[Fwd: (j3.2006) Fortran 2003 compiler features]

[Tobias Burnus]

Hi,

is the following correct? Suggestions?

For the questions, cf.
http://www.kcl.ac.uk/kis/support/cit/fortran/john_reid_new_2003.pdf

Tobias



> The following is sent on behalf of Ian Chivers.  Please reply
> to him at ian.chivers at chiversandbryan.co.uk, not to me.
> --------------------------------------------------------------
>
> I am interested in providing a list of compilers that
> implement one or more features of the 2003 standard.
>
> I want to include this in the April edition of Fortran Forum
> and I'll also add it to the Fortran resource file that Jane and
> I have up at king's.
>
> I've taken the list of features from an article
> by John Reid. I've added the two TRs, 15580 and 15581.
> Should they be there on their own?
>
> I would like each of the compiler vendors to
> provide details of what they implement -
> maybe full, partial, as an option?
>
> If they think there are things that are missing
> I will gladly add them.
>
> I've attached a spreadsheet with the list of features and
> the compiler vendors.
>
> If people could email me with the completed spreadsheet
> I'll pull  things togethor and send copies back to people
> before the article goes to print.


> Fortran 2003 Compiler Features
> *Data enhancements and object orientation*
> Parameterized derived types
>   
NO

> Procedure pointers
>   
NO

> Finalization
>   
NO

> Procedures bound by name to a type
>   
NO

> The PASS attribute
>   
NO

> Procedures bound to a type as operators
>   
NO

> Type extension
>   
NO

> Overriding a type-bound procedure
>   
NO

> Enumerations
>   
YES 4.1

> ASSOCIATE construct
>   
NO
> Polymorphic entities
>   
NO

> SELECT TYPE construct
>   
NO

> Deferred bindings and abstract types
>   
NO

> *Miscellaneous enhancements*
> Structure constructors
>   
NO

> The allocate statement
>   
YES 4.2

> Assignment to an allocatable array
>   
YES, since 4.2

> Transferring an allocation
>   
YES, since 4.2

> More control of access from a module
>   
YES since 4.1

> Renaming operators on the USE statement
>   
YES since 4.1

> Pointer assignment
>   
NO

> Pointer INTENT
>   
YES since 4.3

> The VOLATILE attribute
>   
YES since 4.3

> The IMPORT statement
>   
YES since 4.3

> Intrinsic modules
>   
YES since 4.3

> Access to the computing environment
>   
YES since 4.3

> Support for international character sets
>   
NO

> Lengths of names and statements
>   
YES since 4.1

> "Binary, octal and hex constants"
>   
NO.
(REAL BOZ are not supported, PR29471, PR18026)

> Array constructor syntax
>   
YES since 4.1

> Specification and initialization expressions
>   
NO
(PR29962)

> Complex constants
>   
YES since 4.1

> Changes to intrinsic functions
>   
NO

> Controlling IEEE underflow
>   
NO

> Another IEEE class value
>   
NO

> *Input/output enhancements*
> Derived type input/output
>   
NO

> Asynchronous input/output
>   
NO

> FLUSH statement
>   
YES 4.1
> IOMSG= specifier
>   
YES 4.1

> Stream access input/output
>   
YES 4.2

> ROUND= specifier
>   
NO

> DECIMAL= specifier
>   
NO

> SIGN= specifier
>   
NO

> Kind type parameters of integer specifiers
>   
NO

> Recursive input/output
>   
(don't know)

> Intrinsic function for newline character
>   
YES since 4.2

> Input and output of IEEE exceptional values
>   
YES 4.1

> Comma after a P edit descriptor
>   
YES 4.1

> *Interoperability with C*
> Interoperability of intrinsic types
>   
Only in Fortran-Experiments branch

> Interoperability with C pointers
>   
Only in Fortran-Experiments branch

> Interoperability of derived types
>   
Only in Fortran-Experiments branch

> Interoperability of variables
>   
Only in Fortran-Experiments branch

> Interoperability of procedures
>   
Only in Fortran-Experiments branch

> Interoperability of global data
>   
Only in Fortran-Experiments branch

> ? ISO TR 15580 IEEE Arithmetic
>   
NO

> ? ISO TR 15581 Allocatable Enhancements
>   
YES 4.2