See also: IRC log
https://lists.w3.org/Archives/Public/public-xformsusers/2016Mar/0017.html
https://lists.w3.org/Archives/Public/public-xformsusers/2016Mar/0019.html
Steven: I had an action item to
research what "whitespace" means.
... see my three mails about it.
... The most inclusive seems to be to include the 25 characters
as defined by Unicode
Erik: We should use the most
inclusive possible
... but there may be some worth adding, such as ISO control
characters
Steven: Zero - 31
Nick: What if you enter 0-31,
those are not possible in XML
... I don't know which ones by heart
Erik: If it is not official XML,
are you still allowed to store it in an XML database? Good
question.
... There is nothing restricting characters in our run-time
<ebruchez> https://en.wikipedia.org/wiki/Control_character
Erik: Some of those are already included, such as tab and linefeed.
Steven: Form feed isn't in XML
Erik: stripping those is not hard.
https://en.wikipedia.org/wiki/Unicode_character_property#General_Category
https://en.wikipedia.org/wiki/Unicode_character_property#Whitespace
<nvdbleek> https://www.w3.org/TR/xml11/#sec-xml11
<ebruchez> http://stackoverflow.com/questions/730133/invalid-characters-in-xml
<ebruchez> "Basically, the control characters and characters out of the Unicode ranges are not allowed. This means also that calling for example the character entity  is forbidden."
Steven: We have to serialise instances as XML, so presumably an 0x3 would be refused at the receiving end
Erik: But there are other
encodings such as URL and JSON
... In our implementation we can handle such characters in the
data model
... I wonder if XQuery and so on allow such characters
... I can't find it at the moment
<nvdbleek> https://www.w3.org/TR/charmod/
Nick: The XDM refers to the character model 1.0
Steven: I propose we use this: https://en.wikipedia.org/wiki/Unicode_character_property#Whitespace
Erik: I can live with that.
Nick: So 'preserve' keeps characters, and might cause a problem?
Steven: That has always been the case in XForms; that is the current situation
Nick: Does that change the conformance of existing implementations?
Erik: I don't think we say that control characters are allowed or disallowed on input.
<nvdbleek> I propose rewording ‘preserve: all whitespace is preserved’ to ‘preserve: all whitespace __should be__ preserved’
Steven: I don't think this affects conformance of existing processors.
Nick: I would like to change some
wording in the text.
... I would like "should be" to be used in the definitions
Steven: It is only what is happening to whitespace, it doesn't affect other chatracters
Nick: Ah, OK!
Steven: Are you OK with this text "For
the amount value, by default, whitespace is preserved, but since the
schema lexical space for numbers allows leading and trailing whitespace,
such whitespace (but not embedded whitespace) will not effect validity."
Steven: last week you said that "space space 1 space space" was OK as input for a number
Erik: Well, it works
... but it would be good if the spec referenced that
correctly
... let me see if I can find it
... the schema spec doesn't mention leading spaces in the
lexical space
<ebruchez> https://www.w3.org/TR/xmlschema-2/#integer
Erik: so I don't know why it's working
Steven: [laughs]
... So who does that processing?
Erik: I'm not sure, but Philip
mentioned last week that schema checking of a number with
leading space was working fine.
... it doesn't seem to be in the lexical space, but maybe
something else is at work.
<ebruchez> "The input value is first converted to a value in the lexical space of the target type by applying the whitespace normalization rules for the target type (as defined in [XML Schema])."
<ebruchez> https://www.w3.org/TR/xmlschema-1/#d0e1654
Steven: How about "For the amount value, by default, whitespace is preserved, but leading and trailing whitespace,
does not effect validity "For the amount value, by default, whitespace is preserved, but leading and trailing whitespace,
does not effect validity "For the amount value, by default, whitespace is preserved, but note that leading and trailing whitespace,
does not effect validity."
Steven: How about "For the amount value, by default, whitespace is preserved, but leading and trailing whitespace, does not effect validity."
"For the amount value, by default, whitespace is preserved, but note that leading and trailing whitespace does not effect validity."
Erik: Not sure if that is (officially) true
Steven: Though it worked in two different places last week.
<ebruchez> https://www.w3.org/TR/xmlschema-2/#integer
Erik: I found it!
<ebruchez> "whiteSpace is applicable to all ·atomic· and ·list· datatypes. For all ·atomic· datatypes other than string (and types ·derived· by ·restriction· from it) the value of whiteSpace is collapse and cannot be changed by a schema author; for string the value of whiteSpace is preserve; for any type ·derived· by ·restriction· from string the value of whiteSpace can be any of the three legal values."
<ebruchez> https://www.w3.org/TR/xmlschema-2/#rf-whiteSpace
Steven: Yay!
... perfect!
Erik: We need to discuss the
processing model, at which time the processing occurs
... it is possible to link it to inputs, or recalculate
... We store data as a string.
[Discussion of how data is stored and manipulated]
Erik: Processing should occur at
recalc, or as if, you can optimise.
... we do the broader, during recalc
Nick: Couldn't you get loops then with string calculations?
Erik: No it recalcs it
once.
... You do the recalc, and then the whitespace.
Steven: We've run out of time; I'll update the text and resubmit; please comment.
[ADJOURN]
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/?/:/ Succeeded: s/alo/allo/ Succeeded: s/rime/time/ Succeeded: s/nt/not/ Succeeded: s/how/if/ Succeeded: s/1.1/1.0/ Succeeded: s/gn/ng/ Succeeded: s/one's/ones/ No ScribeNick specified. Guessing ScribeNick: Steven Inferring Scribes: Steven Present: Alain Erik Steven Nick Regrets: Philip Agenda: https://lists.w3.org/Archives/Public/public-xformsusers/2016Mar/0020 Found Date: 30 Mar 2016 Guessing minutes URL: http://www.w3.org/2016/03/30-forms-minutes.html People with action items:[End of scribe.perl diagnostic output]