Commit Briefs
refactor proxy_read
- avoid the ugly `goto copy' to jump back to the start - avoid leaking memory
parse response MIME, check the charset and lang
If the charset is not UTF-8 nor ASCII then we don't know how to handle the body and so rejects it. An abset charset is implied to be utf8, so we're good. In the generated HTML we always use UTF8 charset, which is fine too, since it is a superset of ASCII. Attempt also at saving the lang for later reuse in the generated HTML.
fix request URL
use the matched proxy_name and append the received PATH_INFO. Still undecided wether I should sanitize things here or not.
drop useless proxy_accept wrapper
just call fcgi_accept (with args tweaked)