[Problem]Cutting Divs from Source code String read from saved HTML page.

[Mikeman27294]

Hey everyone,

I am working on a script for which I have a webpage saved. At the start of the program, the program loads the source code into a variable, removes all script and noscript, and then it must cut out a div called product (<div id="product">). The problem is that each time I try to write this function, I have problems detecting how many child divs it has nested in it. I have tried also writing a function that determines the amount of characters in the div, for trimming to the right of the div, but this has not worked. I dont wish to use an embedded browser for my program, I just want to work with the strings. Could anybody give me any pointers?

Thanks.

EDIT

I have found the following forum post:

The problem is though, that (as far as I am aware) it requires you to use the embedded browser or just use internet explorer, which will not help me as I just want to read the source code. I am aware that I can load the source into the browser and then hide the browser, but I would rather just "cut" the div out of the source.

Thanks.

SOLUTION

Thanks to thanlankon. Please note that this also returns the div tags (opening and closing) for the specified div. It also removes the " character from either side of the ID name.

_HTMLGetDIVCode('html code here', 'id of div here')
Func _HTMLGetDIVCode($html_code, $div_id)
    Local $o_htmlfile = ObjCreate('HTMLFILE')
    If Not IsObj($o_htmlfile) Then Return SetError(-1, 0, '')
    $o_htmlfile.open()
    $o_htmlfile.write($html_code)
    $o_htmlfile.close()
    Local $div = $o_htmlfile.getElementByID($div_id)
    If Not IsObj($div) Then Return SetError(0, 0, '')
    Return $div.outerHTML
EndFunc

Edited December 9, 2011 by Mikeman27294

[Xenobiologist]

Please show the text before and after the cutting. That way it should be easy to have a look.

[Mikeman27294]

Ok, I will attach 2 HTML documents.

Before

After

If anybody is in query about whether these files contain viruses or not, then I can email them to you (They are just 2 HTML files called File and Result, respectively). It would also be appreciated if anybody who downloads them could verify that they are virus free (Just to clear the doubts), and thanks to anybody who does.

This is what I was hoping to achieve from the program. I wish to remove the div named product, and write that to a file. I already have my program removing script and noscript, which is also demonstrated in the result file.

If you want to view the original page, it was saved from here:

http://www.oo.com.au/Prima_72_Bottle_Wine_Cooler_-__P4022.cfm

Thanks in advanced for any help.

[Xenobiologist]

Download doesn't work.

_________________________________________________________________________

This file is currently set to private.

When a file is set to private by its owner only the owner of the file can access it. If you are the owner of the file please log into your account to access this file.

If you believe you have reached this page in error, please contact support.

Click here to view our help resources

Edited December 8, 2011 by Xenobiologist

[Mikeman27294]

Sorry, I thought you would still be able to download them. I will fix that.

EDIT

Ok, I fixed that. Thankyou for letting me know.

Edited December 8, 2011 by Mikeman27294

[Xenobiologist]

How do you determine it manually? All I can see is starting at id="product and end at

</div>

</div>

</div>

</div>

Everything between ist what you extracted.

[thanlankon]

I found an object called HTMLFile, like Shell.Explorer.2 (use in _IECreateEmbedded), but the object HTMLFile just work with source code, not display anything like Shell.Explorer.2

How about the code below, Mikeman:

_HTMLGetDIVCode('html code here', 'id of div here')

Func _HTMLGetDIVCode($html_code, $div_id)
    Local $o_htmlfile = ObjCreate('HTMLFILE')

    If Not IsObj($o_htmlfile) Then Return SetError(-1, 0, '')

    $o_htmlfile.open()
    $o_htmlfile.write($html_code)
    $o_htmlfile.close()

    Local $div = $o_htmlfile.getElementByID($div_id)

    If Not IsObj($div) Then Return SetError(0, 0, '')

    Return $div.outerHTML
EndFunc

[kylomas]

thanlankon,

Did you find any MS doc for "HTMLfile"?

I had a VBS routine using the IE parsing engine and this object to extract just what the OP is looking for, however, I have not been able to successfully translate it to AutoIT...finally gave up and used SRE for parsing.

kylomas

[Mikeman27294]

How do you determine it manually? All I can see is starting at id="product and end at

</div>

</div>

</div>

</div>

Everything between ist what you extracted.

