Vista - strange behavior from Regex.Split & myString.IndexOf

I am working in C# ASP.NET framework 1.1 and
for some reason Regex.Split isn't working as expected.
When trying to split a string, Split is returning an array
with the entire string in element [0] and an empty string in element
[1].
I am trying two different ways (an ArrayList and a string array)
and both are doing that. Also, IndexOf is not working,
but StartsWith does.

The code:

using System.Text.RegularExpressions;
...
// some code to find CheckBoxList controls in
Request.Form
// look at Request.Form keys and split out delimiter
// which sometimes is "$" (ie "CheckBoxList1$1")
// and sometimes ":" (ie CheckBoxList1:1)
// (is there any property for what .NET uses?), so to
get delim we
// use a function to return 1st non alphanum char
after control name
string _sIndexDelim =
GetControlNameDelimiter(CheckBoxList1.ID);
foreach (string _key in
HttpContext.Current.Request.Form.AllKeys)
{
//FOR SOME ODD REASON, THIS DOESN'T WORK:
//if (_key.IndexOf(CheckBoxList1.ID +
_sIndexDelim, 1) > 0)
// BUT THIS WORKS (?)
if (_key.StartsWith(CheckBoxList1.ID +
_sIndexDelim))
{
_arrKey = new ArrayList();
_arrKey.AddRange(Regex.Split(_key,
_sIndexDelim)); // does not split!
string[] substrings = Regex.Split(_key,
_sIndexDelim); // does not split!

The values:

?_key
"CheckBoxList1$1"
?_sIndexDelim
"$"

?_arrKey[0]
"CheckBoxList1$1"
?_arrKey[1]
""

?substrings[0]
"CheckBoxList1$1"
?substrings[1]
""

Any ideas why this is happening?

Thanks

On Tue, 01 Apr 2008 17:55:16 -0700, <mad.scientist.jr@xxxxxx> wrote:
On the latter point, I don't understand what you mean. In the code you
posted, IndexOf() is going to be 0 if the string actually passes the test
that StartsWith() applies. But you are checking for "> 0". The two lines
you wrote are definitely not logically equivalent.

As far as Regex.Split() goes, surely this has to do with the fact that the
"$" character is a special regular expression character. If you don't
want to apply an actual regular expression to your string processing,
don't use Regex. That's where the name "Regex" comes from: "REGular
EXpression".

Use String.Split() instead.

Pete

Aha... string.IndexOf is 0 based. Silly mistake on my part. Duh

I can think of other places where I would want to use Regex.split
because it lets you split by more than one character. So you would
need to escape any delimiter characters that would be interpreted as a
regular expression? Did not know that.

Thanks

On Apr 1, 9:02 pm, "Peter Duniho" <NpOeStPe...@xxxxxx>
wrote:

On Tue, 01 Apr 2008 18:17:47 -0700, <mad.scientist.jr@xxxxxx> wrote:
String.Split() does too. Just pass it strings instead of characters.
Yes...if you want characters that would ordinarily be interpreted as a
control character for the regular expression, it needs to be escaped.
Just as is the case any time you might have special characters embedded in
a string (like quote, newline, etc. for plain old string literals).

Pete

Hello mad.scientist.jr@xxxxxx,

This isn't working because $ has a special meaning in Regular expressions
(end of line, or end of input).

To make sure the regex works with any input use string regex = Regex.Escape("$");

which should make it \$.


Jesse
--
Jesse Houwing
jesse.houwing at sogeti.nl