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

Rescheduled: OGF DFDL WG Extra Call (24 Jan 16:00 GMT in No room)
by Steve Hanson 23 Jan '13

23 Jan '13

OGF DFDL WG weekly call dial-in details. Passcode for Participants: 5381214 Canada Toll-Free 888-426-6840 China Toll-Free 10-800-711-1071 CHINA NETCOM GROUP USERS China Toll-Free 10-800-110-0996 CHINA TELECOM SOUTH USERS France Toll-Free 0800-94-0558 Germany Toll-Free 0800-000-1018 India Toll-Free 000-800-100-1176 Ireland Toll-Free 1-800-943-427 Israel Toll-Free 1-809-417-783 United Kingdom Caller Paid 0-20-30596451 United Kingdom Toll-Free 0800-368-0638 USA Caller Paid 215-861-6239 USA Toll-Free 888-426-6840 Other international numbers available - e-mail smh(a)uk.ibm.com. OGF DFDL Home: http://www.ogf.org/dfdl Redmine DFDL: http://redmine.ogf.org/projects/dfdl-wg Meeting was 10am/3pm but only Tim and Steve were on the call

1 0

Clarification when calendarPattern is a superset of the type
by Steve Lawrence 22 Jan '13

22 Jan '13

Assume we have the following element: <xs:element name='date' type='xs:date' dfdl:calendarPattern='yyyy-MM-dd hh:mm' /> So we have a element of type xs:date, but a calendarPattern that specifies date and time formatting symbols. What should happen in this case? Two potential options that I see (there are probably others): 1) Throw some kind of error, signifying that calendarPattern contains formatting symbols not allowed for the type. 2) Truncate the parsed data to fit the type, so the data 2013-01-22 11:59 becomes <date>2013-01-22</date> This seems reasonable, however the spec mentions the following: When unparsing, and property calendarPattern contains a formatting symbol that requires a component of the date/time and the infoset value does not contain that component, it is a processing error. So that means it is impossible to unparse when the calendar pattern is a superset of the element type. Is this ok? Thoughts? - Steve

2 1

Broadcast: OGF DFDL WG Extra Call (23 Jan 15:00 GMT in No room)
by Steve Hanson 22 Jan '13

22 Jan '13

OGF DFDL WG weekly call dial-in details. Passcode for Participants: 5381214 Canada Toll-Free 888-426-6840 China Toll-Free 10-800-711-1071 CHINA NETCOM GROUP USERS China Toll-Free 10-800-110-0996 CHINA TELECOM SOUTH USERS France Toll-Free 0800-94-0558 Germany Toll-Free 0800-000-1018 India Toll-Free 000-800-100-1176 Ireland Toll-Free 1-800-943-427 Israel Toll-Free 1-809-417-783 United Kingdom Caller Paid 0-20-30596451 United Kingdom Toll-Free 0800-368-0638 USA Caller Paid 215-861-6239 USA Toll-Free 888-426-6840 Other international numbers available - e-mail smh(a)uk.ibm.com. OGF DFDL Home: http://www.ogf.org/dfdl Redmine DFDL: http://redmine.ogf.org/projects/dfdl-wg

1 0

Fw: date & time and latest ICU possible issues/conflicts
by Steve Hanson 22 Jan '13

