OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Proposal for language tags according to RFC 4646


Michael Brauer already mentioned there are proposals to come, so here is
my proposal to enable support of any language/dialect/variant in any
script type in any region.

Currently, ODF sections 15.4.23 and 15.4.24 specify the fo:language and
fo:country attributes for style-text-properties-attlist. They refer the
W3C Extendible Stylesheet Language
http://www.w3.org/TR/2001/REC-xsl-20011015/slice7.html#country which in
turn also separated language/country but refers RFC 3066 [see
http://tools.ietf.org/html/rfc3066.html ] language tags where the
combined strings of those language tags don't necessarily result in
a language-country pair, but a sequence of primary-subtag and possibly
empty subtag(s) instead.

Similar to the style's fo:language and fo:country at least the following
are affected:

- "3.1.15 Language", the metadata <dc:language> element

- "7.8.1 Alphabetical Index Source", the fo:language and fo:country of

- "8.6.5 Sort", table:language and table:country of table-sort-attlist

- "14.7.2 Currency Style", number:language and number:country of

- "14.7.9 Common Data Style Attributes", number:language and
  number:country of common-data-style-attlist

- "14.9.3 Bibliography Configuration", fo:language and fo:country of

The current ODF specification is not sufficient as it doesn't allow
script types nor dialects nor variants. A short term solution for script
types could be to "inherit" the XSL script property
http://www.w3.org/TR/2001/REC-xsl-20011015/slice7.html#script but then
again that would not help for variants and may limit us on a long term.

When referring the XSL 1.0 language property for language attributes and
strictly interpreting it according to RFC 3066 we would be formally
bound to the language tags registered with IANA and ISO 639-1 and 639-2
codes, not even the upcoming ISO 639-3, now FDIS, would be possible.
Strictly spoken. OOo already supports languages that are covered by
ISO/FDIS 639-3 only.

I herewith propose full support of RFC 4646 language tags
http://tools.ietf.org/html/rfc4646 that obsoletes RFC 3066.

language ["-" script] ["-" region] *("-" variant) *("-" extension) ["-" privateuse]

The question now is how to do it in the schema. We could add script
attributes similar to language and country, as in

  <attribute name="fo:script">
    <ref name="scriptCode"/>

and the same for privateuse, but how to specify the
zero-or-more-occurrences of variant and extension?

Would we better be off with a full RFC 4646 language tag _string_?
Sounds ugly, but would ease things.

For the ongoing work of the Language Tag Registry Update (ltru) see also
http://www.ietf.org/html.charters/ltru-charter.html and RFC 4646bis


 OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]