[reSIProcate] [resiprocate] New headers and parameters from RFC 3455 and UUI drafts (#6)
Scott Godin
sgodin at sipspectrum.com
Mon Oct 6 09:56:12 CDT 2014
> Do you think that it could also be necessary / useful having a class to
handle tokens without a value like P-Charging-Vector?
Here is the grammar for P-Charging-Vector:
P-Charging-Vector = "P-Charging-Vector" HCOLON icid-value
*(SEMI charge-params)
charge-params = icid-gen-addr / orig-ioi /
term-ioi / generic-param
icid-value = "icid-value" EQUAL gen-value
icid-gen-addr = "icid-generated-at" EQUAL host
orig-ioi = "orig-ioi" EQUAL gen-value
term-ioi = "term-ioi" EQUAL gen-value
Looks to me like icid-value=<?> is actually the token value and not a
parameter (even though it looks like a parameter) - also it is not
optional like other parameters. I think John Gregg's treatment of
this header makes sense. Defining a new header category is possible,
but seems overkill to me.
Scott
On Sat, Oct 4, 2014 at 7:05 AM, Dario Bozzali <Dario.Bozzali at ifmgroup.it>
wrote:
> Hi Scott,
>
> I will try to create the new class next week.
>
> I think that I could take the same approach used for QuotedDataParameter,
> deriving class from Token and adding a data member mQuoted that is set
> while parsing and it is used to encode stream.
>
> Do you think that it could also be necessary / useful having a class to
> handle tokens without a value like P-Charging-Vector?
>
> Maybe I could try to write also a class like TokenWithoutValue or
> ParametersToken (any suggestions for the name would be appreciated): value
> might contain the complete string with all the parameters, but in first
> position we expect a parameter (param=value;...), so when we stream the
> object we don't put ';' at the start of the string.
>
> Best regards,
>
> Dario
>
>
> ------------------------------
> *Da:* slgodin at gmail.com [slgodin at gmail.com] per conto di Scott Godin [
> sgodin at sipspectrum.com]
> *Inviato:* venerdì 3 ottobre 2014 17.50
> *A:* Dario Bozzali
> *Cc:* resiprocate-devel at resiprocate.org
> *Oggetto:* Re: [reSIProcate] [resiprocate] New headers and parameters
> from RFC 3455 and UUI drafts (#6)
>
> Sorry for my mistake in the last email - QuotedDataParameter is for
> parameters not headers. You could use the StringCategory type, but then
> you won't get any special stack handling of header parameters.
>
> I think the proper solution is to create a custom header parser class
> (like WarningCategory) that will handle a token or quoted string and the
> parameters that go with it (named something like:
> TokenOrQuotedStringCategory).
>
> Scott
>
>
> On Fri, Oct 3, 2014 at 6:28 AM, Dario Bozzali <Dario.Bozzali at ifmgroup.it>
> wrote:
>
>> Hello Scott,
>>
>> In draft-ietf-cuss-sip-uui-17 draft (section 4.1) the value of
>> User-to-User header is defined as a token or a quoted string.
>>
>>
>>
>> UUI = "User-to-User" HCOLON uui-value *(COMMA uui-value)
>>
>> uui-value = uui-data *(SEMI uui-param)
>>
>> uui-data = token / quoted-string
>>
>> uui-param = pkg-param / cont-param / enc-param / generic-param
>>
>> pkg-param = "purpose" EQUAL pkg-param-value
>>
>> pkg-param-value = token
>>
>> cont-param = "content" EQUAL cont-param-value
>>
>> cont-param-value = token
>>
>> enc-param = "encoding" EQUAL enc-param-value
>>
>> enc-param-value = token / "hex"
>>
>>
>>
>> Maybe Token is a proper type for UUI value, but like in the case of
>> P-Charging-Vector header we must handle “special cases”. Something similar
>> to the following code:
>>
>>
>>
>> Data uuiData(Data::from(msg.header(h_UserToUser).c_str()));
>>
>> ParseBuffer uuiBuffer(uuiData);
>>
>> if (!uuiBuffer.eof()) {
>>
>> try {
>>
>> // extract the UUI value
>>
>> const char* uuiStart = uuiBuffer.skipWhitespace();
>>
>> const char* uuiEnd = NULL;
>>
>> if (*uuiBuffer.position() == Symbols::DOUBLE_QUOTE[0]) {
>>
>> uuiBuffer.skipChar();
>>
>> uuiStart = uuiBuffer.position();
>>
>> uuiEnd = uuiBuffer.skipToEndQuote();
>>
>> }
>>
>> else {
>>
>> uuiEnd = uuiBuffer.skipToChar(';');
>>
>> }
>>
>> if ((int)(uuiEnd - uuiStart) >= 0) {
>>
>> Data uuiDataValue;
>>
>> uuiBuffer.data(uuiDataValue, uuiStart);
>>
>> // use uuiDataValue
>>
>> }
>>
>> }
>>
>> catch (resip::BaseException& ex) {
>>
>> // handle parsing error, if any
>>
>> }
>>
>> }
>>
>>
>>
>> What is your opinion? Can we continue to use Token type?