22 Jan '13

Clarification from CLDR via ICU on the single 'V' re-use issue: "V" was originally designed for a variant of pattern "z" in CLDR. It had a valid reason to define a pattern different from "z", but we found that CLDR cannot collect such information differentiating "V" from "z". Also, there were no users who requested us (including ICU) to maintain proper data for differentiating "V" from "z". As the result, the data making the difference between "V" and "z" was deprecated, and we no longer have reason to keep pattern "V" except backward compatibility reason. So, strictly speaking, it's a backward compatibility problem. But - It still produces a text representation of the same calendar field. "V" was not adopted by any other known implementations other than ICU. The distinction between "V" and "z" had never worked as users would expect. "z" is a Java compatible pattern for short abbreviated zone format, and it is available from the very beginning. On the other hand, "V" was a recent addition. CLDR technical committee including myself understand the risk, but all of us preferred cleaner pattern definition over the potential backward compatibility problem. Sound like DFDL should withdraw 'V', I doubt that any IBM customer will be using it. 'VVVV' is ok. 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 21/01/2013 18:17 ----- From: Steve Hanson/UK/IBM To: dfdl-wg(a)ogf.org, Date: 17/01/2013 10:04 Subject: Fw: [DFDL-WG] date & time and latest ICU possible issues/conflicts The ICU ticket has been answered, with reference to the following document: http://cldr.unicode.org/development/development-process/design-proposals/ti… 'x'/'X' symbols In a not-too-distant ICU there will be new symbols 'x' and 'X' to handle Z as a time zone for ISO8601 date/times, effectively replacing 'ZZZZZ'. Both 'x' and 'X' will tolerate Z or +00:00 (or variants) on parsing, and on formatting X will result in Z and x will result in +00:00 (or variants). DFDL's use of 'U' is wider than this, as we allow 'U' to appear with any number of 'Z's, meaning that Z is accepted with non-ISO8601 date/times. DFDL also adds the use of 'I' symbol on its own to mean any ISO8601 compliant date/time, and again we allow 'U' to appear with 'I'. However the motivating use case for adding 'U' was IBM MRM which supports this today. But it does so primarily for XML use cases, in particular ISO8601. I am not personally aware of an actual non-XML use case. I suggest we drop the DFDL-specific use of the 'U' symbol in conjunction with 'Z' and 'I' symbols from the DFDL specification via errata, and allow the use of 'ZZZZZ' instead, which at least will accept Z when parsing. When 'x'/'X' support appears in ICU, we can take a future errata to support it or leave until DFDL 2.0. IBM DFDL already supports 'U' but I am ok with deprecating it as I don't believe it will be being used for real. 'V' symbol In a not-too-distant ICU there will be new symbols 'VV' and 'VVV' to handle time zones expressed as Time Zone Ids and localized locations, respectively. We can add that via errata in the future, or leave until DFDL 2.0. However, at the same time the meaning of V is changed slightly. DFDL supports 'V'. I have asked ICU for a clarification. 'O' symbol In a not-too-distant ICU there will be new symbol 'O' to handle localized GMT format variants. We can add that via errata in the future, or leave until DFDL 2.0. 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 17/01/2013 08:58 ----- From: Steve Hanson/UK/IBM To: Mike Beckerle <mbeckerle.dfdl(a)gmail.com>, Cc: dfdl-wg(a)ogf.org Date: 16/01/2013 18:20 Subject: Re: [DFDL-WG] date & time and latest ICU possible issues/conflicts ICU ticket raised as the help does not give an example. https://icu.sanjose.ibm.com/gcoctrac/ticket/469#ticket 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: 16/01/2013 15:11 Subject: [DFDL-WG] date & time and latest ICU possible issues/conflicts Sent by: dfdl-wg-bounces(a)ogf.org Steve Lawrence is on the Daffodil Open Source DFDL team (on CC), and he has dug into date/time types. He raised some concerns to me that I really haven't been tracking at all, so I wanted to put in front of the rest of the group. The date time format syntax for the latest version of icu4j contains a 'U' character, which means "cyclic year name". However, the daffodil spec says the 'U' character, following a Z makes it so a timezone of UTC is represented as Z instead of +00:00. This seems to be a conflict, and would prevent us from ever upgrading to the newest version of ICU (which might be a good idea). I will point out that the latest version of ICU supports ZZZZZ (5 Z's), which is the ISO8601 timezone format. This doesn't add all the functionality that the DFDL 'U' gives. My question is, is this enough? Are there cases where the ZU, ZZU, etc. are necessary? I'm just concerned that the U is going to quite a bit more complexity, and want to make sure the updates to latest ICU don't address the DFDL-WG concerns. And if we still need the 'U', maybe it should change to a different letter to prevent conflicts with the latest ICU4J? I would point out that the ICU pattern language cannot deal with dual-purpose letters very well, i.e., ambiguities are introduced if the same letter both introduces a format, and if following another format string, modifies its behavior. E.g., does ZU mean Z modified by U, or Z first, and then U. So it seems pretty unfortunate if the ICU libraries added a conflicting use of letter U. I believe the point of DFDL's use of the U modifier for letters I and Z was to be absolutely clear on the GMT timezone 'Z' issue, i.e., to indicate that 'Z' is to be used, and -00:00 is not to be output, nor accepted when parsing. The ICU specification ZZZZZ says ISO format, but that allows either 'Z' or -00:00 to be used for GMT timezone, and it's not clear what it means on output. -- 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 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 1

