(Yet another)StringRegExp(SRE) Topic

[BAM5]

Hello all,

I'm creating a library that will allow you to do all things http(hopefully). I'm trying to make this user friendly so a url can be entered as "http://etc", "etc", "http://etc/etc", "http://etc/", etc.

So in order to make this easy on me I thought I'd do some SRE. Unfortunately for me I don't know SRE. So I read up. I read the StringRegExp tutorial by neogia and I downloaded the SRE Tester by Szhlopp.

I've been experimenting and so far I've come up with:

"(?i)(?:http://|)(.*)/"

Which works with http://www.google.com/ But not with http://www.google.com. I understand why, just not how to circumvent it. I also need to make that last forward slash be able to be the end of the line too. "[/|\z]"?

You might also get to answer my future questions about getting pages and stuff, but I think I'll be able to handle it once I have this figured out.

Thanks in advance, hope to get some meaningful replies.

Edited October 23, 2009 by BAM5

[memoryoverflow]

"(?i)(?:http://|)(.*)/"

... "[/|\z]"?

Not absolutely sure what the desired result is, but...

in your first pattern, you wouldn't need the null-string aternative, just a ?-quatifier (but the null alternitive spoils it anyway)

in the latter particle, the aternation specifier is wrong, the charater group [] includes aternation between those charakters.

So, something like

Local $i = 0, $asTest[4] = ['http://www.google.com/','http://www.google.com','www.google.com','www.google.com/sub']
For $i = 0 To UBound($asTest) - 1
    ConsoleWrite($asTest[$i] & ' => ' & StringRegExpReplace($asTest[$i], '(?i)(?>http\://|\A)([^/]*)(?>/\z|\z)','$1') & @crlf )
Next

would strip the surrounds and let you work on with a unified representation.

@ ResNullius: I bet that there's a simpler pattern doing the same job?

edit: typos

Edited October 23, 2009 by memoryoverflow

[ResNullius]

Don't know if it's simpler, just more compact:

Local $i = 0, $asTest[5] = ['http://www.google.com/','http://www.google.com','www.google.com','www.google.com/sub','http://google.com']
For $i = 0 To UBound($asTest) - 1
    ConsoleWrite($asTest[$i] & ' => ' & StringRegExpReplace($asTest[$i], '(?:https*:\/\/)([\w.]+)\/?[\w\.$]*','\1') & @crlf )
Next

[BAM5]

I've adapted this off of your previous posts

'(?:https*://|\A)([\w.]+)'

And it works perfectly, thanks!

Edit:Revised again

Edited October 23, 2009 by BAM5

[GEOSoft]

Or another one

"(?i)([hf]t+ps?:[\w\./]+)"

[BAM5]

All things HTTP, not FTP

[GEOSoft]

All things HTTP, not FTP

No matter. That gets either as well as https:

[SmOke_N]

This may come in handy for you down the road if you're going to be playing with URLs.

_URLSplit : http://www.autoitscript.com/forum/index.php?showtopic=36679&st=0&p=271254&#entry271254

[BAM5]

Ok, now there's more. I want to get the path of the url after the domain so I have this:

(?:https*://|\A)([\w.]+)(/.+)\?{0,1}

It works, but I don't want the question mark to be captured, which it will when there is one.

Thanks for all your help so far!

Edit:

O hey, thanks SmOke_N, I'll look at that when I get home, I have to get to class now.

Edited October 23, 2009 by BAM5