Post Reply 
sidki-config: Feb 13 2009
Apr. 27, 2009, 11:29 AM
Post: #61
RE: sidki-config: Feb 13 2009
WOW!! JD5000!!! hohoho Great to see you!!!
Add Thank You Quote this message in a reply
Apr. 27, 2009, 11:50 AM
Post: #62
RE: sidki-config: Feb 13 2009
(Apr. 27, 2009 11:27 AM)JD5000 Wrote:  Great to see you're keeping your config going old friend. Eyes Closed Smile

How nice to hear from you! I hope you're doing fine.
The config, well, yes, although i'm supposed to do other things...
Add Thank You Quote this message in a reply
Apr. 27, 2009, 12:14 PM
Post: #63
RE: sidki-config: Feb 13 2009
welcome back, Cotter...

(something like that, lol...)
Add Thank You Quote this message in a reply
Apr. 27, 2009, 10:38 PM
Post: #64
RE: sidki-config: Feb 13 2009
hmmmm...............is proxo going to get a major upgrade with all the "old" guys coming back?
Add Thank You Quote this message in a reply
Apr. 28, 2009, 02:51 AM
Post: #65
RE: sidki-config: Feb 13 2009
Perhaps? Wink Having Proxomitron-veterans returning increases the creativity and "morale" of everyone, so I think we can expect more "breakthroughs" and innovation!
Visit this user's website
Add Thank You Quote this message in a reply
May. 21, 2009, 09:02 AM
Post: #66
RE: sidki-config: Feb 13 2009
Maybe a bug in the Manage: Specific Containers on sel. Sites filter.

Original code:
Code:
$INEST(<$TST(\1),((^$TST(script=1*))|$TST(script=1*)(^*</script))*,</$TST(\1))</$TST(\1) >

Notice the </$TST(\1) in the $INEST() will cause below </div-->, which is indeed the ending comment tag, to be counted as the ending div tag, so the filter can't find the real ending tag for the div block.

Quote:<!--div class="it01img"><a href="http://sms.sina.com.cn/magazine/dy/mms.html#health" target="_blank"><span class="it01ib"><img src="http://i0.sinaimg.cn/sms/images/sinapic/27113106.jpg" width="103" height="68" /></span></a></div-->

Modifying the code as below seems solving the problem.
Code:
$INEST(<$TST(\1),((^$TST(script=1*))|$TST(script=1*)(^*</script))*,</$TST(\1) >)</$TST(\1) >

However, after searching $INEST() command in the config file I noticed sidki using the original style a lot so I am not sure what's the difference to ignore the <space>>.

Original discussion and example: http://forum.proxomitron.cn/viewtopic.php?f=3&t=173
Add Thank You Quote this message in a reply
May. 21, 2009, 12:45 PM (This post was last modified: May. 21, 2009 12:46 PM by lnminente.)
Post: #67
RE: sidki-config: Feb 13 2009
I have seen div elements being closed with "/>" and with </div> certainly its nesting is a thing i have very pendant for my adcontainers toggler. A example of a hard to match is the messenger div in yahoo mail classic without ajax, code attached.


Attached File(s)
.txt  div messenger yahoo.txt (Size: 1.26 KB / Downloads: 717)
Add Thank You Quote this message in a reply
May. 21, 2009, 02:39 PM (This post was last modified: May. 21, 2009 02:52 PM by sidki3003.)
Post: #68
RE: sidki-config: Feb 13 2009
whenever, i see the problem, thanks for tracking it down. Smile!

I think it's safe - and a good idea - to modify the "Specific Containers" filter as suggested. In the sina.com.cn example, however, it doesn't do what you may expect. The targeted DIV block is 8624 bytes large. The filter's byte limit is 5000 (for good reasons). Consequently it's falling back to just hide the block.

