Conference turris::c_plus_plus

3487.0. "Vector, template matches no __pl??" by KERNEL::PULLEY (Come! while living waters flow) Wed Mar 12 1997 13:52

Hi,

I've a customer using Digital UNIX v3.2d, C++ v5.5.
I'm just starting to get to grips with C++, so if there's a better title
for this please let me know.
Do I need the code to take this any further?

Below is what the customer has to say.

Thanks,
Steve.

Here is the compilation statement and error message I am getting
from my perfectly good code :

SOURCE=fields/Fields/patchFieldFields/patchFieldFields.C ;
cxx -verbose -DdecAlpha -DscalarMachine -DoptSolvers
-I/usr/users/henry/c++/foam1.7/src/foam/include -Iinclude -I.    -c $SOURCE
-o Make/decAlpha/patchFieldFields.o
cxx: Error:
/usr/users/henry/c++/foam1.7/src/foam/include/FieldMemberOperators.C, line 23:
In this statement, the argument list "((f1)[i])+((f2)[i])" matches no "__pl".
    FOR_ALL_F_OP_F_OP_F(Type, *this, =, Type, f1, +, Type, f2,
----^

The class it is complaining about is tensorField which is properly
declared and does have a __pl operator.  This same piece of code 
compiles fine for vectorField.  The only difference is that vector
is Vector<double> and tensor is a Vector<Vector<double > >, so the
DEC compiler seems to be getting confused by the use of templates.

Can you post either a short code example that demonstrates the problem or a
pointer to the source on the net? I'm not sure there is enough to go on given
the base note info.

If there are user-defined #include files, you can preprocess the user's code
using -P and then reference the .i file produced.

Thanks,
-Jeff

    Sory folks,
    
    I'm still trying to get some source for this.