>>
>>
>>
>> Best regards,
>>
>> Dario.
>>
>>
>>
>> *From:* slgodin at gmail.com [mailto:slgodin at gmail.com] *On Behalf Of *Scott
>> Godin
>> *Sent:* giovedì 2 ottobre 2014 21.44
>> *To:* Dario Bozzali
>> *Cc:* resiprocate-devel at resiprocate.org
>> *Subject:* Re: [reSIProcate] [resiprocate] New headers and parameters
>> from RFC 3455 and UUI drafts (#6)
>>
>>
>>
>> >Scenario 1 related to User-to-User header.
>> >How to parse the following header?
>> >User-to-User: "the quick silver fox jumped over the lazy brown
>> dog";encoding=hex
>> >When I try to get the value of the header I obtain the string ""the"
>> because Token::parse() method skip till first whitespace or semicolon.
>>
>>
>>
>> Perhaps the header type you used is wrong then. If the Data is always
>> quoted you should use QuotedDataParameter instead of Token.
>>
>>
>>
>> Scott
>>
>>
>>
>> On Thu, Oct 2, 2014 at 9:43 AM, Dario Bozzali <Dario.Bozzali at ifmgroup.it>
>> wrote:
>>
>> Hi all,
>> I'm executing some additional tests on new headers. I have a couple of
>> doubts, but I don't know if there is a bug in the pull request or in my
>> test scenarios.
>>
>> Scenario 1 related to User-to-User header.
>> How to parse the following header?
>> User-to-User: "the quick silver fox jumped over the lazy brown
>> dog";encoding=hex
>> When I try to get the value of the header I obtain the string ""the"
>> because Token::parse() method skip till first whitespace or semicolon.
>>
>> Scenario 2 related to P-Charging-Vector header.
>> P-Charging-Value has icid-value mandatory parameter and other optional
>> parameters. I understood that it should be modelled as a Token with empty
>> value, but in this way one ";" is inserted before icid-value.
>> Example:
>> P-Charging-Vector:
>> ;icid-value=B6012049-0080-42D3-982E-473B24816316;icid-generated-at=192.168.56.1;orig-ioi=myserver
>>
>> What is the right way to set header value?
>> Should I use a icid-value parameter to set P-Charging-Vector value?
>>
>> Something like this:
>>
>> Token chargingVectorToken;
>> DataParameter icidValueParam(ParameterTypes::icidValue);
>> icidValueParam.value() =
>> Data("B6012049-0080-42D3-982E-473B24816316");
>> Data icidData;
>> DataStream icidDataStream(icidData);
>> icidValueParam.encode(icidDataStream);
>> chargingVectorToken.value() = icidData;
>> chargingVectorToken.param(p_icidGeneratedAt) =
>> Data("192.168.56.1");
>> chargingVectorToken.param(p_origIoi) = Data("myserver");
>>
>> Maybe John Gregg could help me on this.
>>
>> Thank you in advance and kind regards.
>>
>> Dario
>>
>>
>> -----Original Message-----
>> From: resiprocate-devel [mailto:resiprocate-devel-bounces at resiprocate.org]
>> On Behalf Of Dario Bozzali
>> Sent: martedì 23 settembre 2014 18.34
>> To: 'Daniel Pocock'
>> Cc: resiprocate-devel at resiprocate.org
>> Subject: Re: [reSIProcate] [resiprocate] New headers and parameters from
>> RFC 3455 and UUI drafts (#6)
>>
>> Hi Daniel,
>> Excluding DayOfWeekHash and MonthHash from build I was able to compile
>> without changing DateCategory.cxx.
>> DayOfWeekHash and MonthHash are built because tool configured in project
>> is "C/C++ Compiler Tool" instead of "Custom Build Tool" as per other header
>> files.
>>
>> Regards,
>> Dario.
>>
>> -----Original Message-----
>> From: Daniel Pocock [mailto:daniel at pocock.pro]
>> Sent: martedì 23 settembre 2014 18.24
>> To: Dario Bozzali
>> Cc: resiprocate-devel at resiprocate.org
>> Subject: Re: [resiprocate] New headers and parameters from RFC 3455 and
>> UUI drafts (#6)
>>
>>
>> Hi Dario,
>>
>> I notice that MethodHash is including string.h and ctype.h so I've added
>> those same includes to DateCategory.cxx so they will be included before
>> DayOfWeekHash and MonthHash
>>
>> Can you just confirm, Visual Studio is not trying to compile these cxx
>> files into separate objects? It should only be including them when it
>> compiles DateCategory.cxx so they should all become part of the object file
>> DateCategory.o
>>
>> Regards,
>>
>> Daniel
>>
>> On 23/09/14 18:08, Dario Bozzali wrote:
>> > Hi Daniel,
>> > Below I reported the output obtained using Visual Studio 2005 to build
>> master.
