> It may be possible to write some simple Scheme-implementation-specific
> rules (filename/grep patterns) to produce a of SRFIs supported by that
> implementation, given a tarfile of the implementation's source release.
> I've generally found that automating data extraction is not much more
> work than doing things by hand, and is much less error-prone. It also
> leaves instructions that others can study and run for how you managed to
> find the info.

As a quick example:

curl --fail --silent --show-error
http://synthcode.com/scheme/chibi/chibi-scheme-0.8.0.tgz | gunzip | tar
-tf - | grep -i 'chibi-scheme-[0-9.]*/lib/srfi/[0-9]*.sld' | sed -e
'xxxxxx@.*/@@' -e xxxxxx@xxxxxx@@ | sort -g