[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: abbreviate "stream"?



On Monday, March 17, 2003 10:16 PM, Stephen McCracken [SMTP:samccpa@xxxxxxxxx] wrote:
> The function names in this srfi, while descriptive,
> tend to be pretty long.  I would like to discuss
> possible abbreviations for "stream".  
> 
> Personally, I like "$" as an abbreviation.  It
> suggests "stream" because it looks like an 'S'. 
> Furthermore, it suggests laziness, as '$' is lazy
> application in haskell, and Okasaki uses '$' as
> notation for suspensions in his _Purely Functional
> Data Structures_.  Unfortunately, '$' cannot begin an
> identifier in R5RS.  Here are some legal variations:
> 
> Instead of "stream-map",
> 
>    _$-map
>    S$-map
>    map-$
>    L$-map   (suggests "lazy list")
>    C$-map   (suggests "lazy cons")
>    S-map
>    Phil-map 
>    
> 
> 
> __________________________________________________
> Do you Yahoo!?
> Yahoo! Platinum - Watch CBS' NCAA March Madness, live on your desktop!
> http://platinum.yahoo.com

I'm not sure that R5RS disallows identifiers whose first
character is $.  According to section 7.1.1:

    <identifier> -> <initial> <subsequent>* | <peculiar identifier>
    <initial> -> <letter> | <special initial>
    <special initial> -> a whole bunch of characters including $

I will state that I have always considered non-letter initial
characters to be in bad taste, except for conventions like
*global-name* or :keyword for DSSSL.

R5RS uses stream and map-streams in the Example.  R5RS
also uses the names vector and string to name functions of
those data types, and they are equally as long as stream.  In
the reference implementation, I frequently use strm as an
abbreviation for stream, but not in the names of functions,
only in argument lists.

Since stream-map returns a newly-allocated stream, would
phil-map have to return a newly-allocated phil?  I really don't
want to think about that.

I like stream.  What do others think?

Phil