Zugg Software :: View topic

Newbie Joined: 23 Oct 2007 Posts: 4

Hi, I have been trying to capture a multi line output of a "lore scroll" used on my mud. However, because of the unusual formatting of the output, I am having setting up the script properly. The result is that records in the database suddenly change to 0 and the data has to be re-entered manually. An example would be as follows:

Object 'robe blood stained robes'
Item Type: WORN
Mat Class: cloth Material: wool
Weight : 2 Value : 300

Affects : SAVING_SPELL by -1
Affects : STR by -1
Item is : MAGIC ANTI_GOOD

There are inherently several problems capturing this output. Firstly there is a break in the middle of the information causing my {$} trigger to fire prematurely. Secondly, Affects and by are listed twice so I am unsure on how to parse the data into thier own fields. The current script I am attempting to use is the following:


#ALIAS identify {
#t+ identify
recite lore %1
}



#CLASS {identify}



#TRIGGER Object '(&item.name)' {}
#TRIGGER Item Type: (&item.type) {}
#TRIGGER Mat Class: (&item.matclass) Material: (&item.material) {} <------Two values on same line
#TRIGGER Weight : (&item.weight) Value : (&item.value) {}
#TRIGGER Affects : (&item.affects) by (&item.by) {}
#TRIGGER Item is : (&item.is) {}
#TRIGGER Restricts: (&item.restricts) {}
#TRIGGER Apply : (&item.apply) {}



#TRIGGER {$} {
#if (%find( @item.name) = "") {#new "" @item}
#t- identify
}
#CLASS 0


There is also an instance where multiple peices of data are on the same line. I tried putting both values on the same line seperating them with a # but that had little effect. I was also thinking of trying to define them as variables. i.e. %1 and %2 but I thought it best to check the forums first to gain input from more experienced scripters. Thank you for any suggestions and/or corrections that you may have.

Posted: Tue Oct 23, 2007 11:40 am

The %nn variables probably won't help you here. I personally never use the &variable syntax - I like using specific wildcards - but this is one situation where that won't help.

Firstly, the newline problem, because that's simple enough. Use a multistate trigger:

#TRIGGER {$} {}
#COND {$} {
#if (%find( @item.name) = "") {#new "" @item}
#t- identify
}

This'll only fire after you've received two blank lines instead of one.

Your main problem obviously stems from the fact that your first affect is getting overwritten by the second. If you always have two affects and two affects only and the lines always follow each other, just create another field for the second affect and use another multistate trigger and another database field:

#TRIGGER {Affects : (&item.affects) by (&item.by)} {}
#COND {Affects : (&item.affects2) by (&item.by2)} {} {within|param=1}

If your items can have a variable number of affects, it's going to be more complex.

Newbie Joined: 23 Oct 2007 Posts: 4

These are excellent answers and I think they will work well. I am still a little confused about two items on the same line though. For example:

#TRIGGER Mat Class: (&item.matclass) Material: (&item.material) {}
---------------^-------------------------------^-----------------------------------

What my crudely drawn arrows are trying to show is that sometimes I have to capture two entries from a line. Is there a delimiter necessary for CMUD to understand that two different entries on a single line go to two independent fields? Thus far, my trigger has been capturing both entries and dupkicating them in the fields

matclass metal gold
material metal gold

instead of

matclass metal
material gold

Posted: Tue Oct 23, 2007 5:59 pm

There's not much I can say except that it shouldn't be doing that. Perhaps it's a new bug?

You could try something like

#trig {Mat Class: (%w) Material: (%w)} {item.matclass = %1;item.material - %2}

instead.

Newbie Joined: 23 Oct 2007 Posts: 4

I believe you are right. It is most likely due to syntax which I will review line by line. CMUD has ran pretty well on Vista so far. I am really impressed with how many features it has and the CMUD forum moderators are very smart and helpful as well. Thank you for all of your help I will repost if I find anything new. Mr. Green

Adept Joined: 21 Sep 2005 Posts: 250 Location: Austin, TX

I have similar output from the 'identify <object>' command on my MUD. It doesn't have the issue of internal blank lines, so my script terminates capture on a blank line or a prompt (line starting with '<'). The adaptations shown above would work fine.

As for the multiple AFFECTs, try storing them as a db variable, as shown below. Works for any number of lines. (Actually, my script stores SOME properties in specific named variables, and stores anything else it finds into a catchall db of name-value pairs).

Sample output from my MUD:

Newbie Joined: 23 Oct 2007 Posts: 4

You know that long a** code was kinda what I was looking for. It poses as a great reference to study and adapt. The way the code is set up now it seems to capture some of the information and then send it to a variable instead of to the database itself. I am not entirely sure why it does this so I am still playing with the script. It might be nice if you could post your script after you convert it over to cmud. Maybe it will help give more ideas on top of the already great ones we had so far :-)

Adept Joined: 21 Sep 2005 Posts: 250 Location: Austin, TX