What I was doing is that I had a do while loop, and a variable which incremented each time a div tag was detected, and the opposite when the close div tag. This would return the amount of lines (the starting line is known), and I could then read each line and write it to the file, but this just returned 0 :S

I found an object called HTMLFile, like Shell.Explorer.2 (use in _IECreateEmbedded), but the object HTMLFile just work with source code, not display anything like Shell.Explorer.2

How about the code below, Mikeman:

_HTMLGetDIVCode('html code here', 'id of div here')

Func _HTMLGetDIVCode($html_code, $div_id)
    Local $o_htmlfile = ObjCreate('HTMLFILE')

    If Not IsObj($o_htmlfile) Then Return SetError(-1, 0, '')

    $o_htmlfile.open()
    $o_htmlfile.write($html_code)
    $o_htmlfile.close()

    Local $div = $o_htmlfile.getElementByID($div_id)

    If Not IsObj($div) Then Return SetError(0, 0, '')

    Return $div.outerHTML
EndFunc

That looks good. I will place it in my code and see how well it works, thanks Edited December 9, 2011 by Mikeman27294

[Mikeman27294]

Yep, that code works. I will paste it in a spoiler in my first post for future reference, thankyou, thanlankon.

[thanlankon]

thanlankon,

Did you find any MS doc for "HTMLfile"?

I had a VBS routine using the IE parsing engine and this object to extract just what the OP is looking for, however, I have not been able to successfully translate it to AutoIT...finally gave up and used SRE for parsing.

kylomas

actually no, kylomas. the methods and properties of HTMLFILE like object "document" in javascript, I think so.(of course they have some differences but I'm not sure about them.) I don't know much about VBS but can you post your VBS code here? maybe I can help you to translate them.

Yep, that code works. I will paste it in a spoiler in my first post for future reference, thankyou, thanlankon.

not at all, Mikeman

[Mikeman27294]

Ok, so I was using that code, and it was working quite well, until I found that it was stripping the " characters from div IDs. Are there any work-arounds for that? I dont understand what the problem is with it but I know that before it enters that function, the " characters are there and they arent when the string leaves it :S

[czardas]

I'm not entirely sure what data you are after. I suggest mapping the opening and closing tag positions using StringInStr with the occurence parameter. Then you can grab whatever characters you wish between any two tags. Things like this can also be done using regular expressions.

[Mikeman27294]

Basically, I want to go onto an online webstore, and get data about the product being sold.

I have tried using string in str but that didnt work properly (Not a clue why, I use that function all the time :S). I also dont know how to do RegEx really, so for the time being, I will just stay away from it till I have a bit of free time to learn it.

[czardas]

Here's a simple way to get starting positions of the opening divs. The code is just an example of how you might go about solving this problem. To make it work you will have to make some modifications. I hope it gives you some ideas.

Local $html = "<div><div><div>Hello World</div></div></div>"

Local $iDivCount = 1, $sDelimStr = ""
While StringInStr ( $html, "<div" , 0 , $iDivCount)
    $sDelimStr &= StringInStr( $html, "<div" , 0 , $iDivCount) & ","
    $iDivCount += 1
WEnd
MsgBox(0, "Staring positions", StringTrimRight($sDelimStr, 1))

[Mikeman27294]

Earlier, the one I tried to have a go at basically read each line individually, and if '<div' was in it, it would increment a variable. then if it found '</div>', it would do the opposite.

What I would really need is the beginning, and the ends of each one.

I might have another crack at it soon.

[czardas]

I don't see any other way than adding and subtracting increments. But my suggestion is to simplify the process by first just looking for the positions of the tags. Once you know this you can sort them in order and select between them more easily.

[ripdad]

Something like this might work if you're after the specific info.

Add your own divs afterwards if you need them.

Local $url = 'http://www.oo.com.au/Prima_72_Bottle_Wine_Cooler_-__P4022.cfm'
Local $htm = BinaryToString(InetRead($url, 1))
;
Local $s = ''
$s &= GetPageInfo($htm, 'inner">' & @CRLF & '<h1>', '</h1>') & @CRLF
$s &= 'Price: ' & GetPageInfo($htm, 'Price: <strong>', '</strong>') & @CRLF
$s &= GetPageInfo($htm, 'rrp"><strong>', '</strong>') & @CRLF
MsgBox(0, '', $s)
;
Exit
;
Func GetPageInfo($htm, $tag1, $tag2)
    Local $a = StringRegExp($htm, '(?i)(?s)' & $tag1 & '(.*?)' & $tag2, 3)
    If IsArray($a) Then Return $a[0]
EndFunc
;

[thanlankon]

Ok, so I was using that code, and it was working quite well, until I found that it was stripping the " characters from div IDs. Are there any work-arounds for that? I dont understand what the problem is with it but I know that before it enters that function, the " characters are there and they arent when the string leaves it :S

Because my code use HTMLFILE, the object HTMLFILE automatically remove the " character that wraps value of some properties of the tag like ID. but I wonder what the problem with it is. Although the " character is removed but you can parse the code normally. I think it does not matter much.

[Mikeman27294]

Because my code use HTMLFILE, the object HTMLFILE automatically remove the " character that wraps value of some properties of the tag like ID. but I wonder what the problem with it is. Although the " character is removed but you can parse the code normally. I think it does not matter much.

Ok, thanks.