DSM-G600, DNS-3xx and NSA-220 Hack Forum

Can you shorten it match on "tokkal"? I have no idea if it makes sense as I don't understand the language.

Daro wrote:

Hi,

i browsed through the whole thread but i didnt find solution for my problem. How can i filter this:

Code:

<title>Így jártam anyátokkal - 6x12 [HDTV-720p - Eng - CTU](stevent02)</title>

as you see i have problem with non english characters. I tried with unicode setting like \u00cd as "Í" but pcre doesnt support it.

Can someone help me how can i use the non english chars?

Do you have a public RSS feed for me that contains UTF8 characters so I can test around?

@Daro
Please try this release and tell me if it works now.
Assuming the RSS feed is in UTF8 encoding, make sure your automatic.conf is stored as UTF8 file as well, otherwise the regular expressions won't match.

@Daro

Here's another test build. I tested this with the files you provided and the regular expression matched just fine, so this should work with the feed as well.

Let me know if so, then this will be v0.6.5.

Hi,

i am here again.
The new test version worked great until now. Now it has a problem

Entity: line 130: parser error : xmlParseEntityRef: no name
<title>Law & Order: Los Angeles - 1x08 [Web-Dl-720p - Eng - POD](aljasPOD)</titl

The interesting thing is the rss doesnt contain that torrent what the parser write. I dont know where it gets this xml, and why doesnt download and use the right one. If i am trying in utorrent it works fine, but on dns always listing that law and orders episodes. Maybe this xml saved somewhere and automatic cant overwrite it?

Last edited by Daro (2011-01-21 18:35:26)

KyleK wrote:

This is a problem with your feed: As you can see, the episode name contains an unescaped ampersand, which confuses the XML parser.

Thats ok, but the feed isnt contain anymore that. Thats why i dont understand it. Its an old one.. I dont know where the automatic get it.

KyleK wrote:

Automatic doesn't store any XML, only links to the torrents it downloaded previously (in the automatic.state file).

Please check if Automatic uses the correct configuration file, and not an old one.
Also, can you send me the logfile where this error occurs?

Oops, i mislook, the webpage show only the half xml, if i check the source then there are more and the Law & order still there. So looks like the "&" sign is the problem...

The configuration file isnt changed, so i dont think so it is the problem.

I send the xml, the log and the conf file

---

Attachments:
Test.rar, Size: 11,954 bytes, Downloads: 334

I just wanted to congratulate you with the achievement of programming a brilliant piece of software like this. I'm actually going to buy another NAS and dedicate it to this.

I have one question though: uTorrent has this function for setting an episode number to download from. How can one fix this with regular expressions easily, as some torrents are named [ss]x[ee], and some S[ss]E[ee]?
Also, I would give you a hand in writing a filter/pattern-generator in PHP for some basic functions? Like: fill in series-name, episode to download from, etc, providing the correct text to paste into a conf file?

This post is a mess; encountered some problems, and in the end, found some solutions to imo very common problems

FunFiler wrote:

Just write multiple filters to handle new shows. RegEx is EXTREMELY powerful.
starting Example

"[\w|\.][0-9][0-9]x[0-9][0-9][\w|\.]"
and
"[\w|\.]s[0-9][0-9]e[0-9][0-9][\w|\.]"

That's quite a complex world, but I'll learn it someday, I guess

Is there an easy way to test if my filters actually work?

KyleK wrote:

You can test your regular expressions here: http://www.regular-expressions.info/jav … ample.html

"Download all torrent files with 'Chuck' AND '720p' in any order, somewhere in the filename, but do not download files with 'season' in them".

(?=.*720p)(?!.*season)Chuck

At least, I had to check "Chuck"; "chuck" did not work. I guess the "not case sensitive" feature is achieved by lower-casing all found .torrent files in the first place?

Last edited by wtfok (2011-02-01 11:54:36)

FunFiler wrote:

Even easier:

filter = { pattern => "(?!.*(720p|season))^Chuck\." }

You can build on the exclusion list by using |
I use ^ to indicate "starting with" to avoid it being elsewhere in the string
I also use \. at the end to avoid "Chucking  ..." or some other variation.

Thanks for the help, you really have a point there for the exclusion list. But doesn't | stand for "or"?
Anyhow, I don't think ^ is such a good idea, because some torrents do have a prefix or whatever, and then it would not work. Also, the first character after "Chuck" isn't always a dot, so I don't think that's a good idea either. Maybe \s is an option there? Read something about that in this thread...

Edit: did test "(?!.*(720p|season))^Chuck\.", wouldn't work. What does work, though, is "(?!.*(720p+season))^Chuck\.". My conclusion is that | is indeed OR and + is AND.

Building on that thought, I think this one is quite optimal: (?!.*(720p+season))^Chuck\W. It matches everything except text after the word "Chuck". I'm now finding out how to check everything before that word: nothing OR everything except an alphabetical character.

Last edited by wtfok (2011-02-01 15:37:42)

In my case, I want the "|" which is indeed OR, but use whatever works for you.

I suspect the errors you see indicate a DNS server configuration error.

Automatic parses the XML Document returned from the feed URL, so no, it doesn't start with "http://....". The link to the actual torrent will, but that isn't parsed. As I've mentioned, the ones I've posted above are ones that I actually use and have been using since the day I came across Automatic. If you don't want to limit with "^", then take it out

I download the non 720p versions to save space. The example for Chuck is
filter = { pattern => "(?!.*(720p|micro|repack))^Chuck\." }

If you want only the 720p version, then you can modify and try
filter = { pattern => "(?!.*(season|micro|repack))^Chuck.*720p" }

This excludes anything with "season", "micro" and "repack" (I don't personally want these last 2, you can remove if you want)

As for your other issue, "Couldn't resolve host name" tells me you have a DNS issue unless of course that host does not exist. It could also be a gateway, firewall, ISP issue or the server was down or too busy. Looking at the URLs you are using, are you authenticating? Try a simple (different) torrent site first. You can also increase the level of detail that gets logged.

Last edited by FunFiler (2011-02-05 15:17:22)

Bump up your log level "-v 3". If it still doesn't work Monday, post your config and log files.

Are you using a URL that requires authentication?