And yep, <sometag someattribute[^>]+>$INEST(<sometag,</sometag)</sometag > is my standard way to match specific blocks. For the generic filters this usually means to fall back to Plan B (hide) if encountering </sometag-->.
I have the feeling that i'm opening a can of worms if i change that.

As for the whitepace in the closing tag match, some blocks close like: </td
>
(Mostly for browser compatibility reasons.)


lnminente, in XHTML many tags are supposed to self-close (/>).

In fact, the browser is correcting each empty element like <span></span> to <span/> for its internal representation of the page (a.k.a. the DOM).

However, as long as these documents come with a text/html content-type (except for Ajax requested XML snippets), they are never rendered as XHTML, but as HTML 4.01. That means, the browser is correcting e.g. each <meta /> with <meta>, and is considering e.g. <script src="foo" /> as an open script tag, all following page content till the next </script> being script content.

With "the browser" i mean Firefox here. Smile! IE (6/7/8) doesn't know application/xhtml+xml. Opera is apparently treating XHTML docs coming as text/html partly as XHTML. I don't know how big that part is, however.
Add Thank You Quote this message in a reply
May. 21, 2009, 03:04 PM (This post was last modified: May. 21, 2009 03:09 PM by lnminente.)
Post: #69
RE: sidki-config: Feb 13 2009
Oh i see, an old mistake from me. I copied here the source code after being processed by "The Browser" (hehe) but the code parsed by proxo was another (attached now). Problem solved and removing some WIP tags in some filters Smile! Thanks Sikdi!


Attached File(s)
.txt  original code for the messenger.txt (Size: 397 bytes / Downloads: 723)
Add Thank You Quote this message in a reply
May. 21, 2009, 06:07 PM
Post: #70
Onload Button Appearing in Winamp Media Library
Sidki, thanks for updating your config!

In Winamp 5.552, I now see the red "onload" button in the main pane of the Media Library, when selecting SHOUTcast Radio or SHOUTcast TV in the navigation pane.

In the previous config, this did not happen. In my IncludExclude-U.ptxt file I have the following code, but the "onload" button still appears:

Code:
## allow scripts (if default is "block")    $SET(0=i_script:0.)
## ----------------------------------------------------------------------------
# Winamp
# client.winamp.com/om/sc/             $SET(0=i_level:1.i_script:0.)
# (www|client).winamp.com/(bin|om)/         $SET(0=i_level:1.i_script:0.)
# www.shoutcast.com/scradioinwinamp/        $SET(0=i_level:1.i_script:0.)
# www.shoutcast.com/sctvinwinamp/        $SET(0=i_level:1.i_script:0.)
# www.shoutcast.com/scradioinwinamp/        $SET(0=a_event.i_level:1.i_script:0.)
# www.shoutcast.com/sctvinwinamp/        $SET(0=a_event.i_level:1.i_script:0.)
# ((www|client|media|services).|)winamp(.orb|).com/    $SET(0=i_level:1.i_script:0.)
# www.shoutcast.com/                $SET(0=i_level:1.i_script:0.)

Could there be corrections for this problem in IncludeExclude.ptxt? I'd think most people would want the online playlists to show up without the hassle of clicking a button.
Add Thank You Quote this message in a reply
May. 21, 2009, 06:20 PM
Post: #71
RE: sidki-config: Feb 13 2009
wammie, i'd need at least one link (although one probably isn't sufficient) to a page where this happens.
If there is no address bar, you can get the URL for instance from http://local.ptron/.pinfo/urls/ .
Add Thank You Quote this message in a reply
May. 21, 2009, 06:43 PM
Post: #72
RE: sidki-config: Feb 13 2009
(May. 21, 2009 06:20 PM)sidki3003 Wrote:  wammie, i'd need at least one link (although one probably isn't sufficient) to a page where this happens.
If there is no address bar, you can get the URL for instance from http://local.ptron/.pinfo/urls/ .

