String Management

[hallaplay835]

Hi everyone, I am pretty new to AutoIt.

I have opened a topic here to discuss string management.

I have had a read through the help file under the 'String Management' directory but I still cannot find a way to do what I am going to describe below without using extensive code.

What I want to do is the following. I want to compare the contents of a string to a set of rules.

The string must contain:

1. Less than 32 characters.

2. Only alphanumeric values (e.g Aa-Zz, 0-9). Spaces are also allowed.

3. One of the characters MUST be a letter.

4. However, there are two exceptions to rule no. 2. The string CAN contain also dashes '-' and/or underscores '_'.

Rule 1 is easy. I suppose you could use the StringLen() function.

The rest of the rules I do not know how to account for them in my code.

I suppose you could put all the rules in an If statement with And and Or operators, and everything should work. Then, in the Else section, you could throw an error.

Any plausible ideas as to how to solve this problem? Thanks in advance.

Edited April 14, 2011 by hallaplay835

[MrMitchell]

StringRegExp() should do it. GEOSoft has a nice tool to test it too

Check post and the link to PCRE Tool Kit in his sig...

Edited April 14, 2011 by MrMitchell

[Melba23]

hallaplay835,

Here is my idea of how it might work:

Global $aString[5]

$aString[0] = "123456789 12345678901234567-_A" ; Should pass

$aString[1] = "123456789012345678901234567-_0" ; Should fail - no A-Z

$aString[2] = "123456789012345678901234567890123" ; Should fail - too long

$aString[3] = "123456789 1234567890123456-_A"  ; Should Pass

$aString[4] = "ABCDEFGHIJKLMNOP?URSTUVWXYZ_-"  ; Should fail - contains ?


For $i = 0 To 4
    ; Check length
    If StringLen($aString[$i]) > 31 Then
        ConsoleWrite("Error - String " & $i & " is too long" & @CRLF)
    Else
        ; Check only digits, a-z (case-insensitive), dash, underscore or space
        If StringRegExp($aString[$i], "(?i)[^0-9a-z-_\x20]") Then
            ConsoleWrite("Error - String " & $i & " contains an illegal character" & @CRLF)
        Else
            ; Check there is at least 1 a-z (case-insensitive)
            If Not StringRegExp($aString[$i], "(?i)[a-z]") Then
                ConsoleWrite("Error - String " & $i & " does not contain at least one A-Z character" & @CRLF)
            Else
                ConsoleWrite("Good - String " & $i & " is correct" & @CRLF)
            EndIf
        EndIf
    EndIf
Next

I hope that gives you a start.

M23

Edit: Oh dear, George is looking at this thread - donning tin hat and awaiting incoming!

Edited April 14, 2011 by Melba23

[GEOSoft]

$sString = "The rest of the rules I do not "
_Match($sString)
$sString = "The rest of the rules I do not know how to account for them in my code"
_Match($sString)

Func _Match($sStr)
    If StringRegExp($sStr, "^[[:alnum:]_-\h]{1,31}$") Then
        MsgBox(4096, "Result", "We have a match")
    Else
        MsgBox(4096, "Result", "We have a problem")
    EndIf
EndFunc

If it's possible for the string to be a complete sentence where it could end in punctuation then I would change the expression to

"^[[:alnum:]_-\h]{1,31}[.?!]$"

[GEOSoft]

hallaplay835,

Here is my idea of how it might work:

<snipped>

M23

Edit: Oh dear, George is looking at this thread - donning tin hat and awaiting incoming!

You were close but you went the hard way. Classes are good in this situation and with the [:alnum:] class you don't have to set case insensitive.

EDIT:

I don't remember if I have been in touch since the last update but make sure you are using ver:3.0.1.9 of the toolkit.

Edited April 14, 2011 by GEOSoft

[jchd]

Hi George,

You don't enforce rule #3.

use a lookbehind assertion

[GEOSoft]

Hi George,

You don't enforce rule #3.

use a lookbehind assertion

Possible.

I just ran a quick test at the time and it worked. What did I miss?

[MrMitchell]

What did I miss?

That the string must contain at least 1 letter

Edited April 14, 2011 by MrMitchell

[GEOSoft]

That the string must contain at least 1 letter

I didn't read the post that he needed it that way. Easy fix as soon as I get a chance to look again.

[hallaplay835]

StringRegExp() should do it. GEOSoft has a nice tool to test it too

Check post and the link to PCRE Tool Kit in his sig...

Thank you MrMitchell.

[hallaplay835]

hallaplay835,

Here is my idea of how it might work:

Global $aString[5]

$aString[0] = "123456789 12345678901234567-_A" ; Should pass

$aString[1] = "123456789012345678901234567-_0" ; Should fail - no A-Z

$aString[2] = "123456789012345678901234567890123" ; Should fail - too long

$aString[3] = "123456789 1234567890123456-_A"  ; Should Pass

$aString[4] = "ABCDEFGHIJKLMNOP?URSTUVWXYZ_-"  ; Should fail - contains ?


For $i = 0 To 4
    ; Check length
    If StringLen($aString[$i]) > 31 Then
        ConsoleWrite("Error - String " & $i & " is too long" & @CRLF)
    Else
        ; Check only digits, a-z (case-insensitive), dash, underscore or space
        If StringRegExp($aString[$i], "(?i)[^0-9a-z-_\x20]") Then
            ConsoleWrite("Error - String " & $i & " contains an illegal character" & @CRLF)
        Else
            ; Check there is at least 1 a-z (case-insensitive)
            If Not StringRegExp($aString[$i], "(?i)[a-z]") Then
                ConsoleWrite("Error - String " & $i & " does not contain at least one A-Z character" & @CRLF)
            Else
                ConsoleWrite("Good - String " & $i & " is correct" & @CRLF)
            EndIf
        EndIf
    EndIf
Next

I hope that gives you a start.

M23

Edit: Oh dear, George is looking at this thread - donning tin hat and awaiting incoming!

Thanks a lot Melba23, but I think I prefer GEO Soft's approach using classes. Less code. Anyway, thanks.

[hallaplay835]

$sString = "The rest of the rules I do not "
_Match($sString)
$sString = "The rest of the rules I do not know how to account for them in my code"
_Match($sString)

Func _Match($sStr)
    If StringRegExp($sStr, "^[[:alnum:]_-\h]{1,31}$") Then
        MsgBox(4096, "Result", "We have a match")
    Else
        MsgBox(4096, "Result", "We have a problem")
    EndIf
EndFunc

If it's possible for the string to be a complete sentence where it could end in punctuation then I would change the expression to

"^[[:alnum:]_-\h]{1,31}[.?!]$"

Thank you very much for your answer, GEO Soft. It is all what I expected. However, you do not account for rule 3: the string must contain at least one letter. I am working on that right now. I have also found interesting the toolkit in your sig. Thanks for everything.

[jchd]

Use the pattern with a look-ahead: ^(?=[^[:alpha:]]*[[:alpha:]])[[:alnum:]_-\h]{1,31}$

That assertion is "zero or more non-letters followed by a letter", i.e. "at least one letter". Asserts don't consume anything in the subject, just test the condition.

Contrary to what I wrote in the spoiler few post ago: lookahead is simpler than lookbehind. A lookbehind assertion is restricted to fixed or rather bounded length.

So ^[[:alnum:]_-\h]{1,31}(?=[^[:alpha:]]*[[:alpha:]])$ gives an error due to the * in the assert (unbounded length)

But rewriten this way, it works: ^[[:alnum:]_-\h]{1,31}(?=[^[:alpha:]]{0,30}[[:alpha:]])$

Edited April 15, 2011 by jchd