OGF DFDL WG Call Agenda 2013-01-22
by Steve Hanson 21 Jan '13

21 Jan '13

Please find agenda for call on Redmine at http://redmine.ogf.org/dmsf_files/12991?download= Please note that all 2012 calls have been moved to Archive - 2012 folder ( http://redmine.ogf.org/dmsf/dfdl-wg?folder_id=5486) 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

OGF DFDL WG Call Minutes 2013-01-15
by Steve Hanson 17 Jan '13

17 Jan '13

Please find minutes from the above call on Redmine at http://redmine.ogf.org/dmsf_files/12987?download= Errata document is now at v011 (draft) here: http://redmine.ogf.org/dmsf_files/8104?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

Fw: date & time and latest ICU possible issues/conflicts
by Steve Hanson 17 Jan '13

17 Jan '13

The ICU ticket has been answered, with reference to the following document: http://cldr.unicode.org/development/development-process/design-proposals/ti… 'x'/'X' symbols In a not-too-distant ICU there will be new symbols 'x' and 'X' to handle Z as a time zone for ISO8601 date/times, effectively replacing 'ZZZZZ'. Both 'x' and 'X' will tolerate Z or +00:00 (or variants) on parsing, and on formatting X will result in Z and x will result in +00:00 (or variants). DFDL's use of 'U' is wider than this, as we allow 'U' to appear with any number of 'Z's, meaning that Z is accepted with non-ISO8601 date/times. DFDL also adds the use of 'I' symbol on its own to mean any ISO8601 compliant date/time, and again we allow 'U' to appear with 'I'. However the motivating use case for adding 'U' was IBM MRM which supports this today. But it does so primarily for XML use cases, in particular ISO8601. I am not personally aware of an actual non-XML use case. I suggest we drop the DFDL-specific use of the 'U' symbol in conjunction with 'Z' and 'I' symbols from the DFDL specification via errata, and allow the use of 'ZZZZZ' instead, which at least will accept Z when parsing. When 'x'/'X' support appears in ICU, we can take a future errata to support it or leave until DFDL 2.0. IBM DFDL already supports 'U' but I am ok with deprecating it as I don't believe it will be being used for real. 'V' symbol In a not-too-distant ICU there will be new symbols 'VV' and 'VVV' to handle time zones expressed as Time Zone Ids and localized locations, respectively. We can add that via errata in the future, or leave until DFDL 2.0. However, at the same time the meaning of V is changed slightly. DFDL supports 'V'. I have asked ICU for a clarification. 'O' symbol In a not-too-distant ICU there will be new symbol 'O' to handle localized GMT format variants. We can add that via errata in the future, or leave until DFDL 2.0. 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 17/01/2013 08:58 ----- From: Steve Hanson/UK/IBM To: Mike Beckerle <mbeckerle.dfdl(a)gmail.com>, Cc: dfdl-wg(a)ogf.org Date: 16/01/2013 18:20 Subject: Re: [DFDL-WG] date & time and latest ICU possible issues/conflicts ICU ticket raised as the help does not give an example. https://icu.sanjose.ibm.com/gcoctrac/ticket/469#ticket 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: 16/01/2013 15:11 Subject: [DFDL-WG] date & time and latest ICU possible issues/conflicts Sent by: dfdl-wg-bounces(a)ogf.org Steve Lawrence is on the Daffodil Open Source DFDL team (on CC), and he has dug into date/time types. He raised some concerns to me that I really haven't been tracking at all, so I wanted to put in front of the rest of the group. The date time format syntax for the latest version of icu4j contains a 'U' character, which means "cyclic year name". However, the daffodil spec says the 'U' character, following a Z makes it so a timezone of UTC is represented as Z instead of +00:00. This seems to be a conflict, and would prevent us from ever upgrading to the newest version of ICU (which might be a good idea). I will point out that the latest version of ICU supports ZZZZZ (5 Z's), which is the ISO8601 timezone format. This doesn't add all the functionality that the DFDL 'U' gives. My question is, is this enough? Are there cases where the ZU, ZZU, etc. are necessary? I'm just concerned that the U is going to quite a bit more complexity, and want to make sure the updates to latest ICU don't address the DFDL-WG concerns. And if we still need the 'U', maybe it should change to a different letter to prevent conflicts with the latest ICU4J? I would point out that the ICU pattern language cannot deal with dual-purpose letters very well, i.e., ambiguities are introduced if the same letter both introduces a format, and if following another format string, modifies its behavior. E.g., does ZU mean Z modified by U, or Z first, and then U. So it seems pretty unfortunate if the ICU libraries added a conflicting use of letter U. I believe the point of DFDL's use of the U modifier for letters I and Z was to be absolutely clear on the GMT timezone 'Z' issue, i.e., to indicate that 'Z' is to be used, and -00:00 is not to be output, nor accepted when parsing. The ICU specification ZZZZZ says ISO format, but that allows either 'Z' or -00:00 to be used for GMT timezone, and it's not clear what it means on output. -- 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

