www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - Compiler says cast then when done says no can do

reply Ty Tower <towerty msn.com.au> writes:
I have this line

double balance=values[6]-values[5]; (values 5&6 are char[])

so I change to 
double balance=-(double)values[6]-(double)values[5]; 
compiler says : C style cast illegal, use cast(double)values[6]

so I change to 
double balance=cast(double)values[6]-cast(double)values[5]; 
compiler says: Error: e2ir: cannot cast from char[] to double

So as I see it the compiler message in this case at least is wrong or there is
a problem casting from a char array to a double . Is it a problem with the
compiler message?
May 21 2008
next sibling parent Mike James <null erehwon.com> writes:
Ty Tower Wrote:

 I have this line
 
 double balance=values[6]-values[5]; (values 5&6 are char[])
 
 so I change to 
 double balance=-(double)values[6]-(double)values[5]; 
 compiler says : C style cast illegal, use cast(double)values[6]
 
 so I change to 
 double balance=cast(double)values[6]-cast(double)values[5]; 
 compiler says: Error: e2ir: cannot cast from char[] to double
 
 So as I see it the compiler message in this case at least is wrong or there is
a problem casting from a char array to a double . Is it a problem with the
compiler message?
 
 
I've just tried it - no problems. My compiler version is 1.29. Maybe try cast(double)(values[5]).
May 21 2008
prev sibling next sibling parent reply "Koroskin Denis" <2korden gmail.com> writes:
On Wed, 21 May 2008 15:25:48 +0400, Ty Tower <towerty msn.com.au> wrote:

 I have this line

 double balance=values[6]-values[5]; (values 5&6 are char[])

 so I change to
 double balance=-(double)values[6]-(double)values[5];
 compiler says : C style cast illegal, use cast(double)values[6]

 so I change to
 double balance=cast(double)values[6]-cast(double)values[5];
 compiler says: Error: e2ir: cannot cast from char[] to double

 So as I see it the compiler message in this case at least is wrong or  
 there is a problem casting from a char array to a double . Is it a  
 problem with the compiler message?
What do you want to get by casting from char[] to double? if `values[5]` stores something like "3.14" then cast won't return what you expect, consider using something like double t = to!(double)("3.14"); or toDouble("2.73") Both are defined in std.conv
May 21 2008
parent Tower Ty <towerty msn.com.au> writes:
Koroskin Denis Wrote:
 On Wed, 21 May 2008 15:25:48 +0400, Ty Tower <towerty msn.com.au> wrote:
 I have this line

 double balance=values[6]-values[5]; (values 5&6 are char[])

 so I change to
 double balance=-(double)values[6]-(double)values[5];
 compiler says : C style cast illegal, use cast(double)values[6]

 so I change to
 double balance=cast(double)values[6]-cast(double)values[5];
 compiler says: Error: e2ir: cannot cast from char[] to double

 So as I see it the compiler message in this case at least is wrong or  
 there is a problem casting from a char array to a double . Is it a  
 problem with the compiler message?
What do you want to get by casting from char[] to double? if `values[5]` stores something like "3.14" then cast won't return what you expect, consider using something like double t = to!(double)("3.14"); or toDouble("2.73") Both are defined in std.conv
dmd 1.024 I use Tango to!(double) compiles OK but toDouble does not The char{} values has a changeable string like "300.00" in 5 6 & 7 I tried double olddebit = Float.parse(values[5]); double oldcredit = Float.parse(values[6]); double oldbalance = Float.parse(values[7]); I change the amounts here then double balance = oldbalance + olddebit - oldcredit - to!(double)(values[5]) + to!(double)(values[6]); values[7]= to!(char[])(balance); which compiles but the result is nan (not a number I assume) double olddebit = to!(double)(values[5]); double oldcredit = to!(double)(values[6]); double oldbalance = to!(double)(values[7]); double balance = oldbalance + olddebit - oldcredit - to!(double)(values[5]) + to!(double)(values[6]); values[7]= to!(char[])(balance); this also compiles but gives nan ?
May 21 2008
prev sibling parent "Steven Schveighoffer" <schveiguy yahoo.com> writes:
"Ty Tower" wrote
I have this line

 double balance=values[6]-values[5]; (values 5&6 are char[])

 so I change to
 double balance=-(double)values[6]-(double)values[5];
 compiler says : C style cast illegal, use cast(double)values[6]

 so I change to
 double balance=cast(double)values[6]-cast(double)values[5];
 compiler says: Error: e2ir: cannot cast from char[] to double

 So as I see it the compiler message in this case at least is wrong or 
 there is a problem casting from a char array to a double . Is it a problem 
 with the compiler message?
The error might be in the order of operations. I think the compiler is casting your variable, then applying the index operation. So if you add the appropriate parentheses for default order of precedence, it looks like: (cast(double)values)[6] ... So what you really want is: cast(double)(values[6]) - cast(double)(values[5]); or what might be simpler: cast(double)(values[6] - values[5]); to force the order of operations. I have no idea what the proper order of operations is, so this all might be BS, but it's almost impossible to determine the correct order for normal people (i.e. people who don't write/read grammar expressions regularly). However, the order of operations is sort of built into the grammar on http://www.digitalmars.com/d/2.0/expression.html What makes sense to me is the index operator should be applied before the cast, but apparently that's not the case here... -Steve
May 27 2008