>
> >>>From: Peter Rauch <[log in to unmask]>
> >>>But, I think the question was, how does one "standardize" the names
> >>>(of the subscriptions) so that the punctuation in the "From:" lines
> >>>is constant.
>
> There have been two questions, one about how posts look and the
> other about maintaining the list (e.g. how the GET or REVIEW looks).
> I was addressing the contents of the GET file, not how a post shows
> up to the subscribers.
>
> I think the original question was by a list maintainer, who wanted
> to sort the list subscribers. This would NOT be done on posts, but
> might be done on GET or REVIEW output.
 
I was the one who posted the original question.  Let me restate the
question as best as I can without reposting the original:  I administer
the list ANSAX-L.  Since 1991 I have also been maintaining a full text
database containing an archive of discussions on the list.  (There are
compelling reasons why we elected to go this route, rather than use the
notebook archive function).  The original database was in SPIRES and
modeled on the one developed for HUMANIST.  One of the problems I faced
was in the "From: " line in the mail headers: the SPIRES loader
uses this line to extract the name of the message sender for indexing
(it also extracts the date and time, and the subject of the message
and indexes them as well as key words in the text).  In order to make
this indexing useful, the format of the name and addess need to be the
same in all cases:
 
   From:  Firstname Familyname  <[log in to unmask]>
 
But in fact, in order to meet RFC822 standards, whenever there is
anything other than plain ASCII letters in the name combination, a mail
system will put double quotes around it.  Moreover, other problems
occur, as well.  For example, at least half a dozen subscribers from AOL
have @aol.com appended to the family name.  Some names are in just caps,
and at least in the earlier portions of the database (perhaps pre
RFC822) the order can be reversed:
 
   From:  <email@place>  (givname famname)
 
This combination seems not to occur any longer.  And case is immaterial,
since SPIRES converts everything to upper case anyway, though it outputs
a search in the original case.
 
I am beginning to see that LISTSERV is probably not the tool to
"standardize" this so that my SPIRES indexing will always produce the
same results.  But I'm grateful for all the views that have been
expressed.  And if I am wrong, and LISTERV CAN do this I'd like to know
how.
 
Bill
[Admin/owner or co-owner of:  ANSAX-L, SCRIPTORIUM, INTERSCRIPTA]
 
 
--
Dr. W. Schipper                     Email: [log in to unmask]
Department of English,              Tel: 709-737-4406
Memorial University                 Fax: 709-737-4528
St John's, Nfld. A1C 5S7