Regexp to determine if string contains number

[c.haslam]

I remember that a few months ago someone (I think ProgAndy) posted a regexp which determines whether a string contains a number, including a floating-point number.

I can't find it on the forum. Please would someone point me to it.

[Melba23]

c.haslam,

Not the one for which you are looking, but I have used this RegEx with success:

expandcollapse popup

#include <GUIConstantsEx.au3>
#include <WindowsConstants.au3>
#include <EditConstants.au3>

$iMaxDecimalCount = 2

$hGui = GUICreate('?????? ?????', 220, 180)
$cInput = GUICtrlCreateInput('', 10, 10, 200, 20);, -1, $WS_EX_STATICEDGE)
GUISetState()
GUIRegisterMsg($WM_COMMAND, 'WM_COMMAND')

While 1
    Switch GUIGetMsg()
        Case $GUI_EVENT_CLOSE
            Exit
    EndSwitch

WEnd



Func WM_COMMAND($hWnd, $iMsg, $wParam, $lParam)
    If $ilParam <> GUICtrlGetHandle($cInput) Then Return $GUI_RUNDEFMSG
    Local $nNotifyCode, $nID, $sText

    $nNotifyCode = BitShift($iwParam, 16)
    Static $sPrevText = ''

    Switch $nNotifyCode

        Case $EN_CHANGE
            $sText = GUICtrlRead($cInput)
            If Not StringRegExp($sText, '^(-)?(\d*\.\d{0,' & $iMaxDecimalCount & '}|(\d{1,3}|^)(?:,\d{0,3})+(?:(?<=\d{3})\.\d{0,' & $iMaxDecimalCount & '})?|\d*)$') Then
                GUICtrlSetData($cInput, $sPrevText)
            Else
                $sPrevText = $sText

            EndIf

    EndSwitch



    Return $GUI_RUNDEFMSG
EndFunc   ;==>WM_COMMAND

M23

[mikell]

A little simpler, this one checks if the string contains only a number
Returns 1 for  "75",  ".75",  "0.75", but not "75."

Msgbox(0,"", StringRegExp($sString, '^\d*\.?\d+$') )

 

[c.haslam]

Thank you both.

I thought I had stored ProgAndy's code somewhere on my PC. I had, in an unlikely place. His regexp seems to work for all formats of number:

Func IsNumeric($vValue)
    Return (StringRegExp($vValue, "(^[+-]?[0-9]+\.?[0-9]*$|^0x[0-9A-Fa-f]{1,8}$|^[0-9]+\.?[0-9]*[Ee][+-][0-9]+$)") = 1)
 EndFunc  ;==>_IsNumeric

 

[kylomas]

c.haslam,

Run the following and note the results...

ConsoleWrite(IsNumeric('75.') & @CRLF)
ConsoleWrite(mikell('75.') & @CRLF)

Func IsNumeric($vValue)
    Return (StringRegExp($vValue, "(^[+-]?[0-9]+\.?[0-9]*$|^0x[0-9A-Fa-f]{1,8}$|^[0-9]+\.?[0-9]*[Ee][+-][0-9]+$)") = 1)
EndFunc  ;==>_IsNumeric



func mikell($num)
    return StringRegExp($num, '^\d*\.?\d+$') = 1
endfunc

kylomas

[jchd]

