March 2013 - dfdl-wg - lists.ogf.org

whitespace in DFDL annotations: right now regex is xs:string, expression is xs:token
by Mike Beckerle 20 Mar '13

20 Mar '13

This came up on the call. The schemas I have for DFDL annotations have DFDLRegularExpression as an xs:string, and DFDLExpression as an xs:token. I have no clue what a union of these types behaves like. But we have a union called DFDLExpressionOrPatternOrNothing which is a 3-way union of DFDLExpression, DFDLRegularExpression, and EmptyString (which is also derived from xs:string but has length facet of 0 as well. -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com

3 2

Re: [DFDL-WG] Action 206 - proposed timezone-from-dateTime function - update
by Steve Hanson 19 Mar '13

19 Mar '13

Also noticed that in section 23.5.2.4 we do not state what the type of the return value is from the various date/time functions. They are all xs:integer except for the 'seconds' functions which return xs:decimal as they include any fractional seconds. Suggest a sweep through all the other functions to check the return type is either stated or obvious. Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 From: Steve Hanson/UK/IBM To: Mike Beckerle <mbeckerle.dfdl(a)gmail.com>, Cc: dfdl-wg(a)ogf.org, dfdl-wg-bounces(a)ogf.org Date: 19/03/2013 08:55 Subject: Re: [DFDL-WG] Action 206 - proposed timezone-from-dateTime function Mike This function may be used anywhere, and not just in the context of an element of calendar type. You therefore can't use the value of dfdl:calendarTimeZone. You should just return empty string if there is no timezone component. We also need 3 variations of the function, like XPath 2.0: dfdl:timeZoneFromDateTime($arg as xs:dateTime?) as xs:string? dfdl:timeZoneFromDate($arg as xs:date?) as xs:string? dfdl:timeZoneFromTime($arg as xs:time?) as xs:string? Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 From: Mike Beckerle <mbeckerle.dfdl(a)gmail.com> To: dfdl-wg(a)ogf.org, Date: 18/03/2013 18:29 Subject: [DFDL-WG] Action 206 - proposed timezone-from-dateTime function Sent by: dfdl-wg-bounces(a)ogf.org I examined use of ISO durations, but that introduces a bunch of unnecessary complexity for implementors. I decided to propose instead use of an ISO 8601Time zone designator as an xs:string. Proposed: dfdl:timeZoneFromDateTime($arg as xs:dateTime?) as xs:string? Summary: Returns the timezone component of $arg if any. If $arg has a timezone component, then the result is a string in the format of an ISO Time zone designator. Interpreted as an offset from UTC, its value may range from +14:00 to -14:00 hours, both inclusive. UTC is represented as "Z". If the $arg has no timezone component, then the value of the dfdl:calendarTimeZone property expressed as an ISO 8601 Time zone designator is returned, or ""(empty string) is returned if the dfdl:calendarTimeZone property is ""(empty string). Examples dfdl:timeZoneFromDateTime(xs:dateTime("1999-05-31T13:20:00-05:00")) returns the xs:string value "-05:00". dfdl:timeZoneFromDateTime(xs:dateTime("2000-06-12T13:20:00Z")) returns the xs:string value "Z". dfdl:timeZoneFromDateTime(xs:dateTime("2004-08-27T00:00:00")) with dfdl:calendarTimeZone='America/New_York' returns the xs:string "-05:00". ---------- Forwarded message ---------- From: Mike Beckerle <mbeckerle.dfdl(a)gmail.com> Date: Tue, Jan 29, 2013 at 9:05 AM Subject: Re: clarification needed: datetime and time To: dfdl-wg(a)ogf.org I also believe this means we are missing this function: fn:timezone-from-dateTime($arg as xs:dateTime?) as xs:dayTimeDuration? However, we don't have this result type dayTimeDuration, so rather than add that type, which is a structured thing like: PnDTnHnMnS, we should say it matches the iso 8601 time zone suffix on an iso datetime. So "Z" or "-05:00" etc. Since it is not exactly the existing Xpath function, perhaps this function should not have this name or not be in the fn namespace. Come to think of it. I don't think we ever say exactly what the fn: prefix is bound to in DFDL. On Tue, Jan 29, 2013 at 8:55 AM, Mike Beckerle <mbeckerle.dfdl(a)gmail.com> wrote: In the below. Should xs:datetime have trailing timezone as well? Right now it has no tz component. Seems to me it should to match iso 8601 format which is what XSD uses. Also, I believe we now think ZZZ is not the right behavior, and seems to be deprecated in ICU. The right thing is ZZZZZ. calendarPatternKind Enum Valid values ‘explicit’, ‘implicit’ ‘explicit’ means the pattern is given by dfdl:calendarPattern, ‘implicit’ means the pattern is derived from the XML schema date/time type. Logical Type Default Pattern xs:date yyyy-MM-dd xs:dateTime yyyy-MM-dd'T'HH:mm:ss xs:time HH:mm:ssZZZ Annotation: dfdl:element, dfdl:simpleType -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com -- dfdl-wg mailing list dfdl-wg(a)ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

