Am Sa., 6. März 2021 um 17:03 Uhr schrieb Jakub T. Jankiewicz <xxxxxx@onet.pl>:

On Sat, 6 Mar 2021 15:43:01 +0100
Marc Nieper-Wißkirchen <xxxxxx@nieper-wisskirchen.de> wrote:

> > We can simplify the SRFI by requiring for implementation to work in
> > different
> > files. This is how it was working before I replaced tokenizer in my Scheme
> > that return list of all tokens for whole file with incremental Lexer that
> > is
> > tokenizing while the code is parsed.
> >
>
> How would this look like in a typical R7RS Scheme program consisting of a
> top-level program and a number of libraries?
>
> What happens if two libraries try to declare the same prefix independently
> as special?

The specials are global because the parser is single entity. But what about
SRFI-10 if two libraries define same name (e.g. #,(foo ...)) and they use
different implementations. From what I see SRFI-10 don't use any scope and
don't handle the R7RS libraries. The latest library will overwrite the code
that was defined in libraries that were loaded before it if both defined the
name.

Yes, any SRFIs that define reader extensions usually define them globally. For (quasi) standards, this is less of a problem than for user-written extensions which need to co-exist with other user-written extensions.

For example, assume that I want to use a library (foo). The implementor of (foo) uses some other library (bar), which makes use of your proposed reader extensions, internally. If the reader extensions are only global, (foo) will leak implementation details.

SRFI 10, by the way, is both pre-R6RS and pre-R7RS so it is clear that it doesn't handle libraries. Moreover, I would consider the special syntax used there and the SRFI itself obsolete because since R6RS, `#,` is an abbreviation for `unsyntax`.