>> > Dario.
>> >
>> > 1>DayOfWeekHash.cxx
>> > 1>dayofweek.gperf(5) : error C2146: syntax error : missing ';' before
>> identifier 'type'
>> > 1>dayofweek.gperf(5) : error C4430: missing type specifier - int
>> > 1>assumed. Note: C++ does not support default-int
>> > 1>dayofweek.gperf(5) : error C4430: missing type specifier - int
>> > 1>assumed. Note: C++ does not support default-int
>> > 1>dayofweek.gperf(13) : error C2065: 'Sat' : undeclared identifier
>> > 1>dayofweek.gperf(13) : error C2078: too many initializers
>> > 1>dayofweek.gperf(8) : error C2065: 'Mon' : undeclared identifier
>> > 1>dayofweek.gperf(7) : error C2065: 'Sun' : undeclared identifier
>> > 1>dayofweek.gperf(10) : error C2065: 'Wed' : undeclared identifier
>> > 1>dayofweek.gperf(11) : error C2065: 'Thu' : undeclared identifier
>> > 1>dayofweek.gperf(12) : error C2065: 'Fri' : undeclared identifier
>> > 1>dayofweek.gperf(9) : error C2065: 'Tue' : undeclared identifier
>> > 1>dayofweek.gperf(20) : error C3861: 'strncmp': identifier not found
>> > 1>MonthHash.cxx
>> > 1>month.gperf(5) : error C2146: syntax error : missing ';' before
>> identifier 'type'
>> > 1>month.gperf(5) : error C4430: missing type specifier - int assumed.
>> > 1>Note: C++ does not support default-int
>> > 1>month.gperf(5) : error C4430: missing type specifier - int assumed.
>> > 1>Note: C++ does not support default-int
>> > 1>month.gperf(12) : error C2065: 'Jun' : undeclared identifier
>> > 1>month.gperf(12) : error C2078: too many initializers
>> > 1>month.gperf(13) : error C2065: 'Jul' : undeclared identifier
>> > 1>month.gperf(7) : error C2065: 'Jan' : undeclared identifier
>> > 1>month.gperf(11) : error C2065: 'May' : undeclared identifier
>> > 1>month.gperf(8) : error C2065: 'Feb' : undeclared identifier
>> > 1>month.gperf(9) : error C2065: 'Mar' : undeclared identifier
>> > 1>month.gperf(10) : error C2065: 'Apr' : undeclared identifier
>> > 1>month.gperf(18) : error C2065: 'Dec' : undeclared identifier
>> > 1>month.gperf(14) : error C2065: 'Aug' : undeclared identifier
>> > 1>month.gperf(15) : error C2065: 'Sep' : undeclared identifier
>> > 1>month.gperf(16) : error C2065: 'Oct' : undeclared identifier
>> > 1>month.gperf(17) : error C2065: 'Nov' : undeclared identifier
>> > 1>month.gperf(28) : error C3861: 'strncmp': identifier not found
>> >
>> > -----Original Message-----
>> > From: Daniel Pocock [mailto:daniel at pocock.pro]
>> > Sent: martedì 23 settembre 2014 17.48
>> > To: Dario Bozzali
>> > Cc: resiprocate-devel at resiprocate.org
>> > Subject: Re: [resiprocate] New headers and parameters from RFC 3455
>> > and UUI drafts (#6)
>> >
>> >
>> >
>> > On 23/09/14 17:42, Dario Bozzali wrote:
>> >> Hi Daniel,
>> >> I solved the compiling issue changing Visual Studio project settings
>> to exclude DayOfWeekHash.cxx and MonthHash.cxx from build. I commit this
>> change for every VS version.
>> >> Now I will take a look at conflict issue.
>> >
>> >
>> > That sounds odd. If you checkout the master branch, without any of
>> your changes, does it build with those files included?
>> >
>> > If it doesn't build on master, can you tell me what error it gives?
>> >
>>
>> _______________________________________________
>> resiprocate-devel mailing list
>> resiprocate-devel at resiprocate.org
>> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>>
>> _______________________________________________
>> resiprocate-devel mailing list
>> resiprocate-devel at resiprocate.org
>> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.resiprocate.org/pipermail/resiprocate-devel/attachments/20141006/58c7b9d5/attachment.htm>
More information about the resiprocate-devel
mailing list