2 3

OGF DFDL WG Call Minutes 2013-03-19
by Steve Hanson 19 Mar '13

19 Mar '13

Please find minutes from the above call on Redmine at http://redmine.ogf.org/dmsf_files/13054?download= Errata document is now at v012 here: http://redmine.ogf.org/dmsf_files/8104?download= Action 140 errata addendum at v001 here: http://redmine.ogf.org/dmsf_files/13017?download= Regards Steve Hanson Architect, IBM DFDL, Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848

1 0

proposed errata for argument and result types of expression language functions
by Mike Beckerle 19 Mar '13

19 Mar '13

Proposed Errata for expression language functions to add argument and result type information where it is not otherwise specified. Section 23.5.2.4 is amended to state: All the date/time functions return type xs:integer except fn:seconds-from-dateTime and fn:seconds-from-time are amended to specify that they return seconds (whole and fractional) as an xs:decimal value. Section 23.5.2 is amended to state that all the functions return type xs:boolean Section 23.5.6 function fn:local-name return type is changed to xs:string. (DFDL does not have type xs:NCName) Section 23.5.3 Functions dfdl:contentLength and dfdl:valueLength descriptions are amended to specify the type of the second argument is xs:string. Function dfdl:testBit is amended to specify that the types of the arguments are xs:unsignedByte and xs:integer respectively. Function dfdl:stringLiteralFromString is amended to specify that the return value of type xs:string *containing *a DFDL string literal. -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com

1 0

Action 206 - proposed timezone-from-dateTime function
by Mike Beckerle 19 Mar '13

19 Mar '13

I examined use of ISO durations, but that introduces a bunch of unnecessary complexity for implementors. I decided to propose instead use of an ISO 8601Time zone designator as an xs:string. Proposed: dfdl:timeZoneFromDateTime($arg as xs:dateTime?) as xs:string? Summary: Returns the timezone component of $arg if any. If $arg has a timezone component, then the result is a string in the format of an ISO Time zone designator. Interpreted as an offset from UTC, its value may range from +14:00 to -14:00 hours, both inclusive. UTC is represented as "Z". If the $arg has no timezone component, then the value of the dfdl:calendarTimeZone property expressed as an ISO 8601 Time zone designator is returned, or ""(empty string) is returned if the dfdl:calendarTimeZone property is ""(empty string). Examples - dfdl:timeZoneFromDateTime(xs:dateTime("1999-05-31T13:20:00-05:00"))returns the xs:string value "-05:00". - dfdl:timeZoneFromDateTime(xs:dateTime("2000-06-12T13:20:00Z")) returns the xs:string value "Z". - dfdl:timeZoneFromDateTime(xs:dateTime("2004-08-27T00:00:00")) with dfdl:calendarTimeZone='America/New_York' returns the xs:string "-05:00". ---------- Forwarded message ---------- From: Mike Beckerle <mbeckerle.dfdl(a)gmail.com> Date: Tue, Jan 29, 2013 at 9:05 AM Subject: Re: clarification needed: datetime and time To: dfdl-wg(a)ogf.org I also believe this means we are missing this function: fn:timezone-from-dateTime($arg as xs:dateTime?) as xs:dayTimeDuration? However, we don't have this result type dayTimeDuration, so rather than add that type, which is a structured thing like: PnDTnHnMnS, we should say it matches the iso 8601 time zone suffix on an iso datetime. So "Z" or "-05:00" etc. Since it is not exactly the existing Xpath function, perhaps this function should not have this name or not be in the fn namespace. Come to think of it. I don't think we ever say exactly what the fn: prefix is bound to in DFDL. On Tue, Jan 29, 2013 at 8:55 AM, Mike Beckerle <mbeckerle.dfdl(a)gmail.com>wrote: > In the below. Should xs:datetime have trailing timezone as well? Right now > it has no tz component. > > Seems to me it should to match iso 8601 format which is what XSD uses. > > Also, I believe we now think ZZZ is not the right behavior, and seems to > be deprecated in ICU. The right thing is ZZZZZ. > > calendarPatternKind > > Enum > > Valid values ‘explicit’, ‘implicit’ > > ‘explicit’ means the pattern is given by dfdl:calendarPattern, > > ‘implicit’ means the pattern is derived from the XML schema date/time type. > > *Logical Type * > > *Default Pattern* > > xs:date > > yyyy-MM-dd > > xs:dateTime > > yyyy-MM-dd'T'HH:mm:ss > > xs:time > > HH:mm:ssZZZ > > Annotation: dfdl:element, dfdl:simpleType > > > -- > Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | > www.tresys.com > > -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com

2 1

OGF DFDL WG Call Agenda 2013-03-19
by Steve Hanson 18 Mar '13

18 Mar '13

Please find agenda for call on Redmine at http://redmine.ogf.org/dmsf_files/13054?download= Regards Steve Hanson Architect, IBM Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848

1 0

unordered vs floating - identical?
by Garriss Jr., James P. 14 Mar '13

14 Mar '13

Are these functionally the same? #1 Sequence sequenceKind=unordered Element name=a Element name=b Element name=c /Sequence #2 Sequence sequenceKind=ordered Element name=a floating=yes Element name=b floating=yes Element name=c floating=yes /Sequence

2 1

Re: [DFDL-WG] ISO 8601 restricted profile and optional time zones
by Steve Hanson 12 Mar '13

12 Mar '13

Agreed today that the only change is for the 'I' symbol to allow time zone to be optional in the input data (always output when parsing). Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 From: Steve Hanson/UK/IBM To: Mike Beckerle <mbeckerle.dfdl(a)gmail.com>, Cc: dfdl-wg(a)ogf.org Date: 05/03/2013 14:11 Subject: Re: [DFDL-WG] ISO 8601 restricted profile and optional time zones Today there is nothing in our pattern language that allows fractional seconds and/or timezones to be optional. We don't have a [ ] syntax for example. I can imagine that being very useful. Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 From: Mike Beckerle <mbeckerle.dfdl(a)gmail.com> To: Steve Hanson/UK/IBM@IBMGB, Cc: dfdl-wg(a)ogf.org Date: 05/03/2013 13:23 Subject: Re: [DFDL-WG] ISO 8601 restricted profile and optional time zones Well, the definition of "I" can be whatever we want, but I suggest it follow the iso 8601 format "as practiced" in the real world. However, it can be strict or not about time zone, so long as there is some other calendar pattern which would let us parse "almost iso" data that lacks the timezone, fractional seconds, has one, or both of those. But that can be entirely decoupled from what "I" pattern character means. On Tue, Mar 5, 2013 at 5:56 AM, Steve Hanson <smh(a)uk.ibm.com> wrote: I'd like to discuss allowing time zone to be optional with the 'I' calendar pattern symbol. It looks like the ISO 8601 restricted profile insists that time zone appears if there are time components. See below. Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 ----- Forwarded by Steve Hanson/UK/IBM on 05/03/2013 10:52 ----- From: Steve Hanson/UK/IBM To: "Costello, Roger L." <costello(a)mitre.org>, Cc: David Hardcastle/UK/IBM@IBMGB, "Cranford, Jonathan W." < jcranford(a)mitre.org>, "Garriss Jr., James P." <jgarriss(a)mitre.org> Date: 05/03/2013 10:51 Subject: Re: MBTK doesn't seem to support xsd:time correctly Roger, XML Schema spec talks about what is allowed in the value space of the XML Schema types, ie, what DFDL creates in its infoset. When parsing, DFDL uses a calendar pattern to identify what calendar components appear in the data, using dfdl:calendarPatternKind and dfdl:calendarPattern properties. If the pattern contains fractional seconds symbols ('SSS'), then at least one fractional second must appear in the data else it is an error. If the pattern contains timezone symbols (eg, 'Z') then a time zone must appear in the data else it is an error. If you have data where the appearance of fractional seconds and/or time zone is optional, you might be able to use the calendar pattern 'I' symbol instead. This allows the data to match the time component of any of the 'granularities' in the ISO 8601 restricted profile ( http://www.w3.org/TR/NOTE-datetime) This will work if your data is of the form HH:mm[:ss[.SSS]][TZD]. Does your data match the ISO8601 restricted profile? Testing the 'I' symbol this today I think there is a bug in MBTK where it only allows either no fractional seconds at all or exactly 3 fractional seconds (it should allow any number). I will raise a defect. Also, although MBTK allows time zone to be missing for 'I', I am not sure this is permitted by the restricted profile which seems to insist on time zone being present. I will raise this with the DFDL Working Group, as it seems to me that allowing time zone to be optional is useful. Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848 From: "Costello, Roger L." <costello(a)mitre.org> To: David Hardcastle/UK/IBM@IBMGB, Steve Hanson/UK/IBM@IBMGB, Cc: "Cranford, Jonathan W." <jcranford(a)mitre.org>, "Garriss Jr., James P." <jgarriss(a)mitre.org> Date: 04/03/2013 18:24 Subject: MBTK doesn't seem to support xsd:time correctly Hello Steve and David, Here I declare a time element of type xsd:time: <xsd:element name="Time" type="xsd:time"/> According to the XML Schema specification the xsd:time datatype can optionally specify fractional digits, e.g.: 11:30:30.333 +0500 and timezone is optional: 11:30:30 However, if I use fractional digits I get an error and if I omit timezone I get an error. I looked at the DFDL spec and I don't see any mention about not allowing fractional seconds or requiring timezone. Is this a bug in MBTK? /Roger Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- dfdl-wg mailing list dfdl-wg(a)ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

1 0

OGF DFDL WG Call Minutes 2013-03-12
by Steve Hanson 12 Mar '13

12 Mar '13

Please find minutes from the above call on Redmine at http://redmine.ogf.org/dmsf_files/13034?download= Errata document is now at v012 here: http://redmine.ogf.org/dmsf_files/8104?download= Action 140 errata addendum at v001 here: http://redmine.ogf.org/dmsf_files/13017?download= (Note - the minutes for 2013-03-05 and 2013-02-26 have been updated to include a missing action 206) Regards Steve Hanson Architect, IBM DFDL, Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh(a)uk.ibm.com tel:+44-1962-815848

1 0

testBit function clarification needed
by Mike Beckerle 12 Mar '13

12 Mar '13

What types can the $data argument to testBit be? xs:unsignedByte is clear, but what other types if any? I think the stable design points are any of these: * unsignedByte only * unsignedByte and byte only * types xs:long and xs:unsignedLong and any of their sub-types. * any integer at all ...mike -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com

2 1