1 0

date & time and latest ICU possible issues/conflicts
by Mike Beckerle 16 Jan '13

16 Jan '13

Steve Lawrence is on the Daffodil Open Source DFDL team (on CC), and he has dug into date/time types. He raised some concerns to me that I really haven't been tracking at all, so I wanted to put in front of the rest of the group. The date time format syntax for the latest version of icu4j contains a 'U' character, which means "cyclic year name". However, the daffodil spec says the 'U' character, following a Z makes it so a timezone of UTC is represented as Z instead of +00:00. This seems to be a conflict, and would prevent us from ever upgrading to the newest version of ICU (which might be a good idea). I will point out that the latest version of ICU supports ZZZZZ (5 Z's), which is the ISO8601 timezone format. This doesn't add all the functionality that the DFDL 'U' gives. My question is, is this enough? Are there cases where the ZU, ZZU, etc. are necessary? I'm just concerned that the U is going to quite a bit more complexity, and want to make sure the updates to latest ICU don't address the DFDL-WG concerns. And if we still need the 'U', maybe it should change to a different letter to prevent conflicts with the latest ICU4J? I would point out that the ICU pattern language cannot deal with dual-purpose letters very well, i.e., ambiguities are introduced if the same letter both introduces a format, and if following another format string, modifies its behavior. E.g., does ZU mean Z modified by U, or Z first, and then U. So it seems pretty unfortunate if the ICU libraries added a conflicting use of letter U. I believe the point of DFDL's use of the U modifier for letters I and Z was to be absolutely clear on the GMT timezone 'Z' issue, i.e., to indicate that 'Z' is to be used, and -00:00 is not to be output, nor accepted when parsing. The ICU specification ZZZZZ says ISO format, but that allows either 'Z' or -00:00 to be used for GMT timezone, and it's not clear what it means on output. -- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com

2 1

3.13 on encoding errors - rewording - was Re: Second draft of DFDL Errata v011
by Mike Beckerle 16 Jan '13