expandcollapse popup

#include <Array.au3>

Local $aNb = [ _
    ["0x0", 0, 0, 0], _
    ["0x01", 0, 0, 0], _
    ["0x012", 0, 0, 0], _
    ["0x0123", 0, 0, 0], _
    ["0x01234", 0, 0, 0], _
    ["0x012345", 0, 0, 0], _
    ["0x0123456", 0, 0, 0], _
    ["0x01234657", 0, 0, 0], _
    ["0x012346578", 0, 0, 0], _
    ["0x0123465789", 0, 0, 0], _
    ["0x0123465789A", 0, 0, 0], _
    ["0x0123465789AB", 0, 0, 0], _
    ["0x0123465789ABC", 0, 0, 0], _
    ["0x0123465789ABC", 0, 0, 0], _
    ["0x0123465789ABCD", 0, 0, 0], _
    ["0x0123465789ABCDE", 0, 0, 0], _
    ["0x0123465789ABCDEF", 0, 0, 0], _
    ["123", 0, 0, 0], _
    ["+123", 0, 0, 0], _
    ["-123", 0, 0, 0], _
    ["123.", 0, 0, 0], _
    ["+123.", 0, 0, 0], _
    ["-123.", 0, 0, 0], _
    [".123", 0, 0, 0], _
    ["+.123", 0, 0, 0], _
    ["-.123", 0, 0, 0], _
    ["1.23", 0, 0, 0], _
    ["+1.23", 0, 0, 0], _
    ["-1.23", 0, 0, 0], _
    ["123e4", 0, 0, 0], _
    ["+123e4", 0, 0, 0], _
    ["-123e4", 0, 0, 0], _
    ["123.e4", 0, 0, 0], _
    ["+123.e4", 0, 0, 0], _
    ["-123.e4", 0, 0, 0], _
    [".123e4", 0, 0, 0], _
    ["+.123e4", 0, 0, 0], _
    ["-.123e4", 0, 0, 0], _
    ["1.23e4", 0, 0, 0], _
    ["+1.23e4", 0, 0, 0], _
    ["-1.23e4", 0, 0, 0], _
    ["123e+4", 0, 0, 0], _
    ["+123e+4", 0, 0, 0], _
    ["-123e+4", 0, 0, 0], _
    ["123.e+4", 0, 0, 0], _
    ["+123.e+4", 0, 0, 0], _
    ["-123.e+4", 0, 0, 0], _
    [".123e+4", 0, 0, 0], _
    ["+.123e+4", 0, 0, 0], _
    ["-.123e+4", 0, 0, 0], _
    ["1.23e+4", 0, 0, 0], _
    ["+1.23e+4", 0, 0, 0], _
    ["-1.23e+4", 0, 0, 0], _
    ["123e-4", 0, 0, 0], _
    ["+123e-4", 0, 0, 0], _
    ["-123e-4", 0, 0, 0], _
    ["123.e-4", 0, 0, 0], _
    ["+123.e-4", 0, 0, 0], _
    ["-123.e-4", 0, 0, 0], _
    [".123e-4", 0, 0, 0], _
    ["+.123e-4", 0, 0, 0], _
    ["-.123e-4", 0, 0, 0], _
    ["1.23e-4", 0, 0, 0], _
    ["+1.23e-4", 0, 0, 0], _
    ["-1.23e-4", 0, 0, 0] _
]

For $i = 0 To UBound($aNb) - 1
    $aNb[$i][1] = IsNumeric($aNb[$i][0])
    $aNb[$i][2] = mikell($aNb[$i][0])
    $aNb[$i][3] = jchd($aNb[$i][0])
Next

_ArrayDisplay($aNb, 'What is numeric?', "", 0, Default, 'String|IsNumeric|mikell|jchd')

Func IsNumeric($vValue)
    Return (StringRegExp($vValue, "(^[+-]?[0-9]+\.?[0-9]*$|^0x[0-9A-Fa-f]{1,8}$|^[0-9]+\.?[0-9]*[Ee][+-][0-9]+$)") = 1)
EndFunc  ;==>_IsNumeric

func mikell($num)
    return StringRegExp($num, '^\d*\.?\d+$') = 1
endfunc

func jchd($num)
    return StringRegExp($num, '(?ix)^(0x[[:xdigit:]]{1,8}? | [+-]?\d*\.?\d*(?:E[+-]?\d+)?)$') = 1
endfunc

Signed hex values are uncommon but it would be trivial to deal with.

Edited September 3, 2015 by jchd
Cut & paste typo in regexp

[mikell]

It's definitely not comparable with my code which is far less ambitious 

[jchd]

You're a shy lovely cat...

[guinness]

You're a shy lovely cat...

I am not fooled by that image.

[c.haslam]

Because AutoIt accepts e.g. 1.2e5 I would make one change  in IsNumeric's regexp:

"(^[+-]?[0-9]+\.?[0-9]*$|^0x[0-9A-Fa-f]{1,8}$|^[0-9]+\.?[0-9]*[Ee][+-]?[0-9]+$)"

 

[mikell]

Instead of tinkering you should use directly jchd's expression 

[ViciousXUSMC]

Everytime I open up a good regex thread...

 

Edited September 1, 2015 by ViciousXUSMC

[jchd]

Buy a helmet?

[c.haslam]

jchd,

I am wondering why

[@*{]?

is in your regexp.

[jchd]

It was part of a much more involved expression and I didn't remind/notice it was there. There was also an optional parasitic closing }.

[c.haslam]

Thanks for explaining. I thought that you knew even more than I thought you did about regexps, which is already much more than I know!

