Hi,RFC 3261 BNF section defines Request-URI as follows:
Request-URI = SIP-URI / SIPS-URI / absoluteURI
While it's obvious doing a complete parsing of SIP and SIPS URI, I wonder how
useful is parsing sctrictly an absolute URI. Note that absoluteURI contains a
very complex grammar:
absoluteURI = scheme ":" ( hier-part / opaque-part )
hier-part = ( net-path / abs-path ) [ "?" query ]
net-path = "//" authority [ abs-path ]
abs-path = "/" path-segments
opaque-part = uric-no-slash *uric
uric = reserved / unreserved / escaped
uric-no-slash = unreserved / escaped / ";" / "?" / ":" / "@"
/ "&" / "=" / "+" / "$" / ","
path-segments = segment *( "/" segment )
segment = *pchar *( ";" param )
param = *pchar
pchar = unreserved / escaped /
":" / "@" / "&" / "=" / "+" / "$" / ","
scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
authority = srvr / reg-name
srvr = [ [ userinfo "@" ] hostport ]
reg-name = 1*( unreserved / escaped / "$" / ","
/ ";" / ":" / "@" / "&" / "=" / "+" )
query = *uric
IMHO for a SIP server parsing an absoluteURI is not useful at all (I don't
consider, of course, the case of a server implementing SIP and HTTP and/or
other protocols).
So I wonder if would be possible to relax the absoluteURI grammar making
it "faster" to parse, something like (not pure BNF):
absoluteURI = scheme ":" 1*( any_char - ( "<" | ">" ) )
scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
In this way, the SIP parser would detect the absoluteURI scheme and pass it to
a higher application logic handling protocols more than SIP and SIPS.
Do you consider it correct? any dissadvantage? (Please, consider common cases
implementing a normal SIP server).
Thanks a lot.
PD: I would also consider "tel" URI:
Request-URI = SIP-URI / SIPS-URI / TEL-URI / absoluteURI
--
Iñaki Baz Castillo
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
