $txt = 'aabccccffedbbbfffdaccbefdeeaaacccccccc ccce'
Local $sAcceptChar = "acdef" ; Count only these characters (Note "b" is absent)
Local $sRetStr, $iCount

For $i = 1 To StringLen($txt)
    $iCount = 1
    While (StringMid($txt, $i, 1) = StringMid($txt, $i + 1, 1)) And _
            StringInStr($sAcceptChar, StringMid($txt, $i, 1))
        $iCount += 1
        $i += 1
    WEnd
    If $iCount = 1 Then $iCount = ""

    $sRetStr &= $iCount & StringMid($txt, $i, 1)
Next

ConsoleWrite($sRetStr & @CRLF)

[Andreik]

Thanks guys, for your examples.

@GEOSoft I need to know the number of occurences to can reconstruct the main string later. Thanks anyway for your examples that replace repeating characters with a single character, maybe I will need this too [it's good to know ]