Escape special characters

[water]

Hi PCRE gurus,

I'm looking for a regexp to escape special characters with a "" plus the hex value of the character as described here:

The LDAP filter specification assigns special meaning to the following characters: * ( ) NUL. The NUL character is ASCII 00.

In LDAP filters these 5 characters should be escaped with the backslash escape character, followed by the two digit ASCII hex representation of the character. The following table documents this:

*     2A
(     28
)     29
    5C
NUL     00

Any assistance is greatly appreciated!

Edited December 13, 2013 by water

[BugFix]

To escape hex values, the pattern is: xhh.

In your case: x2A, x28 and so on.

Is it what you looking for?

[water]

I do not need to escape hex values but to replace the character by its hex value. If one of the special characters is already escaped it should be left alone.

Example: * should become 2A, 2A should be left unchanged

I already use a RegExp (for another problem) to escape special characters by prefixing them with a backslash.

Example becomes , remains unchanged.

The RegExp for this is:

$sText = StringRegExpReplace($sText, "(?<!\\)([" & $sEscapeChar & "])", "\\$1")

Edited December 14, 2013 by water

[BugFix]

OK, than try this

expandcollapse popup

$s = _
"*,     \2A" & @CRLF & _
"(,     \28" & @CRLF & _
"),     \29" & @CRLF & _
"\,     \5C" & @CRLF & _
"NUL,     \00"
 
$sPattern = "(\*|\(|\)|\\(?![\dA-F])|NUL)"
 
$sReplaced = _StringRegExpReplace_Callback($s, $sPattern, '_Escape("$1")')
ConsoleWrite($sReplaced & @LF)
 
Func _Escape($1)
    Switch $1
        Case '*'
            Return '\2A'
        Case '('
            Return '\28'
        Case ')'
            Return '\29'
        Case '\'
            Return '\5C'
        Case 'NUL'
            Return '\00'
    EndSwitch
EndFunc
 
 
; #FUNCTION# ====================================================================================================================
; Name ..........: _StringRegExpReplace_Callback
; Description ...: Replaces ByRef all matches in a string by manipulation of the matches in a callback function
; Parameters ....: $sString    String to manipulate
; ...............: $sPattern   RegExp pattern for the match
; ...............: $sCallback  Callback function as string, i.e. "MyCallback('$1')" -- '$1' represents the match
; .....optional..: $sBefore    String to insert before the manipulated match
; .....optional..: $sAfter     String to insert after the manipulated match
; Return values .: Success     1
; ...............: Failure     0  set error = 1
; Author ........: Taz77
; ===============================================================================================================================
Func _StringRegExpReplace_Callback(ByRef $sString, $sPattern, $sCallback, $sBefore = '', $sAfter = '')
    $sString = Execute("'" & StringRegExpReplace(StringReplace($sString, "'", Chr(26), 0, 2), $sPattern, $sBefore & "'&" & $sCallback & "&'" & $sAfter) & "'")
    If @error Then
        ConsoleWrite('_StringRegExpReplace_Callback error! Pattern: "' & $sPattern & '", Callback: "' & $sCallback & '"' & @LF)
        Return SetError(1,0,0)
    EndIf
    $sString = StringReplace($sString, Chr(26), "'", 0, 2)
    Return 1
EndFunc  ;==>_StringRegExpReplace_Callback

[mikell]

?

$str = "this is a star * and a backslash \, these are ( parentheses ) and this is NUL" & @crlf
$str &= _
"*,     \2A" & @CRLF & _
"(,     \28" & @CRLF & _
"),     \29" & @CRLF & _
"\,     \5C" & @CRLF & _
"NUL,     \00"

$res = StringRegExpReplace(Execute('"' & StringRegExpReplace($str, '(\*|\(|\)|\\(?![[:xdigit:]]{2}))', '" & "\\" & hex(ascw("$1"),2) &  "') & '"'), '(NUL)', "\\00")

Msgbox(0,"", $res)

Edited December 13, 2013 by mikell

[jchd]

I see that mikell is having more and more fun with REs

[water]

That is exactly what I was looking for and what I could never have done myself!

Thanks a lot!

[Malkey]

....

Example: * should become 2A, 2A should be left unchanged 

......

Example becomes , remains unchanged. 

....

That is exactly what I was looking for .......

water

An observation.

You have either changed exactly what you were looking for, or, you haven't realized why no one didn't put "" in their test string.

[water]

Thanks for your reply.

I've exactly got what I need. The solution for the problem I described in my OP doesn't need to check for already escaped special characters because the special character is not only escaped but changed too ("(" becomes "28").

The "" is taken from another solved problem where I needed to escape special characters.

The solution now works fine until a user proves me wrong

Edited December 14, 2013 by water

[guinness]

Nice mikell. I don't know if this is a problem as such, but NULL will be converted to 00L, so I would change it to 'NUL(?!w)'. Notice how I removed the unnecessary group.

[mikell]

I see that mikell is having more and more fun with REs

That's it, the teacher making fun of his student

guiness, you're right but in this case I do prefer NULL? (without grouping...)

[guinness]

But that's wrong as NULL shouldn't be converted, only NUL. Your change will convert both.

[mikell]

Ok sorry I misread, I thought that NUL and NULL should both be converted

So I can't see the reason why NULL should not be converted to 00L