Next in thread →
Next in month →
Re: Integral types in OBIX wd36 schema and UML figure; schema and spec inconsistent
All --
Here are the three versions of the UML PNG diagram, one each with
the integral type being
int
integer
long
Craig, depending on the decision on the recommendation below, one
of these should work. Text still needs alignment, as does the
schema.
There's another alternative I hadn't considered:
only the val of Int is long, the others are all int.
Not sure that the potentially smaller sizes justify the
confusion...
Thanks!
bill
--
William Cox
Email:
Web: http://www.CoxSoftwareArchitects.com
+1 862 485 3696 mobile
On 10/21/14 5:01 PM, William Cox wrote:
Toby, Craig, and all --
This is modified from my 16 October email to the list. The most
recent UML diagram and schema use "int" for the integral types.
RECOMMENDATION:
Change all occurrences of "xs:int" in the schema to
"xs:long" consistent with line 516.
Correct capitalization on all of 4.3.1 and subsections -
note that the type on line 516 is "Int" not "int" as in the
text. Table at line 505 has incorrect capitalization in the
Type Name column.
Consistent with the spec, but does the change to xs:long
break anything?
ALTERNATE:
Change all occurrences of "xs:int" in the schema to
"xs:integer". This is a superset of "xs:long" and line 516
should be changed to something like "an integer with at
least 64 bits.
Correct capitalization on all of 4.3.1 and subsections -
note that the type on line 516 is "Int" not "int" as in the
text. Table at line 505 has incorrect capitalization in the
Type Name column. (same as above)
More mathematically elegant, but does it break
applications?
What do you prefer?
Thanks!
bill
INT INTEGER etc
"Int", "int", or "integer"?
We're using XSD types; I've added commas and [number of bits]
to the definitions for readability.
From the XML Schema Datatypes Second Edition:
[Definition:] integer is ·derived· from decimal by fixing the
value of ·fractionDigits· to be 0 and disallowing the trailing
decimal point. This results in the standard mathematical concept
of the integer numbers. The ·value space· of
integer is the infinite set {...,-2,-1,0,1,2,...}. The ·base
type· of integer is decimal.
[Definition:] long is ·derived· from integer by setting the
value of ·maxInclusive· to be 9,223,372,036,854,775,807 and
·minInclusive· to be -9,223,372,036,854,775,808. The ·base type·
of long is integer.
[Definition:] int is ·derived· from long by setting the value of
·maxInclusive· to be 2,147,483,647 and ·minInclusive· to
be -2,147,483,648. The ·base type· of int is long.
The schema uses xs:int, the third definition above - a 32 bit
integer. The UML diagram update I sent a few minutes ago uses
"int" everywhere for consistency.
So should it be "long" or "int" or
"integer"? Depends on what the schema users expect, and the
schema is definitive, so the spec should match the schema
which should match the spec :-)
For the OBIX type Int, note the type of the attributes min, max,
and val.
<xs:complexType name="Int">
<xs:complexContent>
<xs:extension base="Obj">
<xs:attribute name="min"
type="xs:int"/>
<xs:attribute name="max"
type="xs:int"/>
<xs:attribute ref="unit"/>
<xs:attribute name="val" type="xs:int"
default="0"/>
</xs:extension>
</xs:complexContent>
A related issue is on line 516:
ERROR "4.3.1.2 int" is really muddled. it says "val ...
maps to xs:long", NOT xs:int. So the
schema should say xs:long NOT xs:int. And so should the UML.
This is in part a carryover from the short/int/long issues in C.
Thanks!
bill
--
William Cox
Email:
Web: http://www.CoxSoftwareArchitects.com
+1 862 485 3696 mobile
Attachment:
Obj and Subclasses wd37 long.pdf
Description: Adobe PDF document
Attachment:
Obj and Subclasses wd37 integer.pdf
Description: Adobe PDF document
Attachment:
Obj and Subclasses wd37 int.pdf
Description: Adobe PDF document
Next in thread →
Next in month →