BTW for my current project, which is experimental (!), I have mostly abandoned regexps (except for detecting whether a line of code contains a call to a function). I am heeding your advice to keep regexps as simple as possible. Instead, I am using a function that I wrote. It does the same as StringSplit("string", "delimiters",$STR_CHRSPLIT + $STR_NOCOUNT (2)), but also returns the delimiters that it found, and can skip literal strings:

expandcollapse popup

; Like StringSplit except returns delimiters and can skip quotations
; if first char in str is a separator, wordsVec[0] = ''
; if last char in str is a separator, last el of wordsVec = ''
Func StrParseDelims($str,$seps, ByRef $delimsVec,$bSkipQuotations=False)
    Local $wordP=1,$wordPs='',$wordLens='',$q=0,$p,$delims=''
    While True
        Local $ch = StringMid($str,$wordP,1)
        If $bSkipQuotations and ($ch='"' or $ch="'") Then
            $p = StringInStr($str,$ch,1,1,$wordP+1)
        Else
            $p = $wordP
        EndIf
        Local $pdelim = StrScanEq($str,$p,$seps)
        $wordPs &= '?'&$wordP
        $wordLens &= '?'&($pdelim-$wordP)
        $q += 1
        If $pdelim>StringLen($str) Then ExitLoop
        $delims &= StringMid($str,$pdelim,1)
        $wordP =  $pdelim + 1
    WEnd
    Local $wordsVec[$q]
    Local $posVec = StringSplit(StringMid($wordPs,2),'?',2) ; char, 0-based
    Local $wordLensVec = StringSplit(StringMid($wordLens,2),'?',2)
    For $i = 0 to $q-2
        $wordsVec[$i] = StringMid($str,$posVec[$i],$wordLensVec[$i])
    Next
    If $wordLensVec[$q-1]<>0 Then
        $wordsVec[$q-1] = StringMid($str,$posVec[$q-1],$wordLensVec[$q-1])
    EndIf
    $delimsVec = StringSplit($delims,'',2)
    Return $wordsVec
EndFunc

; Get pos of first of $chs chars in $str after pos $pst
Func StrScanEq($str,$pst,$chs)
    Local $ret = 99999
    Local $chsVec =  StringSplit($chs,'',2) ; char, 0-based
    For $el In $chsVec
        Local $p = StringInStr($str,$el,1,1,$pst)   ; case-sens, 1st
        If $p<>0 and $p<$ret Then
            $ret = $p
        EndIf
    Next
    If $ret=99999 Then
        $ret = StringLen($str) + 1
    EndIf
    Return $ret
EndFunc

 

Edited September 3, 2015 by c.haslam

[jchd]

I'm afraid I don't fully get it/ Can you give a spec or at least provide a few couples of typical or pathalogical inputs and corresponding expected output?

[c.haslam]

Here is a (contrived) simple example:

Local $code = "Local $gNowSecs = NowSecs($a,34)"
Local $delimsVec
Local $wordsVec = StrParseDelims($code,' (,)',$delimsVec)
For $i = 0 to UBound($wordsVec)-2
    If StringLeft($wordsVec[$i],1)='$' Then
        Switch $delimsVec[$i]
            Case ' '
                Local $returns = $wordsVec[$i]
            Case ',',')'
                Local $param = $wordsVec[$i]
        EndSwitch
    ElseIf $delimsVec[$i]='(' Then
        Local $function = $wordsVec[$i]
    EndIf
Next
MsgBox(0,'','$returns '&$returns&' $function '&$function&' $param '&$param)

 

[jchd]

I don't believe this can be made robust enough for the general real-world case. I tried:

Local $code = "Local $gNowSecs = Mod(NowSecs($a,34), 5)"
Local $code = "Local $gNowSecs = Mod(NowSecs($a,'34'), 5)"
Local $code = "Local $gNowSecs = 'Mod(NowSecs($a,''34''), 5)'"
Local $code = "Local $gNowSecs = String('Mod(NowSecs($a,''34''), 5)')"

but the result didn't change. Also this falls apart:

Local $code = "Local $gNowSecs = NowSecs('$a',34)"
Local $code = "Local $gNowSecs = NowSecs(""$a"",34)"

What's the actual goal?

[corgano]

I just had to add to this thread, if you want to learn regexp then regexr.com is an awesome place to start. You can paste what you want to match into the main box (html, chat logs, whatever) and then try to make a pattern at the top and it will try to match your pattern on the fly

You can also paste one of the patterns you see here in the top bar and hover your mouse over the various elements, and it will give you a description of what each one does

It is always my go-to tool when trying to make regular expressions, and is well worth bookmarking.

Edited September 4, 2015 by corgano