SHOUTcast Radio:
http://www.shoutcast.com/scradioinwinamp/

SHOUTcast TV:
http://www.shoutcast.com/sctvinwinamp/
Add Thank You Quote this message in a reply
May. 21, 2009, 06:57 PM (This post was last modified: May. 21, 2009 06:59 PM by sidki3003.)
Post: #73
RE: sidki-config: Feb 13 2009
Try to replace below piece of code in proxjs-full.js. Afterwards make sure that your user-agent's cache is empty.

Old:
Code:
// Specific "onload" Events to Button - Paul Rupe, sidki
    if (prxO.oSet.$toButton & 4 && window.onload && !prxO.oAss.asWinTiny && prxO.oAss.asWinNormal) {
      var ldIn = /^\n*function SymOnLoad|^[^;]*((popu(p|nder)|track)\w*|print ?\()|\.location(\.href)? ?=[^=]|Math\.random|float *[^:]|expires ?=/i;
      this.ncButString += this.ncButEvent("window", "onload", /menu|<frameset/i, ldIn);
      if (prxO.oInt.in_onload)
        window.onload = document.body.onload = null;
    }

New (beware of word wrap):
Code:
// Specific "onload" Events to Button - Paul Rupe, sidki
    if (prxO.oSet.$toButton & 4 && window.onload && !prxO.oAss.asWinTiny && prxO.oAss.asWinNormal) {
      var ldEx = /menu|<frameset|\.OnMedia/i,
      ldIn = /^\n*function SymOnLoad|^[^;]*(?:popu(?:p|nder)|track)\w*|location\.href ?=(?!=)|Math\.random|float(?! *:)|expires ?=|print ?\(|\.(?:click ?\(|location ?=(?!=))/i;
      this.ncButString += this.ncButEvent("window", "onload", ldEx, ldIn);
      if (prxO.oInt.in_onload)
        window.onload = document.body.onload = null;
    }
Add Thank You Quote this message in a reply
May. 21, 2009, 07:17 PM
Post: #74
RE: sidki-config: Feb 13 2009
(May. 21, 2009 06:57 PM)sidki3003 Wrote:  Try to replace below piece of code in proxjs-full.js. Afterwards make sure that your user-agent's cache is empty.

Old:
Code:
// Specific "onload" Events to Button - Paul Rupe, sidki
    if (prxO.oSet.$toButton & 4 && window.onload && !prxO.oAss.asWinTiny && prxO.oAss.asWinNormal) {
      var ldIn = /^\n*function SymOnLoad|^[^;]*((popu(p|nder)|track)\w*|print ?\()|\.location(\.href)? ?=[^=]|Math\.random|float *[^:]|expires ?=/i;
      this.ncButString += this.ncButEvent("window", "onload", /menu|<frameset/i, ldIn);
      if (prxO.oInt.in_onload)
        window.onload = document.body.onload = null;
    }

New (beware of word wrap):
Code:
// Specific "onload" Events to Button - Paul Rupe, sidki
    if (prxO.oSet.$toButton & 4 && window.onload && !prxO.oAss.asWinTiny && prxO.oAss.asWinNormal) {
      var ldEx = /menu|<frameset|\.OnMedia/i,
      ldIn = /^\n*function SymOnLoad|^[^;]*(?:popu(?:p|nder)|track)\w*|location\.href ?=(?!=)|Math\.random|float(?! *:)|expires ?=|print ?\(|\.(?:click ?\(|location ?=(?!=))/i;
      this.ncButString += this.ncButEvent("window", "onload", ldEx, ldIn);
      if (prxO.oInt.in_onload)
        window.onload = document.body.onload = null;
    }

That did the trick! THANKS!!! I'll remove all that other code then from my IncludeExclude-U.ptxt.
Add Thank You Quote this message in a reply
May. 21, 2009, 07:26 PM
Post: #75
RE: sidki-config: Feb 13 2009
Good! I'll keep that change then.
Add Thank You Quote this message in a reply
Post Reply 


Forum Jump: