[Date Index]
[Thread Index]

[Date Prev] bullet [Date Next] bullet [Thread Prev] bullet [Thread Next]
bullet
Re: talk_to_html
bullet bullet
  • To: carl@radio.com (Carl Malamud)
  • Subject: Re: talk_to_html
  • From: "Curtis C. Generous" <generous@uucom.com>
  • Date: Mon, 21 Nov 1994 06:53:11 -0500 (EST)
  • Return-Receipt-To: generous@uucom.com
bullet bullet

According to Carl Malamud:
>
> Hi -
>
> Went ahead and made some changes to /web/santa/talk*html. Please
> let me know when its all right to start banging on your header,
> messages, etc....

You can go ahead. I will send you an update on the headers when I get in the office (I"m still @ home right now).

The structure has changed a little. Here is the layout of ~santa/Messages:

Archive/ Context/ Domain/ Footers/ Headers/

Archive/ is obvious.

Context/ is where the files which are included based on the reg. expression matching (which is now done in the santa.pl program btw as you wanted).

Domain/ is new. It will include files which match the domain name. Example: Here is the ls of Domain/

bman[761]% ls santa/Messages/Domain arc.nasa.gov.html edu.html mil.html au.html fr.html nasa.gov.html com.html gov.html net.html default.html it.html org.html

The santa.pl program figures out the originator's domain space, and then attempts to matchfor a valid include file. If the originator domain is host.whitehouse.gov for example, here is the search sequence:

host.whitehouse.gov.html whitehouse.gov.html gov.html, default.html.

The first check (host.whitehouse.gov) is needed for those people who send out domain based email addresses.

Headers/ contains 3 sub-directories: Headers/Santa Headers/Rudolph Headers/Elves Files within those directories are picked at random. No naming convention on the filenames, the script just picks one at random from the appropriate directory.

Footers/ contains closing statements and files are picked at random also from this directory.


> Also, what's the status on where the list of regular expressions
> is kept (e.g, in file or in .pl script?)

In santa.pl:

%regexp = ( "cindy.*crawford", "$PATH/Context/cindy_crawford.html", "cookie.*rudolph", "$PATH/Context/cookie_for_rudolph.html", "power.*ranger", "$PATH/Context/no_power_ranger_dolls.html", );


> BTW, if we match a regular expression, does that match for
> any recipient (e.g., are regular expressions for rudolph
> different than those for santa). I need to know that in
> writing my answers (e.g., if a message is to rudolph and
> we mention dissertation, wouldn't want to start the answer
> with "Rudolph says").

Yes it does, but that is easy to change if you want to go to that level of complexity (it means many more replies have to be written).


> Sorry to bug you on the weekend. :) I'm off for the rest of
> the day, but in early Monday morning. Let me know if I can
> start dripping good cheer and regular expressions into those
> files.

No problem. I was in the office all weekend long anyway.


> Carl

I need you technical advice on a couple of things:

1) Handling of Email vs HTML replies:

Several approaches: [] have seperate reply files for each [] have HTML replies, and strip tags if sent as email [] have email based files and try to add HTML tags

First is simplest. Two is also pretty straight forward but you have to be aware of formatting issues when you generate the files. Three is not worth the effort. What do you think?

Oops, got to go. I'll continue this at the office.

--curtis

bullet
[Date Prev] bullet [Date Next] bullet [Thread Prev] bullet [Thread Next]
bullet