16 Jan '13

*Re: Errata v11 3.13 I concluded that this needs a rewrite entirely, and that the policy we adopt has to allow for a reasonable implementation of scanning that doesn't require re-invention of the entire world of character I/O. I believe no changes are needed to section 4.1.2, so this errata is not about that section anymore, only about section 11. Suggested revised errata 3.13 text follows:* The encodingErrorPolicy property specified in Errata v10 is removed. A new sub-section is added to section 11. *(this is probably 11.2, if 11.1 is about Unicode byte order marks)* 11.2 Character Encoding and Decoding Errors When parsing, these are the errors that can occur when decoding characters into Unicode/ISO 10646. 1. The data is broken - invalid bit/byte sequences are found which do not match the definition of a character for the encoding. 2. Not enough data is found to make up the entire encoding of a character. That is, a fragment of a valid encoding is found. When unparsing, these are the errors that can occur when encoding characters from Unicode/ISO 10646 into the specified encoding. 1. No mapping provided by the encoding specification. 2. Not enough room to output the entire encoding of the character (e.g., need 3 bytes for a character encoding that uses 3-bytes for that character, but only 1 byte remains in the available length. The subsections below describe how these errors are handled. 11.2.1 Parsing: Decoding Errors When Scanning Scanning is searching of text for data matching: - a delimiter (this can be with or without an associated escape scheme) - initiators, terminators, and separators - a literal nil value (nilKind of 'literalValue' or 'literalCharacter') - a regular expression - for lengthKind='pattern' - for dfdl:assert or dfdl:discriminator with testKind='pattern' In all cases, if scanning encounters a decode error, then it stops with the the successfully decoded character immediately preceding the decode-error. This means that neither framing nor content that is scanned can ever encompass data that caused decode errors. It is worth specific note that the "." regular expression character which matches any character, does not allow a regular expression to subsume a decode error. It is also worth noting that the Unicode Replacement Character (U+FFFD) can appear in data as an ordinary character, but the scanning process will never create these from decode-errors. 11.2.2 Parsing: Decode Errors in Elements of Specified Length For elements with specified lengths (see glossary), the content region is processed and decode errors in conversion of this content region to text will result in the replacement of the non-decodable part of the data by one or more Unicode replacement characters (U+FFFD). If lengthUnits='characters', then such a Unicode replacement character counts as contributing a single character to the length. The trimming of padding characters happens after this replacement. 11.2.2.2 Parsing: The Not-Enough-Data Decode Error There is one special case for the 'not enough data' decode error. For lengthUnits='bytes' when the encoding is a fixed-width character set (see section 12.3.7.1.1 Character Width). If the number of bytes is not a multiple of the character set width, then there will be some number of bytes left over at the end of the data which are insufficient to hold an entire character code. In this case no attempt is made to decode a character from these left-over bytes. They are skipped when parsing (and filled with the dfdl:fillByte on unparsing). 11.2.3 Parsing: Unicode Decoding Non-Errors The following specific situations involving UTF-16, UTF-16LE, and UTF-16BE and do not cause a decoding or encoding error. • UTF-16 and unpaired surrogate code-point • UTF-16 and out-of-order surrogate code-point pair • UTF-16 encoding utf16Width='fixed' and a surrogate code point is encountered • UTF-16 byte-order-marks found not at the beginning of the data In all these cases the code-point becomes a character code in the DFDL Information Item for the string. 11.2.4 Unparsing: Substitution Character For unparsing, each encoding has a replacement/substitution character specified by the ICU. This character is substituted for the unmapped character or the character that has too large an encoding. The definitions of these substitution characters can be conveniently found for many encodings in the ICU Converter Explorer ( http://demo.icu-project.org/icu-bin/convexp) An encoding error is a processing error if the encoding does not provide a substitution/replacement character definition. (This would be rare, but could occur if a DFDL implementation allows many encodings beyond the minimum set.) 11.2.4.1 Unparsing: The Not-Enough-Room Encoding Error There is one special case for the 'not enough room' encoding error. For lengthUnits='bytes' when the encoding is a fixed-width character set (see section 12.3.7.1.1 Character Width). If the number of bytes is not a multiple of the character set width, then there will be some number of bytes of space left over at the end of the data which are insufficient to hold an entire character code. In this case no attempt is made to encode a character into these left-over bytes. They are filled with the dfdl:fillByte. (On parsing they are skipped.) 11.2.5 Preserving Data Containing Decoding Errors There can be situations where data wants to be preserved exactly even if it contains errors. It is suggested that if a DFDL schema author wants to preserve information containing data where the data may have decoding of errors, that they model such data as xs:hexBinary, or as xs:string but using an encoding such as iso-8859-1 which preserves all bytes.

1 2

Action 191 - Clarify what constitutes Schema Definition vs. Processing Error
by Mike Beckerle 15 Jan '13

15 Jan '13

Attached is a document about Action 191 which is a list of specfic errors. -- Mike Beckerle | OGF DFDL WG Co-Chair | Tresys Technologies Tel: 781-330-0412

1 2