[RESOLVED] [2005] - Help with regular expression

[nolim8ts]

Hi All,

I am trying to extract 2 values from a string and place them into separate groups.

String:

1.6919</td>
<td class="Normal" align="Right">1.6835</td>
<td class="Normal" align="Right">27 Mar 08</td>
<td class="Normal" align="Right">6 Monthly</td>

Values are 1.6919 and 1.6835.

Below is the code I am using:

Const myPattern As String = "(?<entryprice>[\d.{1}+[\d.{4}])" & _ ".*?" & _ "(?<entryprice>[\d.{1}+[\d.{4}])" & _ ".*?"

For each m As Match In Regex.Matches(mystring, myPattern)
Dim EntryPriceE As String = m.groups("entryprice").ToString
Dim ExitPriceE As String = m.Groups("exitprice").ToString

Next

I have read a whole bunch of regex articles but still can't seem to get it correct.

[vijy]

in all cases the values must be in this same case ah???
*.**** (1.6919 and 1.6835)
i.e with decimal...

[nolim8ts]

Yes, that is correct.

These values are coming from a stock market website so the values could even be:

2.4512
12.1451
12.45
321.4567
4586.1287

etc...

[vijy]

^([0-9]+^).^([0-9]+^)

this Regex enough to find the string in the file.

[nolim8ts]

Using the following pattern's returns nothing:

Const myPattern As String = "(?<entryprice>^([0-9]+^).^([0-9]+^))"

Const myPattern As String = "(?<entryprice>^([0-9]+^).^([0-9]+^))" & _ '".*?"

[stanav]

Please post a larger sample data that includes the text before the 1st value to be extracted.

[nolim8ts]

Hi,

There is no text before the first value because I start the cut at

<td class="Normal" align="right">

and end at

</td>

.

Looking at the source of www.morningstar.com.au, this is the exact data sample I am extracting from.

<tr valign="top" bgcolor="#f5f5f5">
<td><a class="Normal">Sandhurst BMF - Sandhurst Industrial Share Fund</a></td>
<td class="Normal" align="right">1.6821</td>
<td class="Normal" align="right">1.6737</td>
<td class="Normal" align="right">28 Mar 08</td>

<td class="Normal" align="right">6 Monthly</td>
</tr>

[vijy]

Here three RegEx in three diferent script to find the string. its work fine for me, and find the string,.

Ultra-edit:- ^([0-9]+^).^([0-9]+^)
Perl:- \d+\.\d+
VBA:- ([0-9]{1,}).([0-9]{1,})

[nolim8ts]

Cheers vijy for the RegEx. Testing it on www.regexlib.com works flawlessly. The next question I guess is to separate those two values into strings etc... Any suggestions?

[stanav]

The pattern Viji gave you actually only searches for any positive decimal value. It won't pick up the negative sign if the value is negative. Besides, it'll search for all decimal values regardless they are enclosed by <td> tags or not.
Try this:

'Use a webclient to download the source from the link your gave above
        Dim wc As New System.Net.WebClient()
        Dim sourceHtml As String = wc.DownloadString("http://www.morningstar.com.au")

        'Declare a list to hold prices
        Dim priceList As New List(Of Decimal)

        Dim pattern As String = "(?i-msnx:(?<=<td class=""normal"" align=""right"">)[\s+-]?(\d+\.?\d*|\d*\.?\d+)\s?(?=</td>))"
        Dim matches As System.Text.RegularExpressions.MatchCollection = System.Text.RegularExpressions.Regex.Matches(sourceHtml, pattern)
        For Each match As System.Text.RegularExpressions.Match In matches
            priceList.Add(Decimal.Parse(match.Value))
        Next

[nolim8ts]

Cheers stanav for the code. I tried your regular expression but that did not returns any matches (using VS and RegExLib.com).

I have now created a new solution that uses a list (from your posted code), vijy's regular expression and more regular expressions to remove HTML Tags.

This is all working fine, but I have run into a problem with the "Matches" part. I know that each match must contain 2 values. If a user was to enter a invalid stock code, the match would either = 0 Or 1 (From my testing). How can I detect this and log the error? Below is the code I am using

' Read The Response Into A String

                Dim strHTML As String = str.ReadToEnd

                Dim strStartCut As String
                Dim strEndCut As String
                Dim strStartTag As String
                Dim strEndTag As String
                Dim strDesired As String

                ' Cut Out Only What We Need

                strStartTag = "<td class=""Normal"" align=""right"">"
                strEndTag = "</tr>"

                strStartCut = (InStr(1, strHTML, strStartTag, vbTextCompare) + Len(strStartTag))
                strEndCut = (InStr(strStartCut, strHTML, strEndTag, vbTextCompare))

                strDesired = Mid(strHTML, strStartCut, (strEndCut - strStartCut))

                ' Strip HTML Tags

                Dim Pattern As String = "<[^>]*>"

                Dim R As New Regex(Pattern)

                strDesired = R.Replace(strDesired, "")

                Dim Pattern2 As String = "[<]"

                Dim S As New Regex(Pattern2)

                strDesired = S.Replace(strDesired, "<")

                Dim Pattern3 As String = "[>]"

                Dim T As New Regex(Pattern3)

                strDesired = T.Replace(strDesired, ">")

                'Declare a list to hold prices

                Dim priceList As New List(Of Decimal)

                Dim expression As String = "([0-9]{1,}).([0-9]{1,})"

                Dim matches As System.Text.RegularExpressions.MatchCollection = System.Text.RegularExpressions.Regex.Matches(strDesired, expression)

                For Each Match As System.Text.RegularExpressions.Match In matches

                    If Match.Value = 0 Or 1 Then

Me.RichTextBoxError.Text += Date.Now + " - Data Not Available For UNLM Code " + MorningStarCode + vbCrLf

                    Else

priceList.Add(Decimal.Parse(Match.Value))
                        'MessageBox.Show(Decimal.Parse(Match.Value))

                    End If

                Next

                'Dim ExtractedPrices As String = "2B" & vbNewLine & ("00" & MATECode) & vbNewLine & sDate & vbNewLine & ("000000" & priceList.Item(0) & "00") & vbNewLine & ("000000" & priceList.Item(1) & "00")
                'Dim Lines() As String = Split(ExtractedPrices.ToString, vbCrLf)
                'Me.DataGridViewDD.Rows.Add(Lines)

[stanav]

First of all, your pattern is searching for numeric values in decimal format (that is, 123.456 for example). So none of the matches will be 0 or 1. Secondly, match.Value returns a string, and 0 or 1 is an Integer. If you're to test a match against 0 or 1, you have to do this

If match.value = "0" OrElse match.Value = "1" Then

End If

I strongly suggest you to turn Option Strict On. It helps you catch a lot of silly errors like this and it also makes your application run faster.

[nolim8ts]

Cheers for the suggestion in using "Option Strict On". Found alot of errors in my code which were able to be corrected. As for determining if the match holds two values, I used the following code:

 For Each Match As System.Text.RegularExpressions.Match in matches

pricelist.add(Decimal.Parse(Match.Value)

Next

If pricelist.count = "2" then

'add values to datagridview

Else

'Log error to richtextbox

End If