egghelp.org community

[D3FiANC3]

Putting aside the fact that I am a huge Tcl nooblet, where have I gone wrong here? It never binds.

[arfer]

I think if you explain what you are trying to do, it might help us to assist.

From what I can gather, if MrBot (not clear whether that is the botnick on which this script is loaded) says in #mychan something that matches *released*ago* then you attempt to strip text codes with a regsub (I can't see that working) then replace ) ( and " " with %29, %28 and + respectively to form a valid URL from the text. Then append the resultant text to http://www.myurl.com/gotime.php?prerls=

First off, use the Eggdrop Tcl command stripcodes to strip text formatting codes as follows :-

set text [stripcodes bcruag $text]

The channel text which you want the script to respond to is typically :-

<MrBot> (BETA) Beta.v0.27 was released 20h 3m 22s ago

I deduced from this that the URL would be :-

http://www.myurl.com/gotime.php?prerls=%28BETA%29+Beta.v0.27+was+released+20h+3m+22s+ago

I pasted that into a browser and got nothing useful that might tell me what you are trying to do.

Quite aside from the success or otherwise of the coding inside the PUBM proc, if you wish to test if 'things' are happening you could temporarily add output statements at appropriate points. For example, if you want to see if the bind is triggering simply add this as the first line of code inside the proc (and watch the partyline for output) :-

putlog "bind triggered"

You can use this technique to find where your code fails, if there is not otherwise any Tcl error to be seen in the partyline.

A typical http useragent (from an MS Windows machine) would be :-

"Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0)"

[speechles]

Your problem is likely case. You've bound to lowercase words and if any uppercase letters happen to be part of that it will not trigger. Try using something similar to how I've done it below. Let the procedure do the figuring with a series of string equal/matches. I've also correctly added the formatQuery for your text instead of your regsubbing () into html values. User agent shouldnt matter at all as this is your site, you can omit it if you want.

[arfer]

LOL!! OK, don't rub that one in. I never even saw that coming

[nml375]

@speechles, masks of bindings do not care for case.
However, the manual check of the nickname is indeed case sensitive.

As for sanitizing the url for http:get, just use the ::http::formatQuery.

[speechles]

[nml375]

For some very odd reason, I tend to not keep track of the warez-scene :p
In any case, not having a -command (which I assume you were thinking of with -callback) argument means it is running in synchronous mode. That is, the ::http::geturl command will block until the whole webpage is loaded, or an error/abort condition occurs. The content of the retrieved data is then directly returned, rather than returning a http transaction token.

Whenever using the http package with eggdrops, use asynchronous mode, regardless of the script's use.
(No, I did not correct this in my earlier post, I was a lazy boy)
_________________
NML_375, idling at #eggdrop@IrcNET

[D3FiANC3]

Thanks for all the replies. I also appreciate the explanations to go with the code - always better to learn so next time I can do for myself. I have fixed my errors and thrown the log line in to test it out.

The reason why my script sends the data to a php page is as you guessed, to add it to the DB. As for why it is sent unprocessed ... well that is simply because my PHP knowledge is advanced whereas my Tcl knowledge (as you can see) is amateurish.

Edit:

